Ahoy! It’s been a while, matey! How are ye? Slowly gettin’ back t’ normal, I hope? That’s kinda happenin’ here, sort o’. 😛 I’ve started workin’ from the cabin again instead o’ at home. I definitely prefer workin’ at home, as it’s quieter, I don’t ‘ave t’ walk or get the bus, ‘n unlike the cabin galley, me galley has more than jus’ a microwave. So why did I go back t’ the cabin? Basically, ’cause I be nah rich. 😛 I live in a wee flat ‘n ‘ave no dedicated area fer work, so I had t’ ‘ave another keyboard ‘n mouse on the same desk, ‘n another computer by me right leg, ‘n a Mac on me black brew table. ‘n an iPad lying around somewhere jus’ in case. ‘n an Apple TV plugged into a 4K monitor in the ship me TV used t’ be, afore ’twas demoted t’ the floor. ‘n the work PC was wired into me existin’ monitor setup, so I had t’ switch monitor sources every time I needed t’ use a diff’rent computer. So that’s one advantage t’ the cabin: I don’t ‘ave t’ a ton o’ extra crap in me wee flat. 😛 I totally would’ve kept workin’ at home if I had a bigger ship, but I don’t ‘ave that kinda booty. 😛 It’s also nice to be able to jus’, ye know, ask scallywags stuff usin’ me mouth. Instead o’ composin’ a message on Slack ‘n sometimes havin’ to wait about an hour fer a response. ‘n even though it’s quieter at home, I actually found it easier t’ focus at work, ’cause thar’re fewer distractions, meanin’ I can get more work done.
Things aren’t quite back t’ normal ‘ere, though, ’cause nah everyone wants t’ go back t’ the cabin. So I still needs t’ waste time writin’ messages t’ scallywags instead o’ jus’ goin’ o’er ‘n askin’ ’em. Pretty sure they be jus’ usin’ the Death Virus as an excuse, ’cause they had ne’er had any problem gettin’ into work afore, but now, ohhh no it’s too difficult, got t’ look aft sprogs ‘n do things at home. But ye managed afore… 🙄 Oh, ‘n a lot o’ places still demand ye wear a face mask ‘n “check in” or whatever. Guh, I jus’ wish scallywags would grow up ‘n get o’er this flu already. ’tisn’t goin’ t’ walk the plank, it’s impossible t’ belay that from spreadin’, ‘n ye’re goin’ t’ catch it eventually, so jus’… I dunno, ignore it, eh? 😛
But aye, other than goin’ back t’ work ‘n gettin’ vaccinated, nah much has happened wit’ me, mighty. Jus’ recoverin’ from creatin’ Amai: Reaffirmed (please read it) by nah touchin’ DAZ Studio wit’ a barge pole fer the past 7 months or so, ‘n playin’ games instead. 😛 I discovered ‘n took advantage o’ an offer where ye could get three months o’ Xbox Game Pass fer £1, ‘n I’ve been super-surprised by it. I didn’ know they had so many good games on thar. 😮 So I’ve been playin’ a bunch o’ ’em. Tetris Effect: Connected be like enough me favourite so far. It’s like Tetris, Rez, ‘n Lumines somehow had a three-way sprog. 😆 I love all three o’ those games, so combinin’ ’em together be jus’ aaaaaah! It’s such a good game. 😀 It’s got a smart co-op mode, too, which I’ve played a couple o’ times wit’ random players, ‘n once wit’ someone o’er Parsec, a low-latency remote desktop app wit’ a focus on gamin’. Apparently thar was some annoyin’ lag, but ’twas playable. Nah sure who’s fault the lag was, but ’twas prolly at least partially mine. 😛
Anyway, I be off topic as usuaaaal! I’ll get onto the real subject now. 😅
Ye might be wonderin’ why in Davy Jones’ locker this post be written in pirate-speak. 😛 That’s ’cause, matey, today (or at least when I posted this tale entry) be apparently “International Talk Like A Pirate Day”. I ‘ave no idea which thought o’ this, but ahoy, thar ’tis. It’s jus’ another one o’ those things arbitrarily assigned t’ a random day fer no real reason. Like Burger Day, assigned t’ the 28th o’ May. O’ course, it’s also assigned t’ the 26th o’ August. ‘n thar’s Cheeseburger Day, which was apparently yesterday, the 18th o’ September. See? All bullshit. 😛 I’ll make one up meself: Act Like A Mong Day, October 4th, go go go. 😛
While I don’t mighty get why scallywags reckon we shall natter like pirates on this specific day, it’s jus’ a bit o’ fun, innit. 😆 ‘N ahoy, it ended up partially inspirin’ the scene created fer Return T’ DAZ Studio Aft A Lengthy Break Day! Which be today! Along wit’ Natter Like A Pirate Day! Double day! Yarr! 😛
Har, if the things I be sayin’ make even less sense than normal, even wit’ pirate speak, that’s… like enough t’ be expected. 😆 Why’s that? Well, ’cause I be nah sure wha’ kind o’ energy I be runnin’ on at the moment. 😛 Since it’s bin 7 months since I las’ touched DAZ Studio, I was a wee bit rusty ‘n slow. That, combined wit’ the limited time I ‘ave on weekdays, me underestimation o’ how difficult the scene would become ‘n how long ‘twould take, ‘n o’ course, a healthy dose o’ procrastination, I left this render a wee bit too late. 😅 But I mighty wanted t’ get it done, so I jus’ kept on goin’ through the weekend ‘n I’ve sort o’ been awake fer more than 24 hours now. Wants sleep but can nah cuz needs t’ write this. In pirate. Hehehehehehehehehaaarrr
Anyway, aye, Talk Like A Pirate Day was one o’ the things that inspired the scene I came up wit’. ’twas mainly somethin’ else (which I’ll mention later), but this solidified the idea ‘n gave me a date t’ aim fer. By now, I be sure it’s super-obvious wha’ the general theme be. 😆
So without further ado… here it be! HERE BE DRAGONS!
YAAAAAHAHAHAHAAAARRRRR! BOOM! BURNINATE THE SCALLYWAGS!
I din’t sleep on Saturday. ‘n ’twas worth it. 😆 Garr, this one was kind o’ a pain, ‘n turned out way more complicated than I thought ‘twould be, but I be pretty happy wit’ how it ended up. 😀 It’s… nigh-on like wha’ I had in me head. I imagined the dragon cap’n t’ be facin’ the camera a bit more head-on, but I ended up at this angle, eventually liked it, ‘n jus’ sort o’ went wit’ it. Wha’ do ye reckon? How do ye like it? Let me know! 😀 Eh? Wha”s that? “Why are dragons walkin’ ‘n clamberin’ about the ship when they could be usin’ thar wings”? ‘Cause laughin’ ’til I spew grog! I dunno, dragon pirate code o’ honour that bars ’em from flying? Use yer imagination mateyyyy. 😛
Anyway, the process… ’twas fairly straightforward this time, as I was kinda familiar wit’ all the things I needed t’ do. T’ get all the dragons kitted out wit’ pirate clothes, I used the same technique as I’ve used a couple o’ previous renders (Roar, Blue Dragon! ‘n Dragula): the Transfer Tool. I moulded the clothin’ usin’ ZBrush so it fits the Dragon 3’s naught pose, then used the Transfer Tool t’ smoosh the Dragon 3’s rig onto the clothin’ ‘n do an auto-auto-fit. Aye, I meant t’ write auto twice. 😛 The difference this time be that I kiiiind o’ overdid it a wee bit. 😅 Three outfits. Threeeeee o’ ’em. ‘n they all had at least three different pieces o’ clothin’. While I reckon I got away wit’ combinin’ the cap’n’s outfit into a single mesh ‘n shapin’ all o’ it at once, fer some reason or another, I did each piece o’ clothin’ in the other outfits individually. I reckon ’twas jus’ mighty difficult t’ get the arms ‘n legs t’ fit when done all at once.
The only other difference related t’ the process o’ fittin’-human-clothes-t’-a-dragon was that I saved out the morphed outfit as an actual “product”, ’cause I had t’ use ’em multiple times ‘n didn’ wants t’ use a scene subset, ’cause I wanted t’ limit the number o’ times DAZ Studio had a poo on me head. 😛 I can nah remember if I mentioned this in Amai: Reaffirmed’s post, but it’s fairly simple fer a single object. Ye jus’ select the parented outfit ‘n save it as a Support Asset -> Figure/Prop Assets. It’s more confusin’ ‘n complicated if ye wants t’ save an outfit that has things parented t’ it, like shirt buttons. I would explain but I would prolly fall asleep. Sorry. 😅 I might update this later when I be more awake. 😛
One notable thing I encountered was memory usage. Oh boy. Aye, I reckon ’twas a combination o’ things, but this scene’s peak memory usage was around 23GB. Havin’ 11 dragons in the scene prolly didn’ help, but I reckon the biggest problem was the ship I chose t’ use. ’twas mighty good, mighty detailed, ‘n way better than the alternative I had in mind, but fer the love o’ God. I ‘ave ne’er seen a scene use so many textures that the texture dropdown menu covers the entire screen:
Seein’ that memory usage was jus’… beyond ridiculous. Jus’ when I get more friggin’ memory, too! It’s like a curse; I always seem t’ push everythin’ t’ its limits, even if it doesn’t make sense. Like I remember runnin’ mighty low on space when runnin’ a 1TB HDD, so I upgraded t’ a 4TB. ‘n I be still constantly runnin’ low. Hoooowwwwww. 😛
Ah, aye. Ye, uh… ye may be wonderin’ how I was able t’ render a scene that required 23GB o’ VRAM when me usual GTX 1080Ti only has 11GB. That’s ’cause I managed t’ score the ultimate loot. I… managed t’ get an RTX 3090. The Founders’ Edition. From Scan, nah a scalper. 😅
I managed t’ get hold o’ one by writin’ me owns website sniffer. 😛 ‘Twas written partly jus’ a challenge t’ see if I could do it, but when I did manage t’ do it, o’ course I kept it runnin’. 😛 Three weeks later, I got a notification on me phone that ’twas in stock, ‘n after realisin’ it weren’t no false alarm, ‘n I’ve ne’er tried t’ go through checkout faster. 😅 O’ course, me bank tried t’ stop me by repeatedly refusin’ t’ allow the bounty even though I friggin’ approved it wit’ biometrics, but I got thar in the end, thankfully. Don’t worry, this beauty won’t be goin’ anywhere nigh eBay or a blockchain! 😛 I was originally plannin’ on gettin’ a 3080 or 3080Ti, but I jus’ couldn’t ignore that 24GB o’ VRAM. All that memory be literally useless fer gamin’, o’ course, but as evidenced here, it’s super-useful fer renderin’. This scene would ‘ave been impossible without it, thanks t’ the ship model well-nigh blastin’ all 11GB o’ me GTX 1080Ti’s memory out o’ the water alone. 😑
The 3090’s performance be basically wha’ ye might expect. Although it actually exceeded me owns expectations. 😆 Aft spendin’ some time overclockin’ it, I did a benchmark by re-renderin’ “Roar, Blue Dragon!”. The original render o’ that scene was done by a 1080Ti ‘n a 1070, ‘n it took 6:23:38. Wit’ a 3090 ‘n a 1080Ti workin’ on it, it sounded like a jet engine but it took 1:18:40. 396% faster. 🤯
‘N ahoy, since I can nah reckon o’ anythin’ else t’ natter about when it comes t’ the process, now’s prolly a good time t’ move onto the RENDER STATS! 😀 So, o’ course, the 1080Ti had t’ sit this one out ’cause it didn’ ‘ave enough VRAM, but at Quality 2 ‘n 95% convergence target, it took an RTX 3090 2 hours 30 minutes 14.92 seconds t’ render at 4K. Blistering barnacles. 😆
‘N o’ course, now fer the TIMELAPSE! Fairly average-sized one, this time. The render took jus’ under 30 hours t’ complete, includin’ the 2 1/2 hours o’ render time. Ye can watch the timelapse ‘ere! 🙂 I crammed in four tracks this time. First off be We Love Burnin’ Town, by Jake Kaufman, from the Shantae ‘n’ the Pirate’s Curse OST. It’s the chantey played in the openin’ stage. I originally got Shantae ‘n the Pirate’s Curse from a Humble Bundle ‘n kind o’ shoved it t’ one side cuz I’d ne’er heard o’ it ‘n didn’ even know wha’ ’twas. But one day, I finally gave it a try, ‘n ended up lovin’ it. 😀 I’d mighty rarely played Metroidvanias afore, ‘n this was definitely the first one I ever played t’ completion. That soundtrack be jus’ amazin’, too! Anyway, next up be, o’ course, You Are A Pirate from Lazy Town, prolly better known as the chanty used in the Lol Limewire meme. 😛 f ye look closely in the timelapse, ye’ll notice that I replaced the default DAZ Studio “activity image” (the image that shows in the top right o’ the program window) wit’ a reference t’ that meme. Temporarily, o’ course. It’s back t’ Morshu now. 😆 Now, if ye remember, I mentioned that Natter Like A Pirate Day was only partial inspiration fer this scene. The main inspiration was actually the next two shanties in the timelapse… ‘n both o’ ’em are by none other than CHEESY DJ BOBO! 😆😆🧀 Oh aye, yet again, the master o’ cheese inspires me renders. God damn it. 😆 He has an album titled “Pirates o’ Dance”. Unlike his hilarious “Vampires” album, mighty few shanties on this album actually refer t’ pirates, but it’s aye, two be enough. 😛 The first be introduced by the 30-second track Ghost Ship, which seamlessly leads into the actual chanty, Garunga. GARUUUUNGAAAAA 😆😆😆 I end the timelapse on the more upbeat shantey o’ the pair: Pirates of Dance, which goes on ’bout pirates that dance ‘n come from outer space. Funnily enough, that sounds mighty similar t’ t’ Space Channel 5’s Part 2’s Rhythm Rogues. 😛
UPDATE! Ah, crap, I forgot about the redlines I usually share! Here ye go:
Jus’ six preliminaries afore I rendered the final thing. Dunno why, but it felt like way more than that. Maybe ’twas cuz o’ all the damn problems I had wit’ renderin’ towards the end. 😑
I reckon that’s it fer now… maybe. I can nah reckon o’ much else. 😅 Maybe that’s cuz all me noggin’ wants t’ do be make me lie down ‘n melt into me cot fer like 48 hours. But it can nah, cuz I got work on the morrow. Arrrgh. 😛 Anyway, if ye reckon thar’s anythin’ I’ve missed, jus’ let me know! ‘n I hope ye like the render! 😀
(Disclaimer: Sorry, I be a scurvy dog ‘n used this website t’ translate most o’ me stuff into pirate-speak. I would’ve done it meself but I be mighty crap at talkin’ pirate, ‘n I was runnin’ out o’ time like. Uh… matey. :P)
ROUGH BAD ON THE CASE I WANNA HIT YOUR FACE* (*I don’t actually want to hit your face, but H.P. Sauce Baxxter does 😛 )
Lmao, such stupid lyrics. 😆
Anyway, hello again! It’s Valentine’s Day today! Yaaaaaay, meal deals! 😛 I don’t really care about it myself, but I guess it’s nice for people who like it. For whatever reason. Maybe people forget that they have a partner and need this day as a reminder that they should do nice things with them? Shrug. Meal deals. 😛
Even though I don’t actually care about Valentine’s Day, you may remember that a couple of years ago, I created a Valentine’s themed render, Amai, and a short story to go with it. It featured a couple of characters from my never-ever-going-to-be-finished novel, Dragon Tail. I really enjoyed doing both parts of it. The render was fairly simple, but it was fun to make and I just loved how it turned out. Writing the short story was also a lot of fun, and it really helped me flesh out Whirlwind and Emerald’s relationship and gave me an excuse to dig a bit deeper into the world’s lore.
Because I enjoyed it so much, I was thinking of doing the same thing again at some point in the future with something else. But recently… well okay, not recently, it was probably like 8-9 months ago now, but anyway, I wondered… do the render and the story have to be separate? Surely there was some way of combining them. And of course, there are at least two ways of doing that. The first way is through animation, but hahahahahaha no! 😛 Not a chance, not in DAZ Studio. Especially if they think that the 4.14 update included “an advanced Animation Timeline that’s more powerful than ever before”. It’s not hard to be more powerful than a turd. 😛 If I can find a good place to rant, I might talk about the 4.14 update further down, but for now, stay on target, Mongz! 😛 Anyway, yes, no, no animation from me, I don’t have a clue where to start, I don’t have the patience for it, and DAZ Studio sucks at animation. But the second way? Well, I could tell the story through a sequence of still renders. And how best to do that? A comic, of course!
You might be able to tell where this is going now. 😄
I have very limited experience with comics. Comics are cool, sure, but I’ve never really been interested in them, content-wise, probably because 99% of them are superhero comics and I generally don’t like superheroes. The only comics I ever owned were Simpsons comics. Shut up, my mum bought them for me and I liked them and I still have them and shut up. 😅 So, as I don’t really read comics, of course I have zero experience with writing them! I’ve never made a… ahhhhhh no wait, I lieeee! I have made comics before! Holy crap I still have them!! 😆 My DIABLO II COMICS BABY YEEEEAAAHHH!
