What are you waiting for, just click read more already… yeesh.
Well, it’s quite ironic that after all this time, I didn’t made a “Toothless only” post, just now. Heck… He is most of the reason that I’m even here. (My nick name is also inspired by him). I’m waiting for this so long, and I’m so glad that I can finally make one, and I can’t wait to do another one!
Let’s begin.
After I watched HTTYD for the first time (almost 4 years ago) and soon my addiction reached a point, I really wanted to make some 3D animations about him. I had done some pretty nifty 3D stuff back then, (like Desert Wars) so I had the knowledge to do it, but the only problem was, there wasn’t any good 3D models about him.
I've searched every possible model sites, and aside from some very rough or basic Toothless models, there weren’t a single one, that I could work with.
But one day, I found out there is one! But it was in Source Film Maker. Well, I wasn’t gave up just yet, cause I read somewhere with some clever tricks, you can import that model into 3DS Max. The matter of fact, the model was very likely to be made in 3DS Max! So it must have a way!
I downloaded SFM, and the Toothless model, and I was shocked how awesome he was!
 |
| This was my first session of SFM. It dates back to 2014.09.13., about one and a half months after I saw the movie. |
As you can see, he’s very detailed and the rigging is quite decent as well! Sadly I didn’t found a solution to bring him back to Max just yet, two years had to pass until it could happened.
About 2 years ago, when I was randomly searched for Toothless stuff (as my daily routine) I found another possible lead. This time, it was Garry’s mod. There was another model of him. Not that detailed like in SFM, but it was okay.
 |
| Daaaaw’… He looks so cute!!! <3 (His eyes, were originally yellow, but I changed them to green, because they supposed to be GREEN!) |
Garry’s mod was easier, cause there were extractors for the file types that it uses. But, I didn’t forget SFM Toothless. That one was so much better and also had a saddle. I dedicated myself and I wanted to work with that model, so another round began. I searched tirelessly for another source, and eventually, I found the solution. Around that time, someone ripped the dragon models from a HTTYD game called Wild Skies, and to my greatest surprise, it was the same model as in SFM! It had the same detail, the rig, the saddle, everything. I was very happy, and I downloaded instantly. Only downside was, it wasn’t available for Max, only Blender. Well, that’s a bit unfortunate, but I could export him to 3DS Max. And I did. Luckily it was successful and I had Toothless finally in Max no time. But there were drawbacks. The rig (or skeleton) was optimized in Blender. Which really sucked, because when you import the .obj file into 3DS, the bones of the rig will be converted into dummies, which are much harder to work with. Animating with them is even harder if not impossible. You CAN do some very basic animation if you want, but I hope you have a lot of free time and patience, cause you will need it! Another sad fact, that when I played with the rig, I soon realized that it wasn’t linked to the model properly and when I wanted to make more complex poses, the model got very distorted. I don’t say, it’s impossible, but takes a lot of time and you need to adjust things which wouldn't be necessary if the model was rigged properly.
I gave up the animation part, but at least finally I could make some poses, with a lot of trial and error. I made all of my 3D fanart by using that model. One of my first posts on this blog, which was dedicated to Usertoothless’s awesome fanfiction series Dsplash, concluded around 20 carefully posed images of him. Making one pose took me 2 to 7-8 hours! But overall it was a lot of fun, and I was very happy that finally I could work with Toothless! :)
I made several other fanarts with that technique, but I’ve got busy and didn’t had time to make more. (That was the start of a time period when I really struggled to do anything else really and it lasted until just recently).
Meantime, around November, I found something very interesting! I searched again and I found out that, now there is a way to import the SFM model into Max. Finally! I tried it out and it worked! There were good news and bad news. Good news, the rig is MUCH better and it came with proper bone system, but the bad news is, the facial expressions wasn’t working at all. I really wanted to have them in 3DS Max, cause it would open up a lot of opportunities.
And this is what led to me to the half year delay. Not a pretty fun story.
So, this is Toothless in SFM.
You can see there is a list at the left, which contains a lot of control slides that you can operate. These manipulate certain muscles on his face. You can figure out what each of the sliders do. They not just control the facial expressions, but also his claws and the foot pedals.