I almost completely forgot about them. 😆 Yes, back when playing Diablo II, still one of my most favourite games of all time and definitely still my most favourite action-RPG (the only thing that comes close is Path of Exile, it has excellent game mechanics, but it plays too slowly for my taste and its story is almost embarrassingly bad 😛 After finally playing it for long enough to reach endgame, I take it back; Path of Exile is actually a pile of shite, a luck-based wheel spinning game that demands a lifetime’s worth of grinding just to get one decent piece of equipment, and there’s absolutely no game balance at all, and it contains literally the worst endgame boss I’ve ever played in my life, feels like whoever designed it has never played a game before), I’d take screenshots mid-game and add dialogue in Paint afterwards. 😀 Most of them were a single panel and didn’t connect with any other panels, so I’m not sure if you could actually classify them as comics. 😛 I did make a few sequences, though. After creating and sharing my first set, the person I was playing with also started making some, too. I think his were generally better, but the grammar was a bit, uh… painful. 😛 The majority of mine are “meh”, but there are some that I was really happy with. I’m most proud of these two, especially the second:
Ahh, good times. 😀 Although I do wish we could’ve gotten to Hell difficulty. We never made it there for various reasons. 😦 I really need to play it again at some point for old times’ sake and fully complete it. Although there’s no way anyone will join me, and it’s not quite as fun solo. But, having said that, I’ve recently heard of a relatively new mod called Project Diablo II which I definitely need to check out. It looks surprisingly popular, and it seems to add in some of the less crappy elements from Path of Exile, such as endgame maps, which sounds good. But it sounds like they added luck-based item corruption crap, which is crappy. Really crappy. “Want to improve your gear? Gonna have to use this super-risky item! OHHHH too bad RNG not on your side and now your Super Mega Sword of Epicness +98 is a Super Shit Sword of Poo -69 lol go out and grind some more.” And periodic server resets, too; I just don’t understand that. It’s like they don’t realise that some people would like to reach endgame, but can’t make it there before the server is reset, because they have a job. 😛 It might be too vanilla for me, as well. I last played Diablo II with the Eastern Sun mod, which had some great features like gem/rune stockers, an infinite stash, way better item gambling which actually makes gold a useful resource, and crazy bonus dungeons with massively better loot drops, so I’m not sure if I could play Diablo II without that now. 😅
ARGH DAMN IT STAY ON TARGET MONGZ! 😛 Right, back to the present, and the actual subject. 😆 I’ve been thinking about making some kind of comic for a long while now. Using the Dragon 3, of course. 😛 I didn’t plan on doing anything massive, initially, just a four-panel comic or something. I could never think of an idea for what to do, though, so it kept getting kicked down the road a little bit every time I thought about it. Probably didn’t help that I never actively tried brainstorming any ideas, but y’know, whatever. 😛 But then, a little while after uploading Roar, Blue Dragon! to my DeviantArt page, I logged in to check on how it was doing, on the off chance that someone had, y’know, looked at it. 😛 I wasn’t particularly confident of that, and thought that it would be the last time I logged in, because I really don’t feel like I belong there. It’s like… actually no, I was going to post a dumb, extensive analogy but it turned into a rant and I’ve gone off-topic once already. I’ll put it at the end of the post. Basically, DeviantArt is shit in various ways and I don’t like it. 😛
Anyway, so, I logged into DeviantShart, and amazingly, I had a new watcher! Wow! First new watcher since I opened the account about… I dunno, 5 years ago? 😛 They’d also left a message, saying I should do a dragon series… or something along those lines. Ha! Yep, I’ve been thinking about it for a long time. 😆 That message, though… I’m pretty sure it accelerated my thinking. Because not too long after that, around early August, I had a spark of an idea. I was actively thinking about what comic I could make, and thought maybe I could use some kind of celebratory day as both a theme and a target date. Something far enough away to give me a chance to create it, but not too far away, because I would procrastinate. 😛 Conversely, it couldn’t be too close, because I’d never done anything like this before and missing the date would mean I’d have to wait a year for it to become relevant again. And while looking through my calendar, it hit me. The idea, not the calendar. 😛
I didn’t actually scream that, I’m exaggerating as usual, but anyway, yes, Valentine’s Day! Meal Deal Day! Today! Unless it’s not! 😀 I thought it was a perfect thing to base the comic on. I had characters who were in a relationship (Whirlwind and Emerald), and a world (Guira) that I could use, and as a bonus, it was a great way for me to flesh all of that out. Worldbuilding is fun! 😀 Valentine’s Day was also far enough away for me to have a chance of making the date, but not too far away that I’d just go “ahhhh, I’ve got loads of time left let’s play 900 hours of Diablo II”, followed by “crap where’d all that time go”. 😛
So, I made a start on it then, in early August 2020, sacrificing most of my spare time. And now, finally, after half a year of frequent and increasingly painful interaction with DAZ Studio… IT’S ACTUALLY FINIIIIIIIIIISHED!!!! Read it! READ IT READ IT READ IT OH MY GOD PLEAAAAAAAAAAASE READ IT! 😆
Please. Please read it. 🙏 It took me 372 solid hours and required 408 hours of rendering please read it. 😆 I also really want to know what you think! Is there anything you particularly like, or think that I could improve? Any constructive feedback is welcome! It’s my first comic, so there’s definitely room for improvement.
Aaand now I have to talk about it. Sheesh, since it was a huge undertaking that took far longer that I expected and required much more work than I anticipated, I’m struggling to think of where to start. This might end up being pretty long, so if you don’t care that much, you can skip straight to the super-excellent render stats and timelapse using this handy TL;DR link:
Okay… I will probably end up yammering and going in random directions with this explanation, but let’s start with the concept, I guess? 😅 Well, if it wasn’t obvious from the title of the thing, it’s a follow-up to the original Amai, the short story, specifically. Amai (Guiran Draconic for “love”) was about Whirlwind and Emerald’s first Franaie’dianas together, which, as mentioned in the comic please read it, is like the Guiran dragons’ equivalent of Valentine’s Day. Apart from that it’s not commercialised (that’s not a thing among Guiran dragons, there are no “meal deals” 😛 ) and is inclusive of everyone, not just those already in a partnership/matehood. As I probably already mentioned, I really enjoyed writing the original Amai, and – this might sound a bit big-headed, sorry – I love those two dragons. I think they make such a great pair! 😀 After thinking of Valentine’s Day as a potential theme and target date, it made perfect sense to get those two together again for a follow-up, set on their second Franaie’dianas together. 🙂
The original Amai was a lot of fun to write, and it was a really nice change, too. Y’see, Dragon Tail, the story Whirlwind and Emerald originate from, is actually a fairly dark story, so it was great to be able to write a slightly more light-hearted scene in that world. 😀 I definitely wanted to double down on that for Reaffirmed, so I made sure to kick things off on a lighter note, and tried to keep it fun and light-hearted wherever possible. Kinda maybe like a rom-com, or something. Not like I watch rom-coms, (or films in general, lel) so I have no idea if it’s similar. 😛 Soon after I’d committed to finishing the project, though, about four or five pages in, I couldn’t help but feel worried about making a comic out of this kind of story. It’s a weird one: there’s no villain, no conflict, or anything negative like that. It was more of a… “worldbuilding story”. A story that showed a peaceful moment in these two dragons’ lives, while exploring a small part of the huge world they live in. So, because of that, I felt that it might end up being ultra boring. I did my best to avert that by sharing bits of lore that I hoped were interesting, and trying to make it nice, fun, and wholesome, but I simply had to accept that most people might find it boring because no-one gets stabbed. Which is slightly creepy when you think about it. 😛
As I’d never done a comic before (well, not one of this size, anyway 😆), writing the script was a bit difficult. I only ended up writing about half of it to begin with, finishing the rest as I went along. The whole story was in my mind from the beginning, but I didn’t want to write it all out just in case I made a mistake and wrote the dialogue in a way that wouldn’t work in comic format. And as expected, this happened a few times in the early stages. 😛 I had to simplify the dialogue so it would fit a couple of times, and there was one time I was forced to split a page in two to prevent it from ending in a weird spot. I always aimed to end each page at a point where I hoped the reader was looking forward to the next. Kinda like little cliff-hangers. 😀 While it was difficult, it was still great fun to write, and it really helped me to develop and solidify the kind of relationship Whirlwind and Emerald have. (Emerald’s normally very dignified and gentle, but she becomes a little friskier and more playful when she’s alone with Whirlwind, and this behaviour quickly ends up spreading to him, too. 😀 ) Also, it was great to glimpse into Whirlwind’s past, as I haven’t really fleshed that out very much (Whirlwind is, how do I put this… “conditionally amnesiac”. If I say any more then I’ll spoil my will-never-be-finished novel 😛 ). Info about specific locations are really sparse in Guira’s lore, too, so creating the Transient Lake was really awesome. And it helped refine and bolster Guiran Draconic, too! I loved being able to throw Draconic dialogue into the comic here and there, when the time felt right. Hell, it even drove me to pull my finger out of my ass and finally write Draconic Script for the first time, used for the engraving on Emerald’s bracelet! 😀 (That’s a SPOILER but you’ve already read it, right? Please read it, it took 372 hours 😆)
To explain that last bit… well, normally I write Draconic using the Latin alphabet. Y’know, the Latin alphabet, as you can see in this entire blog post. Apart from the emoji. 😛 When Draconic is written in that way, I call it “Romonic”. I know, “lololololol”. It’s a perfect anagram of “moronic”, and I absolutely love it because of that, but would never use the word canonically. 😛 It’s similar in concept to Romaji: Japanese words written using the Latin alphabet. Basically, Draconic Script to Romonic is like kanji (and hiragana/katakana) is to Romaji. It just makes it easier/possible for us Westerners to read. Yeehaw. 😛 I probably explained that badly, but hey, whatever. Explaining it is varok saknankra. Or chou muzukashii. 😛
As for the actual posing and rendering, well, my feelings about that are… mixed, really. 😅 At times, it was fun. Poses would fall into place really naturally, and I made great progress very quickly. Sometimes surprisingly quickly; I remember knocking out about half of Page 21 in less than 3 hours. However, other times, it was turbo-frustrating. I’d struggle to get the dragons posed in the way I wanted to, or I’d finish a pose and then realise that it wasn’t quite right, requiring me to move the entire model and re-pose all four limbs. Sometimes setting up the lighting and camera angle would end up being a massive pain. Then after finally getting it right, I’d realise that single panel took me about 5 hours and I had 3 more panels left to finish, which realllly didn’t help my motivation. 😛
In an attempt to maintain a steady pace and to prevent my motivation from dropping to zero, I did a couple of things. First off, I set up a mini website for a small number of people, namely family and one person from work who has given great feedback on some of my past renders (Dino Disco being one of them, which I was able to improve quite a lot). I also set up a mailing list for them using Mailchimp, which gave me an audience. Sure, it was a tiny audience of five people, but they were active, seemingly interested (they must’ve been; it’s not like I forced them to sign up, and they always had the option of unsubscribing 😛 ), and knowing that they were there and looking forward to the next page basically prevented my motivation from bottoming out. Secondly, I created a project plan, initially using some website called Team Gantt or something, but I found that to be a bit shit (and their terms basically said that they could do anything they wanted with everything you submitted, lol), so I used ToDoList instead. I’ve had that program installed for quite a while, and used it to store ideas for ULSG V14, but never really actively used it. I learned that it had a built-in Gantt chart, though, and it ended up being a massive help for planning and keeping myself on track, especially in the final stages when I knew what kind of pace I could keep. I’d’ve probably never been able to finish on time without it. 😀
As a side-note, making the mini site was an interesting exercise. I’ve never really done much related to web development before, and after scratching the surface, I’ve determined that it’s hell on Earth and I hate it. 😛 Despite that, I think I made a nice little website. I based it off a simple template and heavily modified it, successfully adding a pretty neat gallery called PhotoSwipe, and even managed to add a private livestream with a live viewer count. It was completely private; I streamed using OBS from my main machine to my server box without going through any external provider like Twitch or Facebook or whatever. The live viewer count was implemented using PubNub, which was free for me to use because I didn’t fire many events. Of course, I barely used the thing, and no-one was watching when I did, but whatever! Experimenting is fun and educational! Now I know never to get a job in web development! 😛
Even with the audience and the project planning, it took a lot of willpower to get this thing done. There were a couple of times where I took excessively long breaks, about a month in total, which was really bad. If that had been a week or so longer, I’d’ve blown it, because I finished the final page just three days before the deadline. 😨 I’m not exactly sure why I took such long breaks. One happened just after I returned home after some birthday celebrations, so that might partially explain it. But I’m betting it was just because the constant exposure to DAZ Studio was making me lose what little was left of my sanity and I desperately needed a break from it. 😛
Because this was the longest project I’ve ever undertaken in DAZ Studio, of course, there was no shortage to the number of times DS attempted to shit all over me, the project, all of its users, and basically everyone and everything in the world! And I’m actually not exaggerating quite as much as I normally do, there. 😛 I experienced a massive variety of problems. Some were common things that I’ve endured before, such as GPU memory leaks, Iray suddenly not working properly for no obvious reason, a disappearing mouse cursor, and DS randomly thinking I want to select multiple objects in the Scene tab despite not even touching the keyboard. However, there were some brand new ones, including insanely low viewport performance for no good reason (related to object count rather than poly count, wtf), massive delays when deleting objects (super wtf), decals casting shadows in certain situations that made no sense, very random and very sudden crashes, and worst of all, something I can’t easily explain which made me believe there was a crippling, fundamental flaw with the program. I’ll try and explain the last one, because when I discovered it, I had a very unpleasant feeling that it was going to wreck the project at some point.
Oh, but before I explain, one good thing about this bug was that it made me move all my scenes into OneDrive so they were backed up. It was pretty cool, because I was able to keep my scenes folder in the DAZ Studio content library even though it was nowhere near my OneDrive folder. I did this using a symlink, or more accurately known as a “directory junction” in Windows. Since doing that, I’ve kinda gone a bit symlink crazy, moving a bunch of stuff that really should have been in OneDrive in the first place and symlinking to it from the original location so it feels no different than normal. It’s great! 😀 Of course, you don’t need to use OneDrive. That technique will work with any similar cloud storage provider (Dropbox, Google Drive, etc.).
Anyway, the attempted explanation of the DAZ Studio insanity. 😛 So… it was Page 8, Panel 2 when it first happened. I needed to create the “young” version of Whirlwind, so I loaded the scene containing the adult version of him, duplicated him, and made my modifications there. When I was done, I deleted the adult version, zeroed the young version, and saved it as a separate scene and a scene subset. I went back to the original scene, and loaded the scene subset. When done, it took me a moment to realise that the young version of Whirlwind wasn’t in his zero pose. It had imported him in the pose and position that the adult version of Whirlwind was in… in the previous panel. If that makes no sense, I know, it doesn’t. 😛 I thought I’d eaten something weird. So I tried again, and yep, same thing happened. It was absolutely bat-shit insane, and I genuinely don’t know why it happened, or how it was possible. After realising this, I soon discovered that a massive number of the trees I’d spent 400,000 years positioning manually had moved to positions they used to be in about 2-3 pages ago. Some of them were originally placed quite roughly because they were off-camera, and when they eventually had to be rendered, I moved them to more accurate positions and added additional ones. But then some of them randomly moved to their existing positions, silently and for no goddamn reason! What the CRAP?!? 😕 I swear I saved a video of me demonstrating this, but I must have accidentally deleted it while clearing out the first half of my recordings when I ran out of disk space. God damn it. 😑
There was also the DS 4.14 update, which I foolishly decided to update to halfway through the project. As far as I can tell, it does almost nothing good. It updates Iray which completely breaks the program for some users, adds new bugs, and fixes no existing bugs. This kind of update is extremely common. But, the fact that they ranted and raved about this particular update despite the fact that it’s no better than the last 13 updates they’ve done, throwing tons of bullshit everywhere in the process? That suuuuuper pissed me off. I made a massive list of things wrong with it, and was going to go through them in a video, but I just ended up feeling like I couldn’t be bothered. It wouldn’t make anything better. Basically, Filament is useless because it only works in the viewport and looks like shit, the animation timeline hasn’t changed and is still basically unusable, and Iray 2020.1.1, while nice because it adds support for 3000-series GPUs (even though they’re still friggin’ unbuyable because everyone has gone insane), it’s also not nice, because it removes support for some older GPUs, and this fact is not mentioned anywhere obvious! I’m pretty damn sure their installer doesn’t check your GPU before updating, and you can’t (easily) downgrade after updating!! Sure, this didn’t affect me, because I have a modern enough card, but I can imagine it screwed someone! What in HELL are they smoking?! 😡
So yeah. As usual, DS was crap. All day, every day. 😛
Towards the latter half of the project, I realised that I would need to find a bracelet of some design, and I was divided on what to do. Part of me (the lazy half 😛 ) thought I could find something suitable that was already made, while the other part of me thought that was going to be impossible and I should just create it myself. The latter half was absolutely right. 😅 Bracelet models were quite rare, and none of them looked right. “Besides,” I thought, “if I’m gonna spend so damn long making this comic, I oughta do this right and make it unique”. So, I used Photoshop to sketch out a design that I liked the look of (using its symmetry tool, which is absolutely friggin’ great :D), and finally cracked open Blender, with the intention of using it seriously for the first time. I was a total beginner, but I couldn’t spend weeks and weeks on this bracelet, so I assimilated all the basic tutorials as quickly as possible and started to hammer something out immediately afterwards. And against all my expectations, I think I actually managed to make it look good! Sure, its topology would probably make any real artist throw up, but I don’t care, it’s the first model I’ve created since college, about 14 years ago or whatever! 😀
The last time I opened Blender was way before this project. I’d downloaded it to satisfy my curiosity, but quickly decided that I hated it because its UI felt like it was designed by a sadist. 😛 Nothing about it made sense. I think they’ve updated the UI since then, though, because the version of Blender I used to create the bracelet felt much less… insane. 😛 Apart from the menus that are seemingly only accessible through hotkeys that aren’t mentioned anywhere. But anyway, yeah, my experience with it was generally positive. It felt… clunky, and a bit confusing at times, but unlike a certain other piece of software, I don’t remember anything in Blender blowing up or making me angry. I found out that it can even do high-poly sculpting, like ZBrush! I had no idea it could do that! It’s even got scripting and probably-usable animation tools and fluid simulation and UV editing and texture painting and aaaaaaaaaagh how is this free?! 😮 If only DAZ would create something similar to DS’s GoZ bridge, only for Blender. Sure, DS has a Blender bridge available as a free addon, but it’s only one-way. Maybe I should check it out anyway, see if it’s possible to maybe leave DS behind. Or maybe DAZ could just fix their software already, instead of making one-way bridges to other software. It’s like they don’t realise that DS would actually be a really great program that I wouldn’t complain about if it didn’t have so many massive issues!
So aye, the bracelet creation went way better than I expected! Created the basic 2D shape using my reference, thickened it up to be 3D, refined it over time, then sculpted it into the final, higher-poly shape. That’s pretty much it! Although I did find out much later that more polys doesn’t always equal better quality. I would’ve probably gotten some nicer, smoother reflections on the metal if I’d been more aggressive when decimating the model before exporting it.
After creating Emerald’s bracelet, I thought that would be the last time I’d need to touch Blender. But nope! I did a ME! 😛 At the 11th hour, I ended up revising the script after not being so sure about the ending I’d written. It felt… shallow and meaningless, and a bit sudden. However, the new script made the comic two pages longer, which had a pretty big chance of making me run out of time if it took much longer than I’d estimated. When I got to the point where I had to make a decision about which path to go down, I took a risk and chose the new script. This added the second bracelet and the linked enchantment, which I felt made the gift massively more meaningful. 🙂 But that meant I needed to design and model another bracelet! AAAAGH! 😛 Anyway, it was fine. Due to its simpler design and my newfound (albeit slight) familiarity with Blender, the second bracelet didn’t take quite as long to create.
Let’s see, what else… maybe how I arranged the panels? Well, I used Photoshop for arranging the panels and adding postwork. Before starting a page, I’d refer to the script and block out the panel layout, then measure each of the blocks and use that as the render size. Doing that instead of just rendering at a higher resolution and cropping what I didn’t need probably saved a crapton of rendering time. 🙂 As for postwork, most panels were given a once-over with the blur tool, to smooth out any really sharp highlights or to remove noise the denoiser didn’t catch. I also added blush effects and eye glint to the dragons in some panels. Some panels consisted of two renders, usually a misty and non-misty version, which I combined together. After any corrections, I tweaked the levels, added the border, aaand that was about it. 🙂
Oh, there was one page that had some postwork that I was really happy with: page 8! Panel 2 had a ghostly version of a young Whirlwind and one of his childhood friends, which was supposed to represent Whirlwind recalling memories from his past. I super-liked the idea when it came to me, but didn’t think I could pull it off as well as I’d hoped. But I was wrong! 😛 I rendered the younger dragons separately and with no environment, and slapped it on the top of the main render. Because there was no ground in that render, their feet had to be masked a little bit to blend in, but that was easy enough. To get the ghostly effect, I reduced the opacity, duplicated the layer, and applied a gaussian blur and glow (with its blend mode set to Soft Light) to the second layer, and set that layer’s blending mode to Screen. That was it. I didn’t expect it to turn out that well so easily! 😀
Overall, Photoshop was fine throughout the project. Although it did have a few annoying moments, especially that one where it suddenly became incapable of saving. 😑
Whew… I think that’s about it for the process? As there was so much work involved, trying to think of what to talk about is a bit difficult. 😅 If I remember something, I’ll probably add it here later. Let me know if there’s something I missed! 😀
TL;DR: I made a comic and it was difficult please read it. 😛
Even though it was difficult, I’m so happy with how the whole thing came out. Hell, I’m just happy that I finished it in time. 😆 There were a few panels that I was especially happy with, though, and my favourite panel of them all has to be Page 9, Panel 2. I had a lot of fun coming up with it and posing it, and I absolutely loved the result. It’s so damn cuuute! 😆 I liked it so much that I rendered it in 4K as a standalone thing:
Okay, now for the timelapse! And this time, as you’d probably expect, it’s an absolute monster! 😀 The raw footage was so long that I maxed out one of DaVinci Resolve’s ludicrously long timelines, and I nearly ran out of disk space, so I had to render half of the timelapse to free up some space before I could continue. 😆 Over the course of six months, the comic took a solid 15 days, 12 hours and 41 minutes of human time to create, and required 17 days and 44 minutes of machine time to render. So 781.4 hours between myself and the computer, in total. (although if you include the fact that the machine was also working while I was working, it’s more like 1,154 hours in total 🤯) So, this timelapse shows those 372 hours of human work compressed into 21 minutes. And I definitely need to warn about epilepsy here. 😆 Aye, this video flashes a fair bit as it switches between lighter and darker colours, so seriously, don’t watch it if you suffer from epilepsy! If you don’t, great! You can watch it here! 😀
And yep, there’s a bunch of at least mildly related music to go with it, five songs this time! And because two of the songs I chose are Japanese, I went ahead and added lyrics into the video for them. 🙂 Anyway, to start off, I play the dumb song which begins with the dumb lyrics shown at the top of this dumb post: The Age of Love by Scooter. Mr. Baxxter talks about love, peace, and unity, but begins the song by declaring he wants to hit you in the face, lmao 😆 I ended the blog post of the original Amai with this song’s music video, and I still can’t get over how awful it is. 😛 Next up is Love and Devotion by Real McCoy… but DJ Bobo got involved! I can’t find a link to the song on YouTube, so that’s a Spotify link instead. Hahaha, Bobo’s cheese-flinging rampage CONTINUES! 😛 I thought the original might’ve been just as cheesy, but it’s nowhere NEAR as bad as the remix! How does he do it?! 😆🧀 Following on from the cheese is a track relevant only by title: Sunset, by The Midnight. I’ve been quite into synthwave recently, listening on DI.FM, and this artist keeps popping up every now and then. They’ve become one of very few artists that I can genuinely say I like, and I don’t actually think I’ve heard a bad song from them. 😀 Lol never mind I finally heard a terrible one 😆 The next track is only slightly relevant: a nice song called Starduster by jimmy thumb P, featuring Hatsune Miku, a Vocaloid by Crypton Future Media. It mentions love, stars, and a sunset, so hey, that’s relevant enough for me. 😆 The lyrics and translation in the video are taken from this transliteration and translation by marvelangga. You probably won’t see this, but thank you, marvelangga! 😀 I was going to finish the timelapse with the Reincarnation Mix of Your Affection, from the Persona 4 soundtrack, but at the last moment, I remembered 24h by DAOKO, featuring You Kamiyama, and decided to use that instead. 😀 I first heard this song while playing Dragalia Lost’s Valentine’s Day event, Dragalia Lost being the only mobile game I’ve honestly thought was too damn good to be stuck as a mobile game, and no, it’s not just because I like dragons. 😛 (vaguely related side note: after I was lucky enough to pull the Halloween variation of Mym in Dragalia Lost, I realised soon afterwards that she always transforms into a lingerie-clad version of Brunhilda, and it is amazing 😆) Anyway, the lyrics of the song are a little odd, and I’m not entirely sure it’s a love song, and not sure it matches the comic very much, but hey, it’s got some great energy and I like it, so shush! 😛 The translation was done by Megan Rose, taken from this page on LyricsTranslate. Again, you probably won’t see this, but thank youuuuu! 😀
And now, the render stats! And of COURSE I kept notes on how long every single panel took to render! 😛 Render settings varied, but most panels were rendered at my usual: Quality 2 with 95% convergence. Some, namely the ones with the stupid-expensive mist effect, had their target convergence reduced to around 85-90% because I found that half the time was spent on the last 10% and it made no perceivable difference in quality. Anyway, here’s the render time of the last version of every panel:
SUPER TOTAL: 17 days 44 minutes 20 seconds (425 renders)
You still there? I’m not. 😆 Daaaaaaamn that was boring to type out. 😛 Yep… that’s a lot of rendering. 😛 It doesn’t look like that much in the spreadsheet I have, but after typing it out into that list… yeah. 😆 425 renders! When I bought my 1080Ti way back when, I was all “gaaah my wallet, that was so damn expensive”, but now I’m thinking it was totally, totally worth it. 😀 And hey, since the data’s in Excel, I made a chart visualising it, for no apparent reason:
Whew… that’s about it, I think. Now that I’ve finished the comic, no more rendering for me for about… I dunno, 8 months? 😆 I’ll spend that time playing some games that I’ve been wanting to finish, instead, like Trials of Mana, and Pikmin 3, and CTR: Nitro-Fueled and Disgaea 4 and Diablo II and probably some others!
I really hope you enjoy(ed) the comic! Depending on the reaction and/or how I feel, I might end up doing it again in the future. But with a story that contains explosions. 😛
Have a great Meal Deal Day! 😆
Inane rant about DeviantShart that I skipped earlier:
DeviantArt is like… and here comes a dumb, extensive analogy… on a golden island up in the sky in the world of DeviantArt, you’ve got the genuine artists, people who somehow do it as their occupation seemingly without being involved in games or film. They make really impressive stuff (probably because it’s their job) and everyone and their dog watches them. Then, down on the mortal plane, you’ve got the mildly popular people and total unknowns, some artists, some not really, but there are people down there (like me) who are just hobbyists. They may not be anywhere near as good as the people on the golden island, but they still try their best to create pretty things just so others can enjoy them. However, dotted absolutely everywhere on this mortal plane, there are these gigantic mountains made out of SHIT. 😛 I’ve tried to find stuff by other unknowns that I’m interested in so I can give them a watch, but I can’t find anything good due to all the shit! And there’s no point hoping for anyone on the golden island to notice you and watch you: they never bother looking down because they’re probably too busy being amazing, and probably because all they can see from up there is shit. 😛 And more often than not, the owners of the Shit Mountains are somehow unrealistically popular even though all they do is spam shit, and they even try to get some kind of payment for their shit (or for nothing at all) and for some unfathomable reason, people actually give them money! Often for no logical reason! It makes even less sense than people buying microtransactions in a slot machine mobile game that doesn’t pay out!
And it’s not just the content, either. There’re plenty of other things wrong with DeviantArt, too, like… you can’t submit more than one thing at a time unless you pay for the stupid overpriced-as-fuck Core subscription (completely ridiculous), I’m pretty sure a non-Core user’s deviations eventually become near-undiscoverable (might be a legacy thing, but I swear I remember seeing something that suggested your work stops appearing in the “browsing” sections of the site after a few weeks unless you’re Core, making it only possible to find your stuff through search), their site has only been half-converted from the old, ugly, bogey-green theme, their new site design isn’t much better because of its really confusing layout, and their crappy emojis still feel like they were pulled from random MySpace sites made in the 90s and I absolutely hate the hell out of them. Oh, and if you don’t feel like giving a fave or a comment, there’s a third option for some reason: you can give out a “Llama Badge”. I think it’s supposed to be a positive, fun thing. Because sure, real life llamas are fun and fairly cool. But the first thing I thought when receiving my first Llama Badge was: “Eh? Llama badge? You mean like the thoroughly negativeLlama Medal from the Skulltag source port of Doom given to you when fragging someone who’s badly lagging or typing in chat, which means “lamer”?” So my immediate connection was that people give llama badges out to say you’re lame! And even worse, the llama badge can apparently level up, so the site’s actually giving you a level of how much of a lamer people think you are! When I made this connection, I was like: “Why’s this asshat calling me a lamer? If they could be bothered to go through the trouble of giving me this, why couldn’t they have just given some constructive feedback, instead?” Thinking maybe my work was inferior to theirs or something, I checked their page, and lo and behold, it was another giant mountain of shit, seemingly all made in Paint, and owned by someone doing a “donation drive” for points so they could get Core on someone else’s wallet. “gib money plz”. Why do they even want Core? OH YEAH, to be able to upload multiple things at a time to grow their shit mountain faster. 😑 DeviantSHART. That’s what I call it. 😛
How’ve you been? Good, I hope! I’m pretty much the same as always, sitting in front of a PC, coding for work, then playing games, rendering, and sometimes doing little bits of coding and writing for fun. 😀 Only major thing that’s happened recently is I had my 30th birthday earlier this month, which was great, but now I feel kind of old. Physically, not mentally; mentally I’m still about 8. 😛 Although according to Dr. Kawashima (Brain Training for Nintendo Switch, fitting birthday present 😆), my brain is 23. Just can’t manage the minimum score of 20 because my memory is apparently a bit poop. 😛
Anyway, yes, Halloween! No trick-or-treating this year probably, which is both bad and good depending on who you are, I guess. 😛 Makes no real difference to me, since I don’t get any trick-or-treaters anyway. 😆 But I have got a treat for you, sir/madam/smadam/nullreferenceexception! A Halloween render! Which almost didn’t happen! 😛
Yeah, I kinda barely made it this time around. 😅 I’d had several ideas a while in advance, but I just couldn’t decide on one. And none of the ideas I wrote down were the ones I did; I came up with one final idea while trying to decide, and I decided to pursue that one instead. 😆 So there was that, and there was the fact that I’m working on a, er… secret project right now, which has a strict deadline and requires a crapton of rendering, and if I miss that deadline, the theme of the project won’t be relevant. I scheduled a week to do my Halloween render when planning the secret project, but I’d totally overestimated how much time I had during weekdays. And Halloween landed on a Saturday this year. So I basically had fragments of time during the week, and the Friday evening and Saturday morning to do this. Not helpful when I’m burned out from doing nothing but working and rendering. 😅
But no matter! I may want to just sleep for a year now, but I still managed to finish the Halloween render on time (just 😛 )! Behold! BLEH!
Hahaaa, I got a dragon to wear a tuxedo! That experience I gained from “Roar, Blue Dragon” paid off! 😀 It went way more smoothly than that, too, probably because the suit wasn’t totally broken. 😛 I wasn’t sure this idea was going to work at first, and almost fell back to one of my other ideas, but I think it came out great in the end! What do you think? 😀
When it comes to the process, I can’t really think of much to talk about for this. Maybe that’s because I’m on about 3 hours sleep right now and all my brain wants to do is turn off. 😛 Fitting the suit was pretty much the same deal as “Roar, Blue Dragon”, the difference being that the assets were already in DAZ Studio. I fit the tux to the figure it was intended for (Genesis 8), set up the materials, and then removed the figure, leaving just the clothes. Of course, ZBrush made things a bit difficult afterwards, since you can’t sculpt multiple SubTools at the same time; you have to switch between them. I got around this by exporting the set of clothing items as a single object, then reimporting it into DAZ Studio before sending it to ZBrush. If I had the time, I’d’ve probably sculpted them individually, but… I didn’t, so I didn’t. 😛
The only other thing was the ground mist, which turned out massively better than I was expecting, despite it just being a stupid “blue sky thinking” moment. It’s just a stack of 28 planes, all with a super-simple cloud texture generated by Photoshop, fully translucent and refractive with an refractive index of 1. Job done. 😛 It almost put me in a kind of difficult situation, though. This solution for the mist felt perfect at first, because it was quick to do and worked well, but I’m pretty sure it caused quite a large spike in the render time. And that was no good because if the render took too long, I’d miss the deadline. Thankfully, the render time was still short enough to get it done. Albeit just the 2K version. The 4K version is being rendered as I write. 😀UPDATE!: It’s done now! 😛
Anyway, the RENDER STATS! 😀 At 4K with Rendering Quality 2, it took a total of 15 hours 4 minutes 28 seconds! So, quite a long one, but it was two renders that I combined into one during post. One of the renders had ground mist, which took 10:51:52, and the other one didn’t have the mist, which took 4:12:36. So that mist solution I came up with increased the render time by 150%. Oops. 😛 I still need to try it with caustics on, but I imagine that’ll take at least 24 hours and I still have a secret project to do. 😛 Postwork consists of combining the two renders, masking some nasty dithering with the ground mist, removing some weird highlights on the inside of his cape, smoothing out some stuff the post denoiser didn’t denoise, adding a bit of motion blur to Dragula’s arms, adding an additional layer of fake bloom (even though there was bloom in the scene already, generated by Iray 😆) a levels tweak and a li’l bit of desaturation. 🙂
I’ve got a timelapse for you, too, which you can watch here! 😀 It didn’t take as long as “Roar, Blue Dragon” (just under 20 hours including render time, as opposed to about 70), but it’s still a fairly long timelapse video. Why? Well, because of the cheese-stuffed soundtrack! Seriously, it’s so full of cheese that you’ll think you’re swimming in it. I was only going to put two songs in at first, but I just couldn’t bring myself NOT to put in the third song I found. 😛 The first track is Shadows of the Night, by DJ Bobo. On my 30th burfday, while I was trapped in the car on a journey, my sister played a playlist of birthday songs, and the last one was a track by DJ Bobo, whom we’d never really heard of before (I had, but only vaguely). “WE SING HAPPY BIIIIIIIRTHDAY TO YOU, AND MAY ALL YOUR DREEEEEEAMS COME TRUE!!” Of course, after that, she played nothing but DJ Bobo songs the rest of the way. 😛 It didn’t take long to decide on the next track, because unbelievably, I quickly found out that the same artist has an album literally titled “Vampires”. 😆 So, the next two tracks are from that album! I chose I’m Crazy for the second track, because I am. And because of the cheese 😛 But finally, just when you think I can’t cram in any more cheese, I round off with Dangerous (the video’s a live performance, lmao HOW did this get a live performance, and lmfao he’smiming it 😆). “WE ARE MAD! WE ARE COOL! WE RULE!” 😆😆😆 I couldn’t stop laughing. This album HAD to have been done for a bet. 😛 🧀
I’ve got the progress history here, too:
Six preliminaries this time, which is pretty average, methinks. 😀 Although this doesn’t include anything related to the original pose I made, which you can see in the timelapse. I might render that original pose as a little bonus, cuz I don’t think it was bad or anything. It just didn’t feel Dracula-esque enough. 🙂
Ah, and just in case you’re working about the post title… that’s a reference to a line spoken by Dracula in Monster Bash, a legendary pinball table by Williams. It involves Dracula, The Wolf Man, The Mummy, Frankenstein[‘s Monster], The Bride of Frankenstein[‘s Monster], and The Creature from the Black Lagoon getting together and forming a rock band. 😀 The rule set is just as crazy: you can potentially stack every single feature and play them all at the same time. It has some really funny, well-delivered voice lines (“No I won’t go back to your dog house with you!”), and it has such a great soundtrack, too. 😀 Of course, I haven’t played it for real; I was only 8 when it was released in the U S and A where I don’t live, pinball wasn’t as big in the UK as it was in the US, and it was a pretty rare table to begin with. 😛 I first played it in The Pinball Arcade, and recently picked it up again in Pinball FX 3 on Switch, since I lost all my tables on Pinball Arcade (and they lost the rights to basically all their good tables to Pinball FX 3, looool 😛 ). Anyway, here’s a video of it. 😀
I think that’s all for now! I’ll update this post with the 4K and 1080p renders when they’re ready. But for now, I think I’m going to faceplant my bed. Niiiiight! 😛
UPDATE!! The 4K version finished rendering! I’ve added that to the flavour list, along with the 1080p version. The image itself is slightly different than the one I posted originally; I’d added levels and hue/saturation adjustment layers to the image in Photoshop, but disabled them at some point before I saved it out, and didn’t remember to enable them again, because I’m dumb. Things like this are probably why I can’t get a 20 in Brain Training for Nintendo Switch. 😛 Anyway, I remembered to enable them this time, so the image is a bit brighter and less saturated than before. 🙂
I also updated the timelapse, adding a bit onto the end where I re-do the postwork on the 4K image. Probably not worth re-watching for that. Unless you want to listen to DJ Bobo again. I mean, why wouldn’t you? POWER TO THE VAMPS 😆
Hope things are alright with you! Still apocalypsing it up? 😀 Here in the UK, people are allowed to go to the pub now, or something. I think we’re also allowed to go to the shop without having to queue. Apparently all of this went very well and there were no problems at all, lolololol. 😛 I dunno, I’ve not been watching the news, and the only reason I’ve left the comfort of my flat in the past 2-3 months was to put my rubbish in the bin. 😛 No, I’m not scared of the Death Virus. The only reason I went outside before the outbreak was to go to work. Other than that, I have no real need to go outside, because computers are inside. 😛
So yep, still business as usual for me. Doing work, playing games, making renders, eating, sleeping, repeating. 😀 And on the subjects of renders and games, oh boy, do I have a render for you today! It’s my biggest, most ambitious one yet! And it’s based on a game! A game render! A GRENDER! 😆 NO MONGMASTER NO! 😛
I’m not sure exactly what triggered it, most likely a totally random epiphany that happened while I was showering or daydreaming, but I had this sudden idea for a scene. It was inspired by Dynasty Warriors, probably one of the most polarising game franchises ever. You either love it, or you don’t enjoy hitting thousands of dumb peons with a stick for some reason. I’m the former. 😀 DW3 was the first one I played, and I’ve loved it ever since, although I wasn’t a fan of DW6, and I haven’t played DW9. I really don’t want to know what they did to get such an… immensely negative reaction. 😛 I’ve enjoyed games outside of the core series, as well: the Samurai Warriors crossover game Warriors Orochi was fun, the Monster Hunter-inspired Dynasty Warriors: Strikeforce was interesting, and I quite enjoyed Fire Emblem: Warriors on Switch, too. My favourite game in the series is DW8: Xtreme Legends, though, by a pretty big margin. And even though I enjoyed the older games, I’ll admit, looking back, they were… much, much less good than DW7 and 8. 😛 And the voice acting (well, most of it, some of the actors put on a surprisingly good performance) is still fantastically bad. But that’s part of the fun! Not as much fun as, say, the voice acting in Earth Defense Force 4, but still… 😛
Anyway, to help explain the idea I had, here’s a bit of background for those who might not know. 🙂 There’s a character called Guan Yu: tall, bearded, wrecks face, eventually gets face wrecked at Fan Castle, spoiler. 😛 He’s been my main man since DW5. 🙂 In all games before DW7, one of his weapons was called “Blue Dragon”. It’s also the only weapon that I know of that is referenced in voice lines. In DW5, if you trigger a True Musou Attack while playing as Guan Yu, as the camera spins around, he’ll say “Roar, Blue Dragon!” before starting to pwn every fool in front of him. 😛 He also uses the same quote in DW8 at the start of his Rage Musou Attack. (That link goes to a video I put together for reference. 😀 )
And that weapon was what got me thinking. “What if Blue Dragon was being swung by an actual blue dragon? While wearing Guan Yu’s armour? Holy crap he could be called “Guan Ryu“!” (“Ryu” being Japanese for “Dragon”). Stackin’ wordplay and references like a madman! 😆
I laughed to myself because I am sad. 😛 But even though I loved the idea, I quickly went to dismiss it. “Nah… that would be amazing, but it’s impossible on so many levels.” I thought. “I’d need to rip assets from the game for one thing, and that’s probably not happening. Who’s gonna have written something that can rip assets from Dynasty Warriors games? Hell, even if that was possible and I actually did get the models out of the game and into a format DAZ Studio could understand, I’d need to fit a suit of human armour onto a dragon. That’s the most ridiculous ide… ah, wait, hang on… that thing I did the other day… maybe it actually is possible…?”