I made Toothless angry here and you know how much time it took? About 2 seconds. Previously, when I wanted do the same look on the Blender model, this would take about half an hour or more. And also I would ruin the model for that scene, (unless you make key frames for the control points) because there is no way that you can manually move them back where they were originally.
Also, the expressions are very nicely done! They are very accurate, and you can mix several ones at once! Absolutely fantastic!
The sad part, you can’t bring these into Max… Why? Well… this is where things start to be complicated. But, I try to make it clear as possible.
The model consist of little points, which called vertices (vertex for singular). If you connect all those little points, you’ll end up with a (poly)mesh. If you connect at least 3 of these dots, (without overlapping) you’ll get a polygon. (That sounds familiar doesn’t it?) Now. Every model has a different amount of vertices (and thus polygons). It depends how detailed is. This model has 10.296 vertices. They all have a definite position in a coordinate system. If you move one of them, the model will deform. If you select the vertices that makes up (for example) his upper eye lid, and move them down, it will look like if he want to make a blink or just get bored. These control slides are bound to different groups of vertices. They have a default position and a modified position. The slide is gradually moving them to one position to another.
If you understand that, that’s very great, cause you just learned the fundamentals of character animation! :)
This is where it gets complicated. Try to follow me, if you can.
The facial expressions are stored in a file called toothlessws.vta. It only contains the node ID-s of each of the bones and the vertex coordinates of the different expressions, which are a bunch of numbers, hundreds of pages long. Now. There was an importer for 3DS Max for this .vta file, but when I imported it looked like this. (Warning, disturbing!)
Yes… Not a very pleasant sight. But if you look down to the timeline, (highlighted with red) you can see 41 key frames. Each of the them, represents a different facial expression. And all those key frames are linked to a morph operator (does the same stuff that the slider menu in SFM, - highlighted with blue). It would worked just like in SFM, but as you can see, the model is totally messed up. BTW, when I try to move the sliders, the mesh is changing as it should. I found out that the problem was that the importer didn’t aligned the vertices correctly, so you have to do it manually. Someone from the plug-in’s forum thread, wrote a little Java program for it, and was kind enough to publish the source code. Sadly, I know almost nothing about programming, I can’t even figure out the basic logics most of the time. Fortunately, I had a friend who’s an expert in Java language (He has 15+ years of experience). I showed him the source code, but he told me, that this is very bad coding, and probably it was written for that specific model that the guy was repaired and it require a totally different program for mine. Also, he still didn’t understand how the program aligns the vertices fully. (Which I understand, it require a good amount of 3D skills to that too). So that's probably not going to work. I waited a couple of months, maybe the guy will come up with a solution, but he didn’t. So I went in, and did almost all the expressions myself from zero. It turned out very nice, but it was a lot of extra work, which could've been went into making short animations instead, if the thing works from the beginning. Recently, I made almost every one of them, and now I can finally start to make some proper looking animations about Toothless! This model has everything! Can be used in 3DS Max, has a proper bone system, plus facial expressions!
I would like to compare the two Toothless models to each other and you will see how much better is the SFM version compared to the Blender one. You don’t need to be an expert in 3D, don’t worry. It’s quite self-explanatory. :)
So here are the 2 models. Can you see any difference?
If your guess is no, then you’re right! There is (almost) none! (Fortunately)
How about now?
Now, there is some difference sure, but from a technical point of view, it’s day or night!
Let’s go into more detail!
I start with his head.
As you can see, in the previous version there are a LOT of control points. Most of them are working, (kinda) but they are far from perfect. A few of them doesn’t do anything. (Not even in Blender.) Another annoyance was, that the points wasn’t named properly either. But the biggest problem was the points were bunched up in one spot. Selecting the head or neck was really hard! The points on his face allowed to do some facial expressions, but because they were slightly off, you had to be extremely careful. Especially with the eyes! There are 2 control points in his lower and upper eyelids. Animating them would be a disaster! Sadly, I can’t show you in pictures how terrible it is, but if you’d tried it yourself, you will also get to this conclusion. He has also 2 control points in his big ear flaps. They are much easier to work with, and you have a lot of freedom manipulating them. The two smaller ones on top of his head, has only one control point, and both of them are moving together. You can adjust the side ones too, but only the upper one, for some reason. His jaw is working fine, and also his retractable teeth and tongue. No problems there.