“Wait what the fuck don’t be stupid” I hear someone say all the time, but this particular utterance of the phrase manages to get my attention. 😛 Yeah, it sounds totally stupid, but there is something I learned not too long ago. While dicking around and experimenting in DAZ Studio, I managed to figure out the purpose of a tool I never understood before: the Transfer Tool. It basically takes a “source” figure, copies its skeleton onto a “target” figure, automatically sets up weight mappings and everything, then fits the target figure onto the source figure as if it was a piece of clothing. Theoretically, this means that any item of clothing can be fit to any figure, provided that the clothing fits around the figure in the first place.
After realising this, before I knew it, I was at the computer looking for a way to rip assets from Dynasty Warriors 8. 😛 I thought that would end up being the barrier, but surprisingly, I found a solution quite easily. There’s a tool called Steven’s Gas Machine (the executable is called “xentax.exe”, though, shrug) which can rip assets from a bunch of different games, a lot of them being Warriors/Musou games. It’s pretty much as simple as moving some files to a folder, answering some questions, then waiting for a while. There’s also Ninja Ripper, an almost-global solution which hooks into the game you want assets from, and upon hitting F10, it attempts to dump the models and textures loaded into memory at that moment. To view the ripped assets and get them into a usable form afterwards, there’s Noesis, a model, image, and animation viewer/converter. Its interface isn’t particularly great, but hoooooly shit it sure can convert. 😀
After seeing that it was potentially quite easy to get assets out of the game and into DAZ Studio, and that it was theoretically possible to morph a set of human armour to fit a dragon given enough time and effort, I thought “well, I might as well give it a shot, at least”, and began recording for the timelapse. It took absolutely friggin’ ages, and there were several times when I thought “crap, I don’t think I’m gonna be able to do this after all”. But I kept at it… and kept at it… and then, finally, 70 hours and 234GB of video files later… I actually completed this insane endeavour, producing something that didn’t resemble a gigantic poop! 😛
So, I very, very excitedly present to you… “Roar, Blue Dragon!”:
More flavours here, including an alternate version with a HUD! 😀 No HUD: (1080p) (2K) (4K) With HUD: (1080p) (2K) (4K)
YYYYYYYEEEEEEEEEEEEAAAAAAAAAAAAAAAHHHHHHHHHHH!!!!!!! WOOOOOOOOOOO!! 2,000 KOs! TREMBLE BEFORE THE GOD OF WAAAAAAAARRRRRR!!!
*ahem*. So, ah… yeah. This… this went well. 😆 Honestly, I’m still a bit shocked that I’m able to share it with you, because I thought this project was going to fizzle out as I was trying to fit the armour to Dragon 3. But when I finally got it morphed to the correct shape, and when the Transfer Tool just worked, I started to believe it might actually happen. 😀
Maaaan, I’m so, SO happy with this one! 😀 I think it’s my best render yet! I also think it’s a milestone for me, because it pushed my skills beyond their limits, and I did so much stuff to create this that I’d never done before, or never truly believed I could do. 🙂 But hey, who cares what I think, I’m way more interested in what you think! Let me know! 😀
Before I go on to talk about the process, for those who just screamed at me, YES, I KNOW, the weapon he’s holding isn’t actually Blue Dragon. 😛 It’s Heavenly Dragon (or Moon Dragon in Orochi 4). Getting hold of Blue Dragon would have required me to download Dynasty Warriors 6 (which I don’t own) and use Ninja Ripper to maybe get the correct weapon model. I didn’t really want to go through that. Besides, Heavenly Dragon looks way better than Blue Dragon. It’s actually blue for one thing! 😛 I might do an alternate version with the correct weapon if enough people complain. So if you wanna see that, complain like crazy. 😆
And also, YES, I KNOW, Dynasty Warriors is set in China and the “Ryu” in Guan Ryu is Japanese. But hey, Dynasty Warriors is made in Japan, so mleh! 😛 “Guan Long” would’ve probably been more fitting, but “Long” and “Yu” don’t even share any letters, maaaaaan! I’ve gotta take some concessions when it comes to Epic Wordplay. 😆
“And WHAT ARE THOSE THINGS IN THE SCENE?! HUMANS?!?!?!” Yes, I finally rendered those ugly things. 😛 Hilariously, it’s the first time I’ve ever rendered a human outside of test renders. Hell, even in test renders, I’ve only rendered a human once! 😆 I reckon they were necessary, though; without them, it would’ve been a bit less obvious that the image was a Dynasty Warriors reference/parody/tribute. 😀
Anyway, because quite a lot went into this render, I am going to ramble ALL about how I did it, mainly for my own reference, but I do hope it helps someone else at least a little. 😀 But don’t worry, people-who-don’t-care! I’ve got a very convenient TL;DR link which should hopefully skip you right past it. If you’d rather read more exciting things like the RENDER STATS, go ahead and click the link below. 😀 No, it’s not a link to Never Gonna Give You Up. Okay, it might be. Slightly. Mostly not. 96.5% not. 😛
The first step was getting the armour out of Dynasty Warriors 8 and into a format that DAZ Studio could understand. As mentioned above, I used a couple of tools for this: Steven’s Gas Machine and Noesis. Using Steven’s Gas Machine was simple enough, I just pointed it at the game files, and it did its thing, dumping all the assets into a folder. The hardest part was sifting through all these assets, because of course, they didn’t have any meaningful names (see screenshot). 😛 I used Noesis to go through them, eventually finding Guan Yu and his weapons. It was around this moment that I realised that one part of my joke was going to fall over. Blue Dragon hasn’t been in Dynasty Warriors since DW7. 😦 As a result, I had to settle for the next best thing, Guan Yu’s ultimate weapon: Heavenly Dragon.
When I’d found the models, next was figuring out what format gave the most usable results in DAZ Studio. I tried FBX first, because it’s what we use at work and it seems to work well there. I’m pretty sure it’s a common format, so it surely has some decent support, right? Lmfao no! DAZ Studio did literally nothing when I tried importing it. No feedback at all. 😛 So next up, I tried OBJ. This worked, but the model contained a single material group, meaning that it was going to be a li’l bit of a chore to get all parts of that single material to look good.
The last thing I tried was DAZ’s own format, DAE (Collada). Importing this format gave me no options and made Guan Yu appear as a MEGA GIANT, but weirdly enough, he was separated into many different bits. This made it extremely easy to remove the parts of Guan Yu that were Guan Yu, leaving just the armour. However, this was a trap that contributed towards the pain and frustration and swearing experienced later on in the process. 😣 Of course, I didn’t know it at the time, so I just plodded on thinking “yay, easy!” 😛 After getting the models and textures in successfully, I pulled in Dragon 3, creating a fun little “teaser” render for someone who gave me really useful feedback on Dino Disco. 😀
I wanted to increase the quality of the armour before starting to morph it to fit the Dragon 3, but this is where my respect for ZBrush kinda died a bit. 😛 You see, I found that ZBrush, despite its power, has a pretty glaring flaw. It can’t handle multiple material groups. It claims it can, but really, it can’t. I sent Guan Yu’s armour to ZBrush. It had four material groups when I sent it. I then immediately sent it back to DAZ Studio as a new model, without changing anything. The new model had one material group. It just doesn’t seem to understand that a model can have several material groups. If you split the model into several parts and send that to ZBrush, each of the four “SubTools” (ZBrush’s word for “models”, no idea why because you tend to use tools to shape a model, not brushes to shape a tool, lmao 😛 ) then have their own material, and you can allegedly combine them into a single SubTool but retain their grouping info inside “PolyGroups”. But when you do that, ZBrush DUMBBrush just kinda goes “lol fuck UV maps they’re stupid” and throws them away, so when you send that mesh back to DAZ Studio, even if all the PolyGroups have been separated back into SubTools again, DumbBrush gives DS completely bullshit UV information and you can’t reapply any of the texture maps! AAAAAAAARGH DUMBBRUSH YOU’RE THE LAST THING I EXPECTED TO SCREW ME OVER!! 😦 So yeah. I couldn’t do jack shit with the model quality in the end because I had to choose between losing the material groups or losing the UVs. And they’re both kinda very important. 😛
But anyway, despite that annoyance, I’d at least gotten the model into DS. That was the easy part, though. The first major ordeal followed: getting the armour to fit Dragon 3. I understood the basic process: put the target figure (Dragon 3) in the zero pose at world centre, get the source figure (the armour) morphed so it fits visually, then transfer the source to the target using Transfer Tool. It’s just that the process of getting the armour to fit a creature that was definitely not in mind during the armour’s creation process was… very difficult. 😆 To do it, I used ZBrush’s transformation tools, along with the “Move” and “Move Topological” brushes to stretch the model to fit Dragon 3, frequently exporting it and saving my progress as I went along. Thinking back… although it was time-consuming, it wasn’t too painful overall. The hardest parts were the ornament on his right shoulder, and his hind legs. The legs of the armour intersected with each other, and it was frustratingly difficult to pull them apart so they matched the Dragon 3’s T-pose. As for the boots, I just looked at them, brought D3’s foot next to one, said “heh, no way”, and promptly deleted the polys. I didn’t even want to try that shit. 😛 Besides, I think he looks way better without them! 😀
The morphing process took around 6 hours in total, and I was terrified that the Transfer Tool wasn’t going to work when the moment of truth arrived. 😛 But when it did, and when I applied one of the Dragon 3’s preset poses to test it, I totally felt like Success Kid. It was a true “thank fuck for that” moment. 😆
Of course, it wasn’t immediately a perfect fit (not like it ever fit perfectly 😆), as was evidenced when I broke in the outfit with a more suitable test pose, one based on Guan Yu’s run cycle. There were a couple of things that were ridiculously wrong, and a ton of other things that were less wrong, but still wrong. 😛 The ridiculously wrong things were the shoulder decoration on the right shoulder and the medallion on the left. They were warped to a hilarious degree and needed separating from the main armour.
I imported a fresh copy of the armour, deleted the ornament and medallion on the morphed armour, and deleted everything apart from the ornament and medallion on the fresh armour. I used DAZ Studio’s built-in Geometry Editor Tool for that, which I wasn’t super-keen on at first, but it’s actually pretty easy to use, and I ended up leaning on it quite a bit. 😀 The other problems consisted of some polygons (mainly the bits of fabric that move in-game) poking through each other, making the armour look terrible. This is around the time I realised that something wasn’t quite right with the model, and when that easy solution I stumbled upon when importing the model came back to bite me. Some of the vertices on the model weren’t welded together. They occupied roughly the same point in space, but it was possible to move one but not the other, causing the model to “tear apart at the seams”. This plagued me for the rest of the project. It also made it impossible to apply subdivision to the model, so I was stuck with it at game resolution. 😣 I’m not truly sure what I could’ve done to prevent this. There was a couple of steps mentioned in Steven’s Gas Machine that I didn’t do, because I thought they were optional and I didn’t even have Blender installed and I don’t even know how to Blender. 😛 Maybe they were critical to making the model less janky? I dunno…
Anyway, that was the first hurdle passed. Ah, but there was something with the textures, too. Guan Yu’s texture map was 1024×1024, which is okay for a game, but really quite low for rendering. The weapon texture was even worse: 256×256! 😦 I wanted to upscale them, but knew that just making the textures bigger in Photoshop wouldn’t really do much. That’s when, after having a look to see if there was an upscaler that could upscale more better, I discovered the fantastically-named Waifu2x. 😆 It’s designed for images with an anime-like art style (hence the name 😛 ). It somehow supports photos, too, but probably not as well. Despite Dynasty Warriors having a more realistic art style, I thought I’d give Waifu2x a try and see if it could improve things. It ended up being more like Waifu4x, because I increased the texture size of Guan Yu’s armour to 4096 and the texture size of Heavenly Dragon to 1024. 😛 I also used it on the normal maps. It made a pretty big difference. Definitely keeping that tool in my arsenal for the future. 😀
(I had an Image Compare widget/block thing here before, but it was useless because it didn’t let you zoom in, so here’s a gallery instead 😛 )
Of course, no project goes from start to finish without DAZ Studio trying to shit all over me in some way, and this one was no exception! 😛 Around the point I was splitting the medallion out from the armour, DAZ very randomly and very quietly decided to, oh, sextuple the size of my scene. Yes really. The scene was around 3MB, and then suddenly it was 19MB. I don’t know why. Maybe it was bored? Maybe it thought things were going too well and wanted to waste like an hour of my time? Shrug. I only realised there was a problem when DAZ Studio locked up when saving a bit later on, and then locked up again while attempting to reload the save before the one it just failed to save. It had literally shat out scene files that it couldn’t open again! Aaargh! 😡 It’s a good thing I save so much, because I only lost 30 minutes of work because of that, thankfully. Annoying, but at least it wasn’t hours of work. However, it did make me save the scene extremely frequently for a while afterwards as I tried to figure out what I did wrong. I never figured it out, because it never happened again. So I likely did nothing wrong at all. Thanks DS, you massive anus tart. 😑
The second hurdle was creating the peons that Guan Ryu would smack. The minor soldiers that don’t do much other than get owned by you. I call them “peons”. 😛 At first, I thought I’d just use a figure created by DAZ (Genesis 8 Male), but after realising that there weren’t any outfits available that matched the Dynasty Warriors vibe I was going for, I decided to bite the bullet, and began an attempt to rig the peon model myself. This was a massive first for me, because this is something I’ve been wondering how to do for quite a while, but felt that it was beyond what I was capable of.
First off, I had to get the peon into DAZ Studio in the first place. Not quite as easy as I was hoping it to be. 😆 It turns out that Dynasty Warriors 8 randomly generates peons using various models for the head, body, arms, and legs, pinning them together and animating them on the fly. While there were full models of what looked like soldiers you see in-game, their textures were insanely low-res. I imagine they were guys you could talk to in the camp during Story Mode or Ambition Mode. I realised the peons were randomly generated when I kept finding various bits of a peon instead of a full model, and looked at one of my reference screenshots to confirm it. So, I just picked out some parts to assemble one, exported the whole scene from DS, then reimported it again as a single mesh. Job done!
Apart from that I wasn’t gonna fall into the same trap again as I did with Guan Yu! 😛 I wanted to make sure the model was in good shape before I started rigging it, welding together the vertices to prevent the mesh from falling apart this time. I did this using the Auto Weld tool in the only modelling program I had installed at the time: Hexagon, also created by DAZ. It communicates with DAZ Studio, too, which is nice. Sadly, the program has gotta be about 283 years old, and it’s fucking horrible to use, too. All I wanted to do was weld all the vertices on the model to other vertices close by. But the way the UI works is so utterly, utterly stupid. It’s something like… you select the weld tool, and it immediately takes effect. Uh, I think. If you click anywhere on the model after doing this, there’s a chance Hexagon will just crash. But if you send the model back to DS after seeing the weld tool has done what you wanted, there aren’t any changes. That’s because the change you saw in the viewport is merely a preview! To actually apply the thing you want to do, you’ve gotta press the button over to the right in the Properties window (a button that is always there, even if you have no tool selected) that says “Apply”. No, wait, “Accept”! Apply? Accept? Which one is it again…? God damn. I just had to refer to the raw footage used for the timelapse, which I narrate to help me remember things. 😛 “Apply” either does nothing, or shows you the new settings you chose. “Accept” applies the settings for reals and doesn’t allow you to undo afterwards. Wtf. It’s ass. Like 100% pure ass. I can only assume that the people working at DAZ are actually insane. I mean, this is kinda unrelated, but they did send me an e-mail yesterday saying that I could “Get all Michael base models FREE”. Of course, I quickly found out that they’re only free if you pay! Omfg THAT MEANS THEY AREN’T FREE!! Stop clickbaiting me you insane dumbfucks! 😡
*calm* After getting the model into a good state, I began creating a really basic rig. And it was so much easier than I was expecting. 😀 Sure, there were probably a bazillion issues with my rig, but it was good enough for my purposes. 🙂 I used a couple of DAZ Studio’s built-in tools to accomplish this: the Joint Editor Tool and the Node Weight Brush Tool. Both are now in my toolbar. 😀 The Node Weight Brush is annoying, though. The controls aren’t obvious, and sometimes it refuses to let you erase some weight mapping you’ve done. But yeah, it was quite simple to do. Time-consuming, but simple. You basically create the skeleton using the Joint Editor, then for each bone, you paint weight mapping for each axis of rotation. It lets you copy one axis of rotation to other axes, which is great, but it doesn’t let you mirror weight mappings across bones, though. And that’s really annoying. It lets you mirror it over the same bone, but that makes no sense. Rotating the left hand shouldn’t rotate the mesh where the right hand is. 😛
But of course, this being my first time doing rigging, I fell into another trap. And it was caused because I’d welded the model together, this time! Argh! 😣 The mesh for the legs, just like Guan Yu’s, intersects. That is, the fabric on the left leg cuts into the right leg (and vice versa, of course). Because I’d welded close vertices together, I’d actually accidentally welded his legs together, and didn’t realise this until I’d almost finished the rig. What I should have done is figure out a way to exclude certain vertices from being welded, and made sure to exclude those between his legs. Since I was already 90% of the way through rigging, and redoing the model now would cost me about two hours, I decided to chav it and deleted a large section of his right leg. It was a dirty move and I’m sorry. Not sorry. 😛 Screw you guys, I SAVED TWO HOURS OF MY LIFE! 😛
Weirdly, though, even though you weight map specific bones to specific parts of the mesh, there’s an extra dull-ass step you need to take before you can actually select these bones in the viewport. Of course. 😛 You need to assign “Face Groups” to sets of polygons, and assign those Face Groups to actual bones. When you click on a polygon that’s part of one of these Face Groups, the bone assigned to the Face Group is selected. I’ve no idea why they require you to do this step. I have no idea why there’s no automated process for it, because I’ve already weight mapped the polys to the bones. But it was a fantastically annoying waste of half an hour of my clearly-DAZ-doesn’t-think-it’s-precious time. 😛
After the rigging was complete, I put a ton of them into a test scene with Guan Ryu, in a pose that was close to the final thing. Everything finally felt like it was coming together. This was the first moment when I genuinely believed I was gonna get this thing done, and I was SUPER excited by it! 😀
For some reason, though, there was a two month gap between doing that test render and continuing work on the scene. And I think it was because of the VFX, which was the final hurdle that I could see. I’ve done VFX before, sure. Some examples being the cauldron glow and volumetrics of Fire Burn, Cauldron Bubble, the volumetric disco lights of Dino Disco, and way back when I was still a noob, the fire effect seen in Shoryuken. But they were all quite simple, really, while the VFX required for this scene seemed really, really complicated, and it was making me nervous. I wanted it to resemble the moment I was trying to reference in my scene, the last strike of Guan Yu’s Rage Musou Attack:
And I wasn’t sure I could replicate that, or even get close to it. Definitely not with my 2D art skillz. 😛 But anyway, I eventually manned up, and continued on. I was hoping to rip an environment from the game, but after a lot of searching through the game files, I just couldn’t find the terrain meshes. I imagine they they have some bespoke way of storing info about a map, and generate the terrain from an image, or something crazy like that. No matter, I ditched that idea and found an environment from the store that was a good match for the scene. 🙂 I rearranged the peons, set up lighting, tweaked the pose, fixed more costume issues… things were going really well. But all the while, that feeling kept gnawing at me. “Ya gotta do the VFX sooooon! You might Faaaail It because Your Skill Is Not Enooooough!” 😛
When that moment finally came, I decided to refer to the game’s assets to try and figure out how they did it. Instead of searching through about 90,000 images and God-knows how many meshes to try and find the VFX, I used the excellent Ninja Ripper to target specific moments in the game, and managed to rip the VFX textures and meshes fairly easily. They were a lot simpler than I was expecting, so I just decided to pull them into the scene to see what I could do with them. After arranging them into a shockwave kinda shape, adding emission to the material, and adjusting the colour to be just right, I suddenly became all hyper-giddy at the first pass. The filename of the scene saved just after I got the effect in was “Roar, Blue Dragon! 233 – PRELIMINARY PRELIMINARY PRELIMINARY PRELIMINARY PRELIMINARY NOW NOW NOW.duf”. 😆
So yeah, turns out I was worried for no real reason. Of course, something completely impossible and incredibly irritating had to happen anyway, because I’m me. 😛 There was a problem with the VFX textures, where they were clearly very low-res. I’d upscaled them to try and fix this, not with Waifu2x, just a simple bicubic upscale in Photoshop, with a bit of Gaussian blur added on to soften the edges of the VFX. For some reason, this texture rendered differently to the previous one. It’d lost a lot of its intensity. I tried again but without the blur. Same problem. I tried again without even upscaling the image, literally: put image in Photoshop, save out as JPG. Same problem! I made a copy of that image just using Explorer, and the copy rendered the same as the original, but for some psycho reason, even when following the exact same steps I did to create the image in the first place, it rendered differently! It absolutely blew my mind! I didn’t hash the two files to compare them, but there must have been something different. But then again… it can’t have been that different! They were even the exact same size! Saved in the same format by the same application in the same way! Here’s the panel of my redlines I created after this occurred, hopefully it helps illustrate what I mean. It completely destroyed my head. 😛
I ended up fixing the low-res issue in post with lots and lots of blurring and smudging and motion blurring and radial blurring. So it was okay in the end. 😀 Aside from the swipe, the fire spirals and the hit effects, the only other VFX left to do was a dust shockwave. I first resorted to a pack of “shaders” which couldn’t get anywhere near the effect I wanted. In the end, I realised that I could achieve it very easily myself. All I needed was a primitive torus, a simple gradient, and the ground texture that was already in use. Use the gradient as the cutout, the ground texture as the diffuse, and tweak until nice. Simplez! 😀
So, after getting through the VFX, that was it, Luigi. 😛 Got it rendered, applied all the postwork (mostly motion blur 😆), and finally, the scene I never thought was possible was aliiiiive! 😀 😀 😀 As for the HUD? That was easy; I just grabbed the HUD sprite atlas from the game assets, and spent some time matching everything up using a screenshot. 🙂 Rendered a very quick portrait version of Guan Ryu, nothing complicated, stuck it where Guan Yu’s face was supposed to be, and boom, complete! 😀 I did try to convert the game’s charmap into an actual font file, though, like a .ttf file. I gave up after a little while because it was turning out to be way more effort than it was worth. I came across a promising tool called… uh, FontTrace, I think? Unlike the other tools, I’m not linking to this one because I don’t want you to waste your time. 😛 It did nothing but write some useless zero-byte files. Even using the image it provided as a sample. Which was in a format the tool wouldn’t accept. Lmao. 😛 It was a very old and crusty tool, to be fair, so it probably just broke at some point in time. Anyway, I ended up scaling the charmap up so it matched the letters in a screenshot, then just manually copied and pasted the letters I needed, arranging them into the sentence I wanted to make. 😀
TL;DR: I made a render where a blue dragon hits people with Blue Dragon. 😛
Whew… okay… that was the process. 😆 I have a feeling that I missed something, too, but I’m not sure. 😛 If you actually read all that, firstly, thanks for reading, and secondly, if there’s anything I did miss about the process that you wanted to know about, just give us a shout! 😀
I feel that this scene was accidentally well-timed, too, as I’ve very subtly referenced with an easter egg in the HUD version. 😀 Maybe you’ll see it if you’re an ultra-hardcore fan? I’m not, I didn’t even realise until I had an idle thought about it. 😆 But the KO count I used for the HUD version, 2,083, references a special thing that Dynasty Warriors, or rather, Shin Sangoku Musou, will celebrate soon. Let me explain a little first: Dynasty Warriors 1, known as simply “Sangoku Musou” in the East, was a totally different game to the Dynasty Warriors we know now. It was a 1v1 fighting game like Street Fighter or Tekken or whatever. Dynasty Warriors 2 was the first hack and slash of the series, and the series has followed the same style ever since. Instead of them calling it “Sangoku Musou 2” in the East, they distinguished it as a different game by calling it “Shin Sangoku Musou”, or “True Three Kingdoms Unrivaled”. That’s why they’re one number behind the West (DW8 over here is called SSM7 over there, for example 🙂 ). So when Koei/Omega Force celebrate an anniversary of the series, it’s based on the release date of Shin Sangoku Musou, because that’s when the series truly began. And 2,083? 20-08-03. The series celebrates it’s 20th anniversary on the 3rd of August 2020. 😀 Was the easter egg worth reading all that explanation? PROBABLY NOT!! 😛
Anyway, let’s get on with sharing the TIMELAPSE! And as you might imagine, it’s the longest one yet! 😛 It might be a li’l bit more interesting than usual, though, since I did a lot of things I’ve never done before. Although… the length of it might pull things the other way and actually make it more boring than usual. 😛 Anyway, you can watch it here, if you like! It’s only 17½ minutes long! 😛 (I should probably warn about epilepsy here, the video does flash a fair bit since as it switches between darker and lighter windows about 22,000% faster than normal 😛 )
But hey, a longer timelapse means more music, and this time there’s literally no better way to alleviate the boredom than with some of the tunes from the awesome Dynasty Warriors 8 soundtrack! I crammed in seven of them! 😀 I kick things off with “Fan The Flames“, the music played during the ABSOLUTELY BAT-SHIT INSANE intro to the original Dynasty Warriors 8. I thought the Xtreme Legends intro was crazy, but holy crap, the original intro makes that look like a documentary. 😛 (Highlight: Zhao Yun hits a boat with his spear after flying towards it surrounded by flame arrows, and the boat EXPLODES, then all the boats around that ALSO explode, then he gets on his horse while both him and the horse are FLYING THROUGH THE AIR UPSIDE-DOWN!! 😆) After the intro music, I thought that “Welcome To China (DW 7th Mix)” would be appropriate next. I sent that track to my sister just before she actually went to China a while back. Also, whenever she sent some pictures back by e-mail, I always replied with a stupid Dynasty Warriors reference. She didn’t understand. 😛 Next up is “My Bad Fellow“, a classic from Dynasty Warriors 4, but it’s still part of the DW8 soundtrack, played during one of the hypothetical battles. I remember grooving to it quite a lot when first starting out with Dynasty Warriors. I think it was the track played during Pang Tong’s stage in DW4XL. 😀 Of course, I can’t not have Lu Bu’s theme in the timelapse, so next up is “Theme of Lu Bu (DW 7th Mix)“, the only character in the game with a bespoke theme. The DW 7th Mix isn’t my favourite arrangement of it, as they kinda messed with the classic main melody a bit too much for my liking, but it’s definitely got that familiar vibe. 😀 The penultimate track is the epic “Final Game“, played once during Story Mode, but most frequently during major battles in Ambition Mode. I first heard this track on the Dynasty Warriors 8 promo website when I heard the game was coming out, and it immediately made me want the game. 😛 “Wait, MongMaster, why is FINAL Game not the final track, you gorm?!” Well prawn, that’s because I wanted to round things off with my favourite song on the soundtrack, and in fact one of my all-time favourite pieces of video game music: “Heavenly Strike“. I love the way it fuses rock and traditional Chinese instruments so naturally and perfectly, I love the way it builds up several times as the song progresses, and when it reaches the chorus, I just can’t get enough of that epic lead melody. And it’s so Dynasty Warriors. If there’s one song that can describe the series, it’s this. 😀 I also snuck in “Victory Jubilee (DW 7th Mix)” right at the end, played when you win a battle. Because it’s apt, and I love that jingle. 😀 I used the seldom-heard extended version in the timelapse. It’s not in the sound test menu, and you can only hear it after completing a challenge in Challenge Mode, but it fades out. However, if you press a button while “Challenge Complete” is on the screen, it skips to the results screen without fading out the music, so you can groove to it forever. #themoreyouknow 😛
Oh, and if that wasn’t enough yammering about the timelapse, there’s one other boring thing that I want to say about it. 😛 Normally, I have something in the corner saying how many days I’ve worked on the scene, but for this timelapse, I put the exact amount of time, instead. Well, as exact as I could make it anyway. Sometimes I stop recording to take a break, or to refer to something that happened earlier. 😛 That’s why I know this render took about 70 hours to create. 😀 I did this because I felt that showing the days was slightly misleading before. Decorators didn’t take me six whole days, for example. Anyway, that timer was only possible due to some new video editing software I just recently discovered: DaVinci Resolve.
I normally use Adobe Premiere, but DaVinci Resolve is like a hybrid of Premiere and After Effects, with the critical difference that the basic version (which is ANYTHING but basic) is friggin’ FREE. What the actual hell. Why haven’t I heard about this program before?! 😛 Apart from the interface being all different and requiring me to re-learn where everything is, and it being somewhat… unstable at times, it works quite similarly to Premiere, and can do pretty much everything Premiere can do. It did have a terrible start by crashing silently and forcing me to start from scratch… twice… but it quickly redeemed itself and really showed its worth when I started playing with Fusion, the in-built visual effects tool. 🙂 Also, mine is a niche use-case, but I found the whole program super-accommodating for timelapses. 😀 After Effects claims it can only support compositions up to 3 hours long (even though I imported a Premiere project and it was perfectly fine with a 12 hour long one (shrug)), and while Premiere can handle sequences up to 24 hours long, that’s sometimes not long enough. As is evident by this 70-hour long one. 😆 DaVinci Resolve? Yeah, it kinda just shits all over them both and can support both timelines and compositions that are as-near-as-makes-no-difference 256 hours long. 😛 Although in the Fusion visual effects tool, I found that expressions break when you go past frame 1,000,000. Which is weird. Maybe I should file a bug report and give them a good laugh. 😛 I fixed that by using multiple compositions less than a million frames long and setting the start time of each of them differently. 😀
Anyway, enough about the timelapse! Now it’s time for the RENDER STATS!! 😀 Amazingly, this render, even with the environment and lighting and peons, was only taking around 4 minutes to render in the early stages. That kinda changed after adding the VFX and upping the render quality and resolution. 😆 Sooo, at 4K and Rendering Quality 2, rendering it took a GTX 1080Ti and a GTX 1070… 6 hours 23 minutes 38.32 seconds! So, not too bad. 😀 Although in hindsight, I have no idea how it went from taking minutes to taking hours. 😛 Oh, and as for the portrait render done for the HUD version? I have the time for that one, too! It took 15.97 seconds. Lmao. 😆 As for postwork? Aye, there’s a fair bit. I cleaned up some graininess, fixed a problem with Guan Ryu’s outfit on the left shoulder, removed two of the specular highlights on the gem embedded into his helmet, applied motion blur to the swipe effect, Guan Ryu’s right arm, and the peons in the foreground, added a fake bloom effect, and added some focus lines and a vignette. And of course, assembled the HUD for the alternate version. 😀
Ah, I nearly forgot: the progress history! It’s quite… tall, this time. 😆 I ended up doing 5 test renders and 17 preliminary ones before the final render. Here it is, you’ll probably wanna click on it to be able to see it properly. 🙂 Not sure why it’s so blurry before clicking on it, though. Probably WordPress’s fault. 😛
Since I know I can do this kind of thing now, if I ever decide to make a follow-up, and I’m not saying I will (although despite the duration and how much effort was needed, it was still really fun to do, and very satisfying 😀 ), I definitely know who I’ll be having the Dragon 3 cosplay as next. That goon over there. Zhang Jiao. 😀 Why? He’s my second-favourite character. 😀 I love how broken he is, gameplay-wise. If you equip him with the Throwing Knives (which are fully compatible and give him the maximum attack bonus), they can hit an insane number of times, filling 1 unit of Musou in a single combo if the group you attack is big enough. Then you can use that to fuel his secondary Musou Attack, which hits so many times that you can get his Rage from zero to maximum in a single attack. Then, if you trigger Rage and hit Rage Musou while his secondary Musou is still going, you can get to True Rage instantly. Repeat all this, and the amount of time he spends in Rage mode is ridiculous. 😆 Not only that, his voice actor is a legend. 😆 Here, I made a quick video to demonstrate. THE TIIIIIIIME FOR JUDGEMENT HAS COOOOOME!!!! 😆
Anyway, I think that’s all. Hope I didn’t bore you too much, and I really hope you like the render! 😀 I will finish this post with the following GIF. And by saying WordPress is shit. 😛
How’re you doing? Hopefully you’re using your mad Ninja Skillz and dodging all that invisible and ultra-deadly COVID-19 while at the shop only to buy bread, milk, fruit, and Cheesestrings. (cringe) 😛 Me? Yeah, I’m totally fine, better than usual, actually! I really don’t get what all the panic is about. It’s no worse than the flu, right? We get that every year, but that doesn’t cause people to have a massive panic attack and buy 40,000 toilet rolls and hand sanitisers. 😛 I dunno, I completely shut myself off from the news months ago, and Adbook not long afterwards. It’s all scaremongering and fake news. Sure, I’ll stay away from other people out of respect (even though I don’t have the virus, and even though I know they don’t have the virus, because if they did, they’d be at home dying, because apparently almost no-one recovers, lmao), but I’ll think for myself, thanks. 😛
I’ve not really had to adjust to the “always stay inside” rule because I do that anyway. 😛 There are only two differences, the first being that I now have to work from home, which I greatly prefer because it’s quieter, because I’m an anti-social bumface, and because I don’t have to give the bus company that I hate ~£12 a week just to get to work. 😛 But the second difference is that I now have to spend an hour and a half getting to the shops and back, and can only carry about one week’s worth of supplies, because everyone in the country is panicking and suddenly wants to use online grocery delivery and all stores are booked solid for three weeks and people are staying up past midnight just to snipe a friggin’ delivery slot for three weeks in the future. God damn really? 😑 But yeah, other than those things, it’s business as usual for me; just staying inside with my best friend (the PC 😛 ) and playing many many games. Although some spaz is trying to scare people out of that, too, by posting an unrealistic model of what gamers will look like in 20 years. Lol, 20 years? I’ve had my gaming PC for about 12 years now, and I’ve been gaming longer than that, but I still look pretty much the same as I have done for a while. So in about 8 years I’m apparently going to be fat only in the lower part of my torso and have a long head with a crevasse in it? Lmfao those “researcher” gorms need to get out. But when they do, they’d better stay 2m away from other people. 😛
Anyway, yep, the world has gone crazy, but I’m still the same. Still crazy. 😛 Despite me trying to ignore everything, I couldn’t bring myself to ignore the very appropriate but potentially tasteless render idea that appeared in my mind the other week. Because it was funny. And funny is good. Usually. 😛
It took some self-nagging to bring myself to get started on it because of what I thought was going to be a mahoosive pain. The idea required me to construct a mound consisting of many duplicates of a single object, and just thinking about doing it was boring. 😛 I was also unsure about how it would turn out when I had done the really boring part. But nevertheless, I kicked myself up the arse, reattached my leg, and started work on it on Easter Monday. Then, just the next morning, I was greeted with THIS!
More flavours below, including a special, stylized, dark flavour! Regular: (1080p) (2K) (4K) Dark: (1080p) (2K) (4K)
Because OF COURSE!! Morons hoarding a ton of toilet paper got you down? Fix it with a dragon hoarding a ton of toilet paper! 😛
Hahaha, despite the mild pain that was the first 30 mins or so of creating this (and the moderate pain that followed), this was so worth it. 😀 Hell, what little pain I did feel while creating this render was pretty short-lived anyway, because this scene only took about 7 hours to make, which is the fastest I’ve been able to create a fully-fledged scene in quite some time! 😀 Hell, if you include the render time and postwork (render stats spoilers! 😛 ), it only took about 9 hours, start to finish! 😮
So, that “mound consisting of many duplicates of a single object” I mentioned earlier? Yeah, that was all the toilet rolls. I ended up referring to it as Mt. TP. 😛 I put a roll of toilet paper in the scene and created 100 instances of it, arranging them into the base of the mountain. To help, I created a sphere which vaguely represented the final mound, and very roughly posed the Dragon 3 so he was sprawled on top of it to ensure it was about right. Then I realised I didn’t need to create the “core” of Mt. TP, I just needed the shell, so I rearranged the toilet rolls so they followed the perimeter of the guide sphere. It wasn’t too bad when I got into the rhythm of it, and it wasn’t particularly hard. Just… dull. Needed way more than 100 rolls, too; the final scene ended up having 343 rolls in there. 😛 It only took about half an hour to finish the basic structure, though, so that was good. 😀
After that, it was just a matter of posing the Dragon 3 on top, which took a bit longer because I couldn’t quite decide how I wanted him to lie. I wanted him to be sprawled out, but at the same time, I wanted him to look happy and cozy, so I settled on a hybrid of the two, where his upper body was curled and he was resting his head comfortably on his claws, but his hind legs were sprawled out a bit, accidentally kicking a couple of rolls off Mt. TP. But he clearly don’t care. He stupid-comfy. 😀
Deciding on an environment was a little tricky. In fact, I already had a shot at making this scene a few days before doing it for real. During that session, I managed to narrow my environment choices down to a few options, the last one being “make a custom cave”. About 5 minutes after experimenting with ways to do that, I quit, because everything was being a dick. 😛 The rocks I was trying to use all had inverted normals so you couldn’t see them in the viewport without using negative scale, and instances of those objects with the same transform parameters were appearing in a different position for no reason. 😑 Anyway, thankfully I didn’t have to resort to creating my own. I found a little nook in one of the environments I chose which was just what I was hoping for. 😀
I learned something pretty strange while doing this scene, though: a “solution” to the inverted normals problem I had. Let me explain a bit: each polygon on an object has a “normal” which indicates what direction the polygon is facing. If this is towards the camera, that’s okay. If the polygon is facing away from the camera, however, that’s called a “backfacing” polygon. Usually, programs save processing power by not rendering backfacing polys, but some programs decide it’s necessary that the user sees them, and do render them. DAZ Studio is speshiful, though. It does render backfacing polys, but instead of drawing the material on the poly as normal, it simply makes the poly solid black, regardless of lighting! Neither of DS’s core renderers (Iray and 3Delight) have a problem with showing backfacing polys like this; they render perfectly fine. But when all the viewport renderer bothers to show you is a silhouette of the object, it is a problem! A massive problem! ARGGGH!
There’s an option in the preferences called “Backface Lighting”, but it appears to do shit all. But hey, it’s fine, cuz I managed to fix the problem by… heheh, God damn… sorry, I… I still can’t believe the way I fixed it. It’s so, so dumb. 😛 Okay, so… DAZ Studio has several viewport render modes, most of which are never used unless you’re really weird or have some kind of need for it. Two of those are “Wire Shaded” and “Wire Texture Shaded”. They’re helpful if you need to see the polygons of your scene, or, well… they would be useful, if they didn’t completely tank the viewport framerate. Seriously, it takes DS about 10-20x longer to draw the wireframes than it takes to draw the object on its own. 😛 For some insane, impossibly stupid reason, though, in these modes, if you have the “Backface Lighting” option on, backfacing polys actually render properly. I just… why. WHY!
Of course, if you do decide to use the “Wire Shaded” or “Wire Texture Shaded” mode with the Backface Lighting option on just to get around the problem where backfacing polys are black no matter what, you then have to deal with the ludicrously low viewport framerate. But hey, there’s a stupid way around that, too! If you go to the Draw Settings window and set the wireframe opacity to 0, yay! Framerate becomes normal! So uh… if I’m in Wire Texture Shaded mode but with no wireframe… WHY exactly does that act differently to Texture Shaded mode?! What are you smoking over there, DAZ?!
Ugh. Whatever. 😛
I also poked dForce again in an attempt to get some stray sheets of toilet paper lying in the scene, dForce being DS’s solution for cloth simulation. It was my first time using it properly, and it’s, uh… shit? Yeah, “shit”, that’s the word. 😛 Maybe it’s okay for clothes and stuff? I dunno, I don’t render humans, so I don’t render clothes. But seeing as how the following can happen sometimes when trying to apply cloth simulation to a simple plane, I can’t see how clothes would work any better:
Maybe I was using it “wrong” or something, but seeing as there’s almost no documentation for this feature (yet again!), I’m not exactly sure how I can learn to use it. Maybe I should render a human with dForce clothing just to see if I can figure out how it works?
Anyway, I think that’s all I have to say about the process. 🙂 Despite DS’s stupid brokenness and dForce’s stupid brokenness, everything went pretty smoothly, which is always nice. 😀 But now, it’s time for the render stats, which may have already been spoiled! 😛 Despite all the toilet rolls, thanks to instancing (maybe), this scene didn’t need much memory, so both my GPUs were able to party this time! 😀 So, at Quality 2, the 4K render tooook… 1 hour 35 minutes 20.36 seconds! Pretty nippy! As for postwork, I covered up some poke-through on the toilet paper hooked onto the dragon’s horn, fixed some weird artefacts on a different sheet of paper, added motion blur to a few rolls of toilet paper, blurred out some noise that the Post Denoiser didn’t quite denoise, tried fixing some nasty-looking texture seams on the cave wall, added a very mild bloom effect, and added a vignette. 🙂
I got a timelapse for you, too! Since the render didn’t take long to create, it only contains two songs this time, both by the very funny and very crazy comedy/metal band Psychostick. 😀 I’m glad it only has two songs because I couldn’t think of a third. 😆 Anyway, first up there’s The Dumb Song, because everyone is being dumb and hoarding craptons of toilet paper for no good reason. Hoarding it won’t stop you from getting the virus. 😛 The timelapse ends with Fight To The Death, because of one perfectly appropriate lyric near the end of the song, which I also used for the post title. 😀 You can watch the timelapse here!
I also have the progress history below, and it’s very short! 😀
Only four preliminary renders before I felt like I could render the final thing! 😀 After the first preliminary render barely showed any issues, I got very suspicious and triple-checked it to be sure I wasn’t losing it. 😆 It never goes that well! 😛
Anyway… I think that’s all from me for now! Hope you liked the render. 🙂 See ya later, don’t die and stuff, and enjoy your pasta and toilet paper casserole or whatever you do with those things that I can’t buy any more because everyone else is suddenly buying all of them. 😛
(in all seriousness, plz stay safe, you’re too cool to get owned by a microscopic virus 😆)
I AM! I haven’t heard even a hint of that stupid song yet! I can taste victory! 😀 😀 😀 UPDATE!!:I WON!!
Anyway… harrooo! Merry incoming Christmas! Or whatever it is that you celebrate! 😀 Exactly like last year, when this is posted, I’ll be sliding down a snowy hill on a toboggan in an attempt to beat my sister to the bottom, and losing every time because my toboggan is crap! 😛 Saw lots of people crashing and wall-riding last year because they were too dumb to slow down and/or turn for the corner, which was funny. Hopefully there’ll be more of the same this year. 😀 Oh, and we’ll be sitting in giant inflatable rings and sliding down some ice, too. 😀
I also went to Cadbury World in Bournville with the whole family a couple days ago. 🙂 The best part was… well, I was gonna say the shop, but no, the best part was an accident that happened during one of the shows. One of the actors was sat on a barstool-style chair when the top cushion part of it kind of collapsed, allowing the frame of the stool to tear through the fabric and pelt him (seemingly) directly in the gonads. 😆 He rolled with it so well that it almost felt like it was staged, but it can’t have been. That stool looked totally phwoacked afterwards, and the wicked witch completely fell out of character and could barely stop laughing. Understandably, of course. It was hilarious. 😀 The second best part was the shop, which is now my new favourite shop. I was expecting it to be a total ripoff, but holy crap, there were some really great deals in there. I ended up buying way more than I thought I would. 😀 The rest of Cadbury World? Nah. Don’t bother. Especially with kids. Unless you’re crazy about history or the Cadbury brand (as a company), it’s suuuuper-boring. 😛
Anyhoo, aye, despite having the uninteresting history of the Cadbury brand forced into my ears, it’s been fun. I hope you’ve all had fun so far, too, and I hope you get some nice presents tomorrow! 😀 And hey, I’ve got a slightly early present for you: a Christmas render! Whether it’s a nice present is totally up to you, but hey! PRESENT! Don’t look a gift horse in the mouth! Its breath is terrible! 😛
(where did that stupid saying even come from 😛 )
So, this year, I was a bit worried, cuz I couldn’t think of a solid idea for ages. Usually I have something solid before I start working on a render, because I really don’t like scrabbling around in the dark, so to speak. It’s not fun to work on a vague idea for hours just to realise that it ain’t gonna work out well, or at all. 😛 But I did eventually think of one, and amazingly, it was DeviantArt which kick-started the idea this time.
One day, when I decided to log in just to see if I’d happened to get any views on my work in the past few months (I hadn’t 😛 ), I noticed that they were doing an event where they take Christmas cards made or decorated by artists and give them to people who are in hospital over Christmas. If you participated, they’d give you a free month of Core, too. This was interesting, because it felt like the first time I’d seen something on DeviantArt that I might actually be able to participate in. Normally all they do is competitions clearly aimed at high-level artists. But just as I was thinking about doing it, I realised. I don’t give a damn about DeviantArt’s stupid Core membership, I probably wouldn’t be able to get the render done in time, the price to send a letter overseas would probably be way more than I’d be willing to pay, and I don’t really consider myself to be an artist. They might’ve rejected it for being too weird or not arty enough. I dunno if they look gift horses in the mouth or not. 😛
But the event did make me think… why don’t I do it anyway? Why not base my scene on something that would look nice on a Christmas card? And hell, why don’t I make actual cards out of it and give them to people I know, people who will probably appreciate it? 😀 Kind of an odd idea, really, because I haven’t done Christmas cards since primary school. I think they’re a bit pointless. 😛 Oh well, my brainworks in odd ways, so I guess the idea isn’t so odd for me. 😛
So, I booked my remaining days off work, and when they came around, I used them to work on the render. Then the weekend, too. And the odd day after work the next week. Eventually, after tons of fannying about, I ended up with this! BEHOLD!
Oh maaan, I’m so happy with this one. 😀 And I’m glad, because as you might’ve deduced from what I said up there, it took way longer than I was planning. I initially thought it was a simple enough idea that I’d be able to do it in two days, but nope! It ended up taking six. Not six whole days, of course. I had annoying things like “work” and “social life” that I had to deal with in-between. 😛
Although thinking about it, overall, it probably didn’t take that much longer than normal. I didn’t spend the entirety of every day doing the render, so… it probably only took about 4 whole days, as opposed to my usual 2-3. I think it took longer because of indecision and the need to keep adding things to the scene, not because of things that went wrong. 😛 The worst thing was choosing the environment, and at first, I thought the render would never see the light of day because of it. I tried one which looked quite good and that I felt would work, but when I set it up and rendered it, I found that it didn’t look anything like the promo images… which is a very common occurrence, but still irritating. 😛 The next environment I tried didn’t look bad, but it was mahoosive, and all the detail seemed to be focused on one small part of it that I didn’t plan on using. Thankfully, I got hold of another environment which was a much better fit. I ended up combining the HDRI from one of the other environments I’d tried with the geometry from the new environment, which worked very nicely. 😀
After that, it was just tweaks and mostly unnecessary experimentation. 😛 The most time-consuming one was probably the custom morph I did for the wyverns. There’s a supermorph which transforms the Dragon 3 into a wyvern, but I’ve never properly used it because it never looked “right” to me. The body’s fine, and I don’t mind the bulkier muzzle (even if I do wish it was optional), but the feet really bug me for some reason. I always saw wyverns as having three front toes and optionally one rear toe, so they can grab things easily, kinda like a bird of prey. But this supermorph keeps him as a four-toed dragon, but with an inflated index toe tipped with an oversized talon, and it just makes him look stupid. Like some sort of deformed velociraptor. 😛 So, I made a really rough extension to the morph in ZBrush, in which I scale his fourth toe down to zero (because one cannot simply remove it without rerigging the whole model, and I have no idea how to do that 😛 ), and make all his talons big. Massive overkill, because I reckon no-one will give a damn, and the morph is rough as hell, but hey, I like it and I think it looks better, so shush. 😛
There were other annoying little things I had to do, too. Because the tree and decorations were all one big mesh (not separate), I had to extract some baubles from it to place in the claws (and maws 😛 ) of the dragons, as well as in the decoration box. I ended up having to cut the top off a crate using DS’s lame-o geometry editing tool to make said decoration box, because I didn’t have one that could open. 😛 It was also really tricky to make the tinsel look nicely shiny and metallic. I had to use glossy anisotropy and set the rotations just right to get the effect I wanted.
I got the A5 version done in a few days, and that was printed onto the cards I sent out to people, but it didn’t actually end up being the final version. Because of time pressure (I had to get something reasonably close, otherwise the cards wouldn’t be delivered in time), I’d forgotten to add mist into the background, which made it look like the foreground and background were melding together, even with depth of field. Also, I then realised that when I extended the picture to 16:9, the left and right sides looked… dead. There was no life outside the middle third of the image. So, yep. There was still more work to be done! 😛
I came up with some nice “scene extensions” for the left and right sides, and learned a way to add mist to the scene. I basically just used a bigass cube set to be completely refractive, with a refractive index of 1 (the refractive index of air), a high scattering distance, and a very, very faint light blue scattering colour. It pretty much doubled the render time, but it was worth it. 😀
Wow, that’s a lot of blabbing, but I’m done now! 😀 Lots of tweaking and additions, but I think it was worth it in the end. What about you? Do you like it? 😀
Now for the RENDER STATS! And… I don’t have them! 😛 Not right now, anyway. I know that the render was done in 4K at Quality 3 (one more than I usually do) and it think it took about 12 hours. I’ll get the accurate numbers when I get home, and update later. 🙂UPDATE!!: Lmao oops. I completely forgot, and now the render time’s been truncated from the DAZ Studio log file. 😆 I’ve also since deleted the raw footage used for the timelapse, so the only number I have is from the timelapse itself. Rendering in 4K, it took around 18½ hours to reach 95% convergence at Quality 3. Normally I do Quality 2, but this time, my early test renders were completing really quickly, so I thought I’d spend a bit more time for a tiny increase in quality. When I shifted to 4K, added the new dragons, and especially when I added the mist, the render time spiked, but I didn’t go back to Quality 2 in response. I was probably thinking “naaah it’ll be fine, placebo quality go go go”. And I was probably stupid-tired, too. 😛
As for postwork, there isn’t much, really. The main things are the subtle falling snow and the vignette, but I also removed/masked a few distractions and tiny mistakes, added a cooling filter, and tweaked the levels. I do want to tweak the vignette again because I think the desaturated corners are way more distracting than I intended, but that’ll have to wait until I get home. 🙂
Since the render took longer than usual to create, I made a slightly longer timelapse for it, too. Sorry. 😛 But if you’re still interested, you can watch it heeah! 😀 As always, it comes with vaguely related music. I didn’t put any crappy Christmas music in the video because I can’t stand it, and all the winter-related songs I could find were all friggin’ depressing, so I had to resort to using songs from games I’ve played. 😛 First off, there’s a very wintry-sounding piece from Pokémon Mystery Dungeon: Gates to Infinity’s final dungeon (I think): Glacier Palace (Great Spire). Spoilers. 😛 After that, I stick with the wintry theme by playing Snowy Mountain from Sonic Adventure before kicking it up a notch with the snowboarding music from the same game: Be Cool, Be Wild and Be Groovy! 😛 Then, there’s a song featured in Hatsune Miku: Project Diva F 2nd: Decorator by livetune(kz), which I chose because the name of the track describes exactly what’s going on in the scene. 😛 I finish with Wonderful Dream (Holidays Are Coming) by Melanie Thornton, which is both full of cheese and just makes me think of the old Coca-Cola adverts, preventing me from taking the song seriously and making me laugh every time because I have a broken sense of humour. 😆
Also, here’s the progress history, too! You’ll probably need to click on it to enlarge if you wanna read any of my scribbles (unless you have eyes of GOD 😛 ).
I think that’s all for now. Time to schedule this post… if I can figure out how to do it on this mobile app! 😛 Have a great Christmas, and a happy new DECADE! 😀 See yaz!
UPDATE!! I added a few bonus pics to the post in an attempt to make the huge, likely boring wall of text a bit more interesting and less eye-bleedy. 😛
Hiii! Happy Halloween! Or All Hallows’ Eve! Or Allhalloween! Or All Saints’ Eve! Or Saints Row Eve! Not that last one! 😛
How’s your day been? If you’re a kid, I hope you obtained/will obtain many free sweets from creepy strangers! And if you’re an adult, I hope you successfully make it through your evening without any kids ringing your doorbell. 😀 Me? Well, I bought a box of treat-size sweets with the words “ideal for trick-or-treat” on it. And then began to eat them all. 😛 Like hell am I givin’ out sweets to random kids that interrupt my evening! Not like anyone’s ever buzzed my doorbell on Halloween, anyway, and not like I’d answer it if they did. 😛
Ah, seeing as it’s Halloween, I’ve got a spooky story about something that happened recently. Well, it’s not really spooky. Or particularly interesting. 😛 But anyway, for some reason that I can’t quite figure out, one of my videos on YouTube – just one – has seemingly gone viral, because the view count on it has exploded. It’s the last stage in my Streets of Goong series, Stage 8. I finished that series about 9 years ago now, and although it didn’t get a monstrous amount of attention, I was still really humbled by the response. It got far more views than I ever expected it to, around 5K-14K per video. However, all of a sudden, Stage 8 has gotten a monstrous amount of attention, achieving more views than all of my shitey videos combined: 636K, and it’s still rising. :O What shocks/scares/confuses me the most is that seemingly because of this, my subscriber count has jumped from about 100 (which I was genuinely amazed by) to 2,000 (which completely blows my mind), and it’s still going up! I don’t get it! That series is 9 years old! Why’d I get so many subscribers from it now? What’m I gonna do with them?! And why is everyone only watching the last video in the series?! That’s not even the best one in my opinion! 😛
Heh… typical, ain’t it? Back then, I had a ton of free time, so I used to post crap on YouTube fairly frequently, but it only got a relatively small amount of attention. Now I barely post one video a year, and suddenly I’ve got this massive audience for the stuff I posted ages ago. 😛
Y’know, talking about this viral explosion has just reminded me of another one I had a while ago, when a Starbound dev used one of my videos in a news post on their blog for a reason that didn’t really make any sense. The difference was that that ended up being very annoying, mainly because of the fanboyish comments I got which were directed straight past me as if it wasn’t even my video they were commenting on. Instead, most of the comments on the video were directed at the dev who shared the video, in a tone that falsely insinuated that they were bestest buddies, and in a way that made it sound that they believed the dev created the video or something. I didn’t mind much at the time, because I felt honoured that one of the Starbound devs thought my video was funny and random enough to share on their blog for a joke, but now I just get super pissed off when I think about it. Because Starbound, that borderline scam of a game that I backed when its first round of crowdfunding opened (with three others, and I was the one who suggested it, which made me feel bad when it was delayed to the point where we nearly forgot it existed), turned out to be a gigantic pile of turd and one of my most hated games of all time, easily described with words such as “slog”, “dull”, “untested”, “late” and “iwannaplayterrariainstead”. I know that last word isn’t a word. That just symbolises the developers’ lies. I don’t want anything to do with the shithead company who created it and if they shared any of my videos again I would immediately take down the video and re-upload it with external sharing disabled, like a kind of Chucklefish exorcism. Lmao “Chucklefish”, more like “Chucklefuck“. 😛 I encourage everyone to pirate every goddamn thing that has their name on it, and here is a gif just for those shithead devs:
ANYWAY, off-topic, MongMaster, VERY off topic and very ranty. The readers don’t care because they’re too cool for that, so get back to Halloween. 😛 Yes, despite the start of this post looking like some kind of trick, I didn’t come here to do that. I came to give you a treat*, in a kind of Soviet Russia-esque trick-or-treat! And that treat* is… a Halloween render! 😀
*your perception of whether this “treat” really is a treat may vary 😛
I didn’t do a Halloween render last year for some reason. In fact, I didn’t even realise I hadn’t done one last year until this year. 😛 Maybe I didn’t have any ideas for it, or maybe I was busy. To be honest, at first, I wasn’t sure if I was gonna create one this year, either. I wanted to, but couldn’t think of any ideas for days. It was probably just because my brain was a bit fuzzy, cuz when I decided to get serious, sit down, and brainstorm some ideas, I got one quite quickly just by thinking of stuff associated with Halloween. 😀
I ended up writing down “pumpkins”, “the Moon”, “werewolves” and “witches”. I didn’t feel that a pumpkin carving scene was good enough, and I felt like I’d already used the Moon in my last Halloween render, Ghost Stories, with its “ghost stories under the moonlight” theme. I liked the werewolf idea, but decided to put that on the back-burner in favour of a scene based on witches. 😀 And of course, you can’t really think of witches without thinking of the famous “Song of the Witches” scene from Macbeth, which I ended up using as inspiration:
Double, double toil and trouble; Fire burn, and cauldron bubble. Fillet of a fenny snake, In the cauldron boil and bake; Eye of newt and toe of frog, Wool of bat and tongue of dog, Adder’s fork and blind-worm’s sting, Lizard’s leg and owlet’s wing, For a charm of powerful trouble, Like a hell-broth boil and bubble. Blahdeblahdeblah.
I’d say it’s more of a “poem” than a “song”, and that’s only part of it; it goes on for three verses, which is about two verses too long, and the second verse (written above) is the least annoying and most readable of the three. 😛 IMHO, of course. I really hate poetry like this, and I’m not a fan of Shakespeare and his broken English (yes I know how stupid and ironic that sounds 😛 ), but I can’t deny that it’s a classic scene.
Anyway, so I finally had a solid idea in my head, but I still felt kinda unsure whether it would come out good or not. The mini-Arnold in my head said yelled “STAHP WHINING AND DOOEET NAOO”, so I picked a weekend (the only weekend left before Halloween :P), sat my ass down, accidentally stayed up until what was effectively 6AM (because of daylight savings 😛 ) and came up with THIS!
As always, there’re more flavours here! (1080p) (2K) (4K)
DRAGON WITCHES, BITCHES! WOOOOO! 😀
Ahhh, this one came out so gooooood! XD Honestly, I thought I was going to Fail It because My Skill Was Not Enough, but the basic composition of it turned out to be really close to how it appeared in my head. 😀 Beyond that, the little details and pose decisions, such as the jar of eyes and the right witch looking unsure about the recipe, just happened on the fly, as they tend to do. 😀
Not much went wrong during the process, this time. In fact, things went unusually well, which made a nice change after all the difficulties and indecisions I had while creating The One True Egg. 😀 Despite not having a clue about how to create textures properly and not having any of the tools for the job, I created a relatively decent (for me 😛 ) texture for the eyes in the jar. Stacking up about 40 of them in that jar also went really well; it wasn’t too painful, and it turned out great, especially since I thought they wouldn’t show up very clearly in the final render.
After having the central witch stir the cauldron and the left witch add ingredients, I was struggling to think of what the right witch would do at first. I came up with two ideas, and posed both of them to compare and decide:
I ended up going with the second one, because I liked the pose more, it matched up with the rest of the scene a bit better, and I felt that it had a bit more humour to it, in a subtle kind of way. 😀
The only technical problems I came across were pretty easily fixed. The first issue was with the “Instancing Optimisation” setting, which I’ve had a problem with before, back when creating Monster. Hunter. (see right). For some unfathomable reason, when set to “Memory”, it mangled the texture of the Dragon 3’s left eye, even though both eyes used the same material!
Turns out that this time, it was caused by having Instancing Optimisation set to “Auto”, whatever the fuck that means. 😛 I would’ve expected that “Auto” would select between “Speed” and “Memory” depending on how much memory the scene was eating, but no, it appears that Auto does something completely different to the other two options!
This “Auto” option seems to have been added in DAZ Studio 4.12.. or it might’ve been 4.11, but I never actually clicked the dropdown for that option in 4.11 because I never had a problem with it while using that version. Anyway, it was a minor annoyance. I just picked “Speed” and everything was fine. 🙂
The second problem was that I wanted to try dForce, DAZ Studio’s cloth simulation, to get some realistic collision on the hats. I’ve never really used it before because I only render things that don’t wear clothes (that’s dragons, dinosaurs, and the occasional mammal, before you go and make a funny 😛 ), but seeing as I got a dForce-compatible witch outfit so I could have the dragons wear the traditional witch hat, I thought I’d try dForce out. Turns out that the hat, despite being part of the “dForce-compatible outfit”… wasn’t actually compatible. Trying to make it compatible by Doing Stuff(tm) ended in a funny way:
So, yep. I didn’t bother using dForce after all. 😀 I just placed the hats as best I could, and did some morphing in ZBrush where necessary. 🙂 Part of me thinks I should have at least tried to dress them up in the witch robes, but it would’ve probably taken an immense amount of time to get right, and I only had one weekend to finish the scene. I can’t bring myself to pose after work; I feel way too tired and unmotivated. 😛
Also, I was originally planning to use an emissive material for the contents of the cauldron, but I actually found that using a point light and placing it inside the cauldron contents created a massively better effect. Probably because the emissive material just made the whole mesh glow with the same intensity, whereas putting a physical ball of light in the middle of the mesh caused some smexy refraction and stuff. It was a nice surprise. 😀
Anyway, now for the render stats! This was kind of an odd one. At Quality 2, the 4K render took 14 hours 14 minutes 28.57 seconds to reach its iteration limit of 25,000, getting to 73.11% convergence, which was about 22% shy of the target. That’s never happened before! Normally, a render at least gets close to the convergence target before it hits that many iterations, but this time, it was way off! I might have to do another render both overnight and while I’m at work to see if the outcome is any better, but hell, even at 73% convergence, it came out well! UPDATE: Before this post was scheduled to go live, I did another render, and this time it went on for about 27 hours before hitting a doubled iteration limit of 50,000, reaching a convergence ratio of 83.76%. It was increasing by about 0.01% per minute by the time it ended, and as convergence increases slower the higher it gets, it would probably take several more days to reach the target of 95%. 😛 The result of all that extra GPU burning was a microscopic increase in quality that I only noticed by flicking back and forth between the old render and the new one, so I felt that it was pointless to update the final image. 🙂
As for postwork, there was quite a lot this time. I added motion blur to the eyes falling into the cauldron, corrected a massive blunder with the front of the cauldron (there was a very obvious seam caused by the smoke effect from the cauldron, so I had to do some colour tweaks to mask it), did my best to try and make the dragons look more female (painting dark lines around their eyes, it was the only thing I could think of, and I really don’t think it worked 😛 ), removed some nose horns and minor light distractions, added a subtle fake bloom effect, smoothed out some noisy areas, added a vignette, and did some levels and brightness tweaks. 🙂
So… what do you think? Do you like it? Spotted something that looks off? Know an easyish way to make a Dragon 3 look more female? Lemme know! 😀 Honestly, I’ve done it before, but I don’t know how. I’m pretty sure it was a miracle that I made Emerald look as female as she did when I created Amai. 😛
Hmmm… I’m not sure why I use OneDrive to share my timelapses. Maybe I should put them on YouTube… 😛 Oh no no no, I know why, because if I put them on YouTube, I’d get slapped for using copyrighteted musics. 😛
Ah, also, as is also now tradition, I drew notes over my prototype and preliminary renders to keep my thoughts straight and to try and prevent me from forgetting anything, which makes a nice little history of how the render evolved. 🙂 Here it is:
Anyway, I think that’s enough from me for now. I hope you enjoy the render, and the rest of your Halloween! I’m off to trick or treat myself again. 😛 Bye!
‘Sup? Hope everything’s well on your end! 😀 Me, I’ve been somewhat busier than normal, as of late. My mum came down to visit a few months ago, and we visited London because we think Brighton is a boring shitehole. Yes, London is also kind of a shitehole, but at least there are things to do there. 😛 Anyway, that was good, but then the month after that I went to London again with my friends. Then the month after that, my sister came down to visit, and we did an escape room, which we cleared with only 4 minutes left. 😀 We learned afterwards that it has quite a high failure rate, so we were pretty chuffed with ourselves considering there was only two of us. 😀 It was really awesome, and I totally recommend it if you ever find yourself in Brighton for some reason (have a look for Bewilder Box). Oh, and we rode on some bikes along the bike lane on the seafront, too. It’s been at least 9-10 years since I rode a bike, and I’ve never ridden one in public before, and it was genuinely terrifying, especially since the bike lane was about 4 inches wide in places and the bikes were god-awful. I constantly felt like I couldn’t control the thing. 😛 Anyway, two weeks after that, I had another weekend with my friends, this time mainly just hanging out, playing board games and vidya games.
So, this excessive amount of human interaction, combined with work and lack of sleep due to the insane heatwave (it got so bad that I actually bought an air conditioner 😛 ), means I’ve felt pretty burned out as of late. Uh, no pun intended. 😛 Various bits of technology have decided to start being a dick to me recently, too. My printer randomly stopped working immediately after someone printed something with it last weekend, so I had to buy a replacement which happened to have the most irritating Wifi setup process ever. The new Windoze 10 update was near-impossible to install, and when I managed to do it, all it seemed to do was slow my PC down a lot and break Groove Music. And the Wifi on my router has randomly started plebbing out and refusing connections for no good reason. And my idiot firewall recently decided to ignore all its rules again, blocking things I told it to allow and allowing things I told it to block. It’s all so very, very annoying. 😛
To help myself chill out, I decided to book a week off work so I could sleep a lot. And it’s gone way too goddamn fast. 😛 And I haven’t really used it to chill out or sleep as much as I planned, because I’ve spent more than half of it doing a new render. But this is a good thing! I like doing da rendarz. 😀
This time, I decided to make something based on a weird office meme that started recently, between me and a few others. The story goes something like this: one day, one of my co-workers said “Lewis, can you test my branch?” from across the office. Lewis, one of our QA guys, somehow misheard this to a degree no-one could ever have imagined, replying with “What, egg and be true?” Shortly after that, I drew the best egg I’ve ever drawn on the nearby whiteboard, and wrote “Egg and be true” next to it.
Aaaand that how we began a cult. 😛
I’ve been meaning to make a render based on it since then. I came up with an idea for it while bored and daydreaming during a totally unnecessary meeting, but never found the time to put it together. So, with this week off, I thought I’d spend a day creating it, but because of things, that day turned into like 3 days. But I’d say it was time well-spent, as it is a joy to honour the One True Egg! PRAISE BE:
Ahhhh, so much zen! 😀 This was a really fun scene to do, and the pose was interesting to create, too, but when I began, I wasn’t sure it was possible to do it with Dragon 3. Sure, he can stand upright on his hind legs, and he can look pretty good when doing that, but having him sit down on his butt is a totally different story. Not only does his tail get in the way of him sitting down, there’re bend limits on his legs which prevent them from crossing over like that, and I had no choice but to break them by a massive amount (they’re twisted about 130 degrees, which is 50 degrees more than the limit 😛 ). It makes his thighs look really warped and weird, but I think I got away with it here because of the camera angle. His shins and feet block most of the problems, thankfully. 😛 There were other problems too, like his left foot intersecting his right calf, and the fact that his biceps looked huge when raising his arms. I had to fix those by creating some custom morphs using ZBrush.
Oh, and speaking of the raised arms… that gesture the dragon is doing with his claws is our form of greeting to members of the cult. Originally, the dragon was holding the egg in front of him, but I thought performing our greeting gesture was more apt. 😀 It’s performed by forming the top of a triangle with your hands and moving it over your head. The triangle represents the truth, and your head represents the Egg. The motion simulates the Egg piercing the truth. 😛 Nope, that was the QA guy’s idea, not mine. 😛
Originally, I expected to spend an evening doing the posing and lighting, have it render overnight, and do the postwork the next day. Annoyingly, though, a lot of things went wrong, and I kept having second thoughts and random ideas along the way, so I ended up spending way longer on it than I planned. 😛
On the first day, I had the scene assembled and the pose was pretty much finished, although it took a bit longer because I had to re-pose his arms after getting the idea to use the greeting gesture. The lighting wasn’t going great, though, so I had to spend another day on it. On that second day, though, I started getting a lot of random ideas. I changed the candle holders to be black on one side and white on the other, I tried out making a gloss map for the dragon’s tail to show that it was wet from being in the water (which didn’t go very well for some reason), added rim lighting behind the dragon, and struggled with adding lights around the egg to simulate that it was glowing. That’s when I remembered seeing something called “Canvases” in the advanced render options, and thought that it would let me render different parts of the scene separately so I could do some things in post that I wouldn’t normally be able to do very easily. Lol no. 😛 When rendering a Canvas, you can only render one at once, it seems to render them in a totally different way than normal, and it spits out some random garbled junk in addition to your render, which very, very vaguely resembles what I wanted to render:
yay whitespace I can’t get rid of, thanks wordpress, you piece of shit
A long time after rendering these, I realised that these are 32-bit images, which probably explains why the files were so massive. Really not sure why they display like that, but whatever. Converting them to 8-bit gives you something that looks better, but at the same time… no it doesn’t. 😛 It looks awful. Like… all glowy and shit. And the background is gone, too. And judging by the fact that two of each type of candle rendered as a black silhouette in the left image, it seems that Canvases don’t support instanced objects, either:
wordpress is shit
And of course the Canvases feature isn’t documented. 😛 So I gave up on it. It’s really annoying, though. It feels like this would help with things like motion blur, and open up a lot of post-processing possibilities, but nope, no documentation about it, and it’s impossible to figure it out by poking it like I normally do. Maybe I’ll fight with it again in the future.
Anyhoo, after ditching Canvases, I then had the idea to make the egg gold instead of silver… then as I was doing that and trying to decide whether I liked gold or silver more, I accidentally made the egg transparent and thought “AH, A GLASS EGG!”, so I did that instead. 😀 I think it turned out well in the end, but making the glass egg caused another problem which ended up costing me a lot of time. I first noticed it when adding a light behind the egg to re-create some nice-looking water reflections that I’d lost after switching the egg to be made of glass. Doing this made me discover a very… uh, interesting quirk with the Iray shader. If you set a light emitter to have geometry, and place it behind something that’s refractive… even if the light emitter’s geometry is set to be hidden, the refractive material will render it. Basically… you can see invisible things in refractive materials. 😛 This meant you could clearly see a black circle in the egg, that black circle being the emitter I was using to get the water reflections around the egg cup.
Trying to think of ways to fix this without cheating pretty much drained about 4 hours of my time. 😛 I even went back to the Canvases feature in desperation, trying to figure it out so I could fix the issue I was having, but nope. In the end, I had to cheat and render the scene twice, once with the emitter on, and once with it off, and combined the two images in post to get the reflections without the black circle. It made me feel a bit dirty. 😛
After realising that cheating was my only option, and fixing some problems where the lighting changed completely when enabling the Caustic sampler, I was ready to go. But of course, I had to make two more foolish mistakes. 😛 Firstly, I forgot to increase the maximum render time again. DAZ Studio has idiotic defaults where it will stop rendering regardless of the convergence rate if either two hours pass or Iray does 5000 iterations, and you still can’t change the defaults! It’s very rare for any of my renders to take less than two hours, goddamn it! (angry emoji) But my second mistake after that which was even more annoying? I trusted a friggin’ plugin I’d discovered a few days before. 😛
It’s a render queue. Basically, something DAZ Studio should have, but doesn’t, because they’d rather spend years adding “strand-based hair” and other much less useful shit first. 😛 As you can probably tell, the render queue lets you render one scene after another without input. I was a bit skeptical, but in my tests, it actually seemed to work, so I trusted it with rendering these two scenes. When I came back to it the next day, it claimed it had finished, but I could only find one render. That’s when I realised that whoever coded this made two massive schoolboy errors!
My scenes were called something like “The One True Egg 45.1.duf” and “The One True Egg 45.2.duf”. What this idiot plugin does is look at the name of your scene file, find the first dot, and discard everything after it to get the render file name. Presumably, it does this to strip the file extension, but for fuck’s sake, this logic is completely wrong!! When you do that to the scene names above, you get “The One True Egg 45” for both of them! You’re supposed to check for the last dot and strip everything beyond that! I didn’t notice this idiocy before rendering because, as you can see in the screenshot above, the window didn’t show the full target path! I just assumed the target paths would be correct, because why wouldn’t I?! But after scrolling across the window after about 8 wasted hours of rendering, BOOM! That’s when I realised that the coder was a dumbass!
And hey, as if that’s not enough, it doesn’t even bother to check if the file it wants to save already exists before saving it! So the plugin just goes “yep just save with that file name I’m sure it’ll be fine” and shits all over what was already there without giving half a damn! Oh my God, why do you not check if the file already exists?! Both of these errors are so ridiculously basic that all I can imagine is that the author wrote the code, then never looked at it again! Never considered any kind of edge case! That just seems like an “I don’t give a crap” attitude, to me. So yeah, I was pleasantly surprised by this plugin at first, and was gonna keep it after I’d confirmed it working in a real-world situation, but hah, not any more! No chance. I just can’t deal with half-arsed plugins that waste my time. I’ve already asked for a refund. 😛
Anyway… I’m calm. I’m calm… FUCK YOU I’m calm. Come on, MongMaster, zennnn… sit like the dragon and praise the One True Egg… thaaat’s better. 😛
That was the last thing that went wrong, thankfully. 😀 After that, I re-rendered the scene four times (twice for the character-focused version, twice for the desktop version), and all was good and stuff. 😀 At Quality 2, with the Caustic and Architectural samplers on, the character-focused render (6.7MP) took 7 hours 6 minutes 20.2 seconds in total, which was much longer than I initially expected. And that was using both GPUs, because the scene was quite light on GPU memory. As for the 4K render (8.2MP), it took 3 hours 52 minutes 29.3 seconds in total, which makes NO SENSE. It has 22% more pixels but it rendered 80% faster! :S Maybe it’s something to do with the image dimensions? I dunno. Magic. 😛
As for postwork, that consisted of merging the two different renders together to get some nice light around the One True Egg’s eggcup without having a black circle within the Egg, blurring out some noise that the post-denoiser didn’t get for some reason, adding some glint to the dragon’s eyes, tweaking contrast, levels and saturation, and adding a vignette. Here’re the two different renders before postwork, so you can hopefully see what I mean:
fix your shitting editor wordpress oh my god
Continuing the recent tradition, I created a timelapse of this render, too, adding some mildly-related music behind the video. 😀 I really struggled for tracks this time, though, because for SOME reason, no-one wants to sing songs about eggs. 😛 Still, I just about managed to scrape together enough egg-related songs for the video. First off, there’s “E.G.G.M.A.N” from Sonic Adventure 2, followed by “Do The Funky Chicken” by Rufus Thomas (sorry, not sorry 😛 ), and finally, “Egg Shaped Fred” by Mansun. 😀 I was tempted to include “Tails and Dr Eggman Fight but Dr Eggman Dies In An Explosion” by Sonic4ever on there, but I thought I’d use Eggman’s theme instead. 😛 You can check the timelapse out here, and watch me keep changing things and flail around with Canvases and slowly lose my sanity (not like I had much in the first place 😛 ). Yes, I’ve tried embedding the video in the blog post again, but it’s still not working. And I don’t want to use YouTube because they’ll be all like WAAAAAAAH COPYRIGHTED MUSIC, even on unlisted videos, and even though I’m not profiting off using it. 😛
Oh ya, also, like I did with Dino Disco, I drew some notes over my preliminary renders that pointed out problems and things that were going well. I’ve found that this really helps keep my thoughts straight and helps prevent me from forgetting something that I need to fix before the final render. And when the render’s finished, it becomes a nice history of how the render evolved over time. 😀 Here it is:
I also thought I’d try doing something else new and dumb with my renders, for cheap lols important research purposes. 😛 I’m not sure if you know, but Microsoft released something a while back called CaptionBot. It’s an AI which looks at an image you give it and tries to describe it, usually with terrible results. 😛 So, I thought I’d feed CaptionBot my renders and see what it says it is. I’m hoping to be surprised at some point, but I’m mostly just expecting it to fail. 😛 Here’s what it thought about this render:
Boooring. You could’ve at least tried to describe it. That would have been funnier. 😛 But hey, it recognised that something was sitting down! Not sure where it got “cake” from, though. 😛
Oh, but there’s something similar to this that I’ve found really impressive, though: Clarifai. It does the same kind of thing, but instead of trying to form a sentence, it just spits out a list of words that match the concept of the image, along with a “confidence” value. It has all sorts of other features, too face detection, a colour breakdown of the image, and can even determine if it’s NSFW… and now I have a huge urge to try and break that feature for some reason. 😛 Aside from the explicit content detection, I’m not exactly sure what kind of serious use-case all this stuff has, but anyway, there’s a section of the site that lets you try it out, and it’s amazing how close it can get:
“Meditation”, “Zen”, “Spirituality”, “Candle”… there’re so many accurate terms there, it’s crazy! :O However, it didn’t see “dragon” (slightly understandable) or “egg” (bit of a blunder there 😛 ). And yes, it says it’s SFW with 96.8% confidence. 😛 I might make this a thing. CaptionBot vs Clarifai! 😀 Ahhh, I’m such a gorm. 😛
Anyway, I kind of went off-topic at the end there, but I hope you like the render! Let me know what you think. 😀 For now, though, I think I’ll stop writing junk and go and shoot some more giant insects on EDF 5, which thankfully came out on Best Console recently (thank you Sandlot/D3Publisher/whoever else was involved! 😀 ).
Egg and Be True, everyone! 😀
UPDATE!! Firstly, many thanks for the likes! Honestly, it means a lot. 🙂
A couple days after finishing and posting this render, I found myself thinking about it. To me, this scene represents light, and the good side of things… but what about the dark side? I wondered… could I pull a 180 on the atmosphere and make a dark, evil version of this scene?
To find out, I decided to revisit the scene, and began a corruption spree. I took everything I could think of, and flipped it around. White turned into black, sky blue became blood red… and before long, I found it: the Dark Side of the Egg!
So, so EVIL! 😀 I just took the words “evil” and “corruption”, and ran with them. The environment changes consist of the red lighting, the red “Dark Egg”, flipping the candle positions around and giving them red flames, as well as darkening the rocks and sand around the edges of the pool. As for the dragon, aside from the very obvious black scales, I made his eyes and wing membranes red, gave him new horns with a modified texture, made his facial expression more aggressive, lowered his arms, lengthened his talons, and gave him a bit more bulk. Good experiment, methinks! I had fun with it, at least. 😀
I’ve added it to the Render Flavour List under the main render near the top of the post, but of course, you can see the full version by clicking on it above. 🙂 I have the desktop version of the render in progress, and I’ll post it at some point tomorrow. I had an idea for a dual-monitor desktop wallpaper version of this, as well. Nothing special: just a simple compilation of both renders, side-by-side. If it doesn’t end up being a turd, I’ll post that here, too. 🙂
As Ulala would say: staaaay tuned!
UPDATE!! (2) I created the desktop versions of Dark Side of the Egg. You can get them from the Render Flavour List above, where the main render is. 🙂
I also created a basic dual-monitor version, where both renders “blend” together slightly. However, I realised that it’s not going to work for everyone unless I make about a million different versions of it. 😛 For two monitors of the same resolution, it’s easy, but if the monitors are different resolutions, it gets tricky.
I made Double 4K, Double 2K and Double 1080p versions, though. 🙂 I also made a 4K + 1080p version to put on my desktop at work, and if you have a look at that version, I think you’ll understand what I mean when I say it gets tricky, and why I decided to stop there. 😛 Having said that, if you do want a specific version for your monitor setup, just ask, and I’ll create it for ya when I get some time. 😀 Unless it’s triple monitors or something crazy like that. 😛
…when a 👫 o’ 👨s, they were 👆✌️🚫👍, 🌟🐻 makin’ trouble in my neighbourhoo– yeah, that. 😛 (I had to write that on my phone, cuz that’s where all the emoji that the kids use these days are. 😛 )
Harro! How are you this fine weekend, or weekday if you didn’t read the post on a weekend? 😀 Me? I’m good, nothing particularly notable’s happened recently. Although I do need to phone Virgin Media because they “need information” about something or other. I’m upgrading to 350Mb from 200Mb after finding out that these titfaces have been overcharging me for pretty much no reason at all. I went onto their site the other day and found that I could upgrade to 350Mb for 13p more a month than I was paying for 200Mb. Shitheads. 😛
Anyway, whine bitch moan no-one cares. 😛 I do have something you might care about, though; a brand new, fresh, shiny, tasty* render! *oooooooooooh* (*not edible)
This one came about from this bit of conversation with a friend:
This was awesome, because no-one’s ever requested a render from me before, and that alone makes me happy. 😀 But anyway, dragons playing basketball? Hell yeah! Totally my style! 😀
…Apart from that I didn’t have much of a clue about basketball. 😛
All I knew is that some very tall men in some vests bounce an orange ball around and try and get it through a hoop to score variable amounts of points depending on how far away from they hoop they were when they threw the ball. 😛 So, I did some research, because I probably wasn’t going to get anywhere if I didn’t. 🙂 I searched for pictures of people playing it for pose ideas and mostly just found pictures of basketball riots, then tried to watch some NBA game replays for scene composition ideas but was denied by Fox because I don’t live in America, so I settled for some random friendly match (which didn’t seem very friendly 😛 ) on YouTube that wasn’t in English. 😛 I also read about basketball rules and a page containing shit-tons of lingo on Wikipedia.
I was originally looking for some kind of lingo where I could play around with the words (or not) to make a double entendre (the clean kind 😛 ) for the title. I was originally going for “monster jam”, but that’s just something I heard on Rocket League once, and it’s more commonly known as a motorsport event featuring monster trucks. But one of the terms on that Wikipedia page ended up giving me the general idea for the scene: “fast break”, where players try to advance the ball and score as quickly as possible, so I just went for that with the title, too.
Man, this one was so much fun to make! 😀 I was originally thinking of a scene with some action near the hoop, like a dragon leaping up to score while another leaps to defend, but after some playing around and discovery of the term “fast break”, I really liked the idea of a dragon confidently running up the court while the defenders prepare to intercept. It gave a much clearer focal point. 😀
This render contains the largest number of subjects I’ve ever rendered in one scene, including test renders: five main and technically four in the background, so nine in total! :O When I say there’s “technically” four in the background… that’s because I used DAZ Studio’s instancing system. I created four unique audience members and created two additional instances of each of them, which (apparently) saves resources when rendering. This means there’re actually 12 dragons on the bleachers, but eight of them are duplicates. I did this because… well, you’ll almost certainly never notice, because they’re waaaaaaay off in the background, miles into the camera’s depth of field annihilation zone. Would you have known if I hadn’t told you? 😀
I also realised there was a practical problem… team identification. I wasn’t going to give them clothes because A: I don’t like the idea of that, and B: the only items available are armour and saddles. 😛 I was thinking of spraying player numbers onto their scales, but I didn’t go through with that because I thought it’d look shitey, and I’d had a better idea anyway: armbands! I thought they’d make it very clear without overdressing the dragon and making them look weird. 🙂 So, I grabbed an armband that was intended for humans and hammered the shit out of it using my pathetic ZBrush skills, making a unique morph for each dragon so it would fit well. After doing armband morphs for four of them, I realised… you were only ever going to see two of the armbands. What a pleb I am. 😛 Ah well.
As for postwork, I did quite a bit of it this time. I applied motion blur to the red dragon’s legs and his left arm, as well as the green dragon’s legs. I also added motion blur to the ball. I would’ve done this in Iray rather than Photoshop, but you can’t. 😦 3Delight supports motion blur, but I have no idea how I’d get the motion blur out of that and slap it on top of an Iray render without it looking like… like 💩. 😛 I also removed the nose horns from the red and green dragons, because you can’t remove them from the model itself… well, okay, I’m sure you can, but that would involve things like UV mapping which I don’t fully understand yet. 😛 Finally, I added a normal vignette, and then perhaps most obviously, some manga-like “focus lines” as a secondary, stylized vignette. It took me ages to decide whether to go through with those focus lines, because I found them really distracting. Eventually, I softened them up and made them very transparent, which made them kind of “blend into” the image a bit more, complementing it and guiding your focus rather than stealing it. 🙂 Also, instead of my usual “typed” signature, I used the signature I made for my sister’s Amai canvas print. I’m probably gonna do that from now on. 🙂 For some weird reason which makes no sense, I… kinda prefer not having the title of the art there now. Shrug. 😛 Might be because it looks a bit cleaner.
Oh, oh, yeah! There was something I was supposed to mention after doing Amai, but I totally forgot to mention it then, and I almost forgot to mention it here, too! 🙂 Ever since starting to use Iray, I’ve had this irritating problem with the renders it spits out, where certain areas – usually reflective or out-of-focus – come out as grainy, fuzzy chunks of poo. The way I’ve dealt with the problem has evolved over time. Initially, I used a Gaussian blur in Photoshop, then began to use a combination of 0.6 radius Gaussian blur followed by a 0.6 radius Smart Sharpen. Most recently, I’ve been applying a stronger blur and sharpen to a Smart Object version of the image, then painting the effect mask depending on how strong I wanted the blur in certain parts of the image. Now, though? In the DAZ Studio 4.11 Beta, they’ve added this godsend, called the Post Denoiser:
Just click to make it available, click to enable it (not sure why there’re two steps, lel, it’s not a nuke 😛 ) and boom, it kills that noise like the police killing a noisy party. 😛 It can even make Iray viewport renders more useful by removing the initial noise early into the render so you can get a good idea of how the render will turn out in much less time! 😀 One thing, though: the default start iteration is 8, and while that’s great for viewport renders, I don’t know whether I can trust a number that low for final renders. If the denoiser starts early, it makes the image appear very, very messy early on, and I’m not sure if this is improved as time passes. For Fast Break, I first performed a render with the denoiser off to get a vague idea of how many iterations it would need. After it reached about 33% convergence, I stopped the render and set the start iteration to about the number of iterations it had performed, which, for this render, was about 2500. Thinking about it, though, I used the denoiser on Amai, but never adjusted the start iteration, and that turned out alright. So… maybe I’m worrying for no reason. 😛UPDATE 2!!: Yep! I did an experiment and found out that having the Post Denoiser on from an early iteration doesn’t make any difference to the final quality of the render; it comes out just as sharp as it would if the denoiser was set to begin at a later stage. So it’s a win-win; a more useful Iray viewport experience and no noisy renders! Yay! 😀
Anyway, whaddya think of the render? 😀 I hope you like it as much as I enjoyed making it. Special thanks to Ben for the request! 🙂 I now understand basketball just a tiny bit more than I did originally. 😛 To me, it’s like a crazy “sleight of hand” game, with all the insane passing techniques and dribbling tricks and stuff. 😀
Oh, crap, of course! Render stats! Yaaaaaaaay! 😛 So, at 4K, with Rendering Quality 2, 98% convergence and everything nice on (Caustic and Architectual samplers), my 1080Ti and 970 pushed out a render in 9 hours 42 minutes 34.23 seconds, which, when you consider the 1080Ti, is a really long time! Some maths would suggest that on my old cards, this render (if it had been able to fit in the GPU memory, which it wouldn’t!) would have taken about 18 hours 16 minutes! :O I LOVE MY 1080TI! Best purchase ever. 😀
Right, I’m gonna bounce on outta here. Yeah, I went there. 😛 I’ll leave you with one of the songs that started playing while I was making this render. I found it funny when it came up, because the lyrics say “make more rhymes than Michael Jordan make slam dunks”. XD
UPDATE!! First off, I want to say thank you for the likes, everyone! It means a lot! 🙂
While I was creating this render, I had PlayClaw recording with the hopes of creating a timelapse video for no apparent reason. 😛 I’ve tried doing this a few times in the past, but the videos always had something or other wrong with them, usually PlayClaw or Shadowplay silently crashing when I started a render, or sometimes for no reason at all (Shadowplay -.-). This time, though, the videos recorded with no problems, so I stuck them all into Premiere and made a… er… 5258.17% timelapse video, which you can check out here, if you’re interested! 😀 Almost everything is there, excluding the initial scene setup bit (because I forgot, oops 😛 ), and the bits where I stopped recording to have a break to eat dinner or send messages or talk on the phone or poo. 😛
So, it’s my sista’s burfday today. 🙂 A while ago, when I posted Amai (the render) on Jokebook, she mentioned that she’d have it as art, as in, to put on a wall. I lolled to myself and thought “yeah right, I don’t even consider it to be art”, but as I’m a (un)funny man, I jokingly replied with “burfday present sorted”. 😛
Turns out she was actually serious, though, so with no other burfday present ideas… I got it printed on canvas for her. It arrived way earlier than I expected and she opened it EARLY, NAUGHTY GIRL:
I did a special 8K render for this, which may have been massive overkill, but hey, the site had a 40MB upload limit and I wasn’t going to take any chances. 😛 I’ve added it to the previous post, just in case someone wants to check it out. Made a proper signature too (took me half an hour, top lel 😛 ). It was printed by HelloCanvas (not an advertisement, just a recommendation 😛 ), and it looks like they did a good job! I haven’t seen it in real life, but from what I can tell, it came out pretty much as the preview suggested it would.
I’m glad she likes it so much! 😀 But she’d better not feel like she shouldn’t take it down when it becomes an eyesore. 😛