In the new rig, things are much simpler… waaay too simpler. If you didn’t made the facial expressions yourself, you’re in trouble! There are no nodes for the eyes, eyebrows, etc., so you have to manipulate the vertices to have facial expressions (which is a pain). But if you have them, it’s better in all means! Let’s see, what we have in this version.
We have a very distinct neck (yellow) and yaw (blue) node! The head node (brown) works like a charm! Much better than what the old rig had! The tongue is also working fine (red). As for the appendages, we have two very nicely done nodes for the big ear flaps (cyan), and they are flexible as they were in the previous model. There are some changes for the smaller ones though. The two between his big ears (green), has one node for each plate! So now, you can animate them separately. Awesome! The side ears (light green) unfortunately, have to share only one node for each side, but at least all 2 side ears are working now (even if they’re moving together). BTW, I love these ears/plates/flaps! They're such a great touch to his design, and makes him look so much more unique! I also love how he uses them for his expressions! I adore them, and can’t wait to animate them! :)
The next section is the torso.
In the new rig, his spine is looking pretty much the same as the older version, but he has now 3 spine segments (violet), not only 2. Which is very good, cause that reduces the risk of serious distortions and some artifacts.
And then, what a surprise! See those little purple octagons along his spine? Yes, you can now animate his back spikes! (Well… 6 of them). And I have to stop here for a moment. If I have to choose what’s my favorite body feature on a dragon/reptile, I would definitely choose back spines! I always were a sucker for them and I always will! I maybe sound like a crazy person, but it’s true. Usually when someone creates an evil creature, it often gives spikes on its back to make it more scary, intimidating, fearsome and thus look and appear more repelling. Funny thing is, that it has the opposing effect on me! :D I like those creatures much more, that has this awesome body feature, which doesn’t. Toothless isn’t an exception (fortunately, and I would be VERY disappointed, If he hasn’t had any)! And I gotta say, he has one of the most beautiful and cool looking back spines, I ever saw! Now… I’m not biased because he’s my favorite character, no. Really! I would gladly go into detail explaining why, but that would be several pages long, (no joke!) and I’m afraid, then you will definitely think I’m mental, so I’m gonna draw the line here. :D
So, yeah. This plus feature is absolutely amazing, and I LOVE it! Also can’t wait to animate those beauties! <3
Now, the wings.
Well, this is the trickiest part. I only had experience with the older model and all I can say is, it’s far from good. The control points are KINDA working, but only for posing. (Half of the time went posing the wings alone!) Animating them would be a real nightmare with this skeletal system! If you’d spent a whole day trying to do a single wing flap, it would still look awkward. Yeah, it’s that bad!
The new one looks kinda scary though. It’s too simple. There are only 5 points for the “fingers”, so there are less opportunity for details, but we’ll see.
Toothless has another smaller set of wings, which he uses for gliding and stabilization. They work surprisingly well with the old model! (If only I could say that for the main wings too… sigh). They only have one control point for each wing, but with that, you can open and close them perfectly! Very nice!
For the new rig, the secondary wings has now a point for each “finger”, (per wing) instead of a single main one. Opening and closing them is a bit more difficult, than with the previous rig, but I don't mind if I have to put more effort for their animation. They deserve the extra work, cause I really like them, and I think they look pretty awesome! :)
The tail fins are a bit more complicated. In the older rig, they also had a single one and worked decently well, (Not as perfect, like the secondary though)
but the newer rig has some serious problems. It has 2 points, one at the tail and one on the end of the tail fin, and they are pretty broken. I have to fix it somehow, because in this state, it couldn’t be animated very well.
Next, the tail.
The older model had 5 points, that worked very well. So well, that I can even say it's proper for animation! When you try to rotate one of the points, it not just bends the tail sharply, but it smooths out the part behind it.
That's not the case with the newer rig. This is an example, that sometimes the newer doesn't mean better. We lose a control node at the base of his tail for some reason (where the secondary wings are). Pretty annoying, cause the tail is almost never straight, and with less nodes the bends are more choppier. Also, when you try to rotate the nodes, they bend the tail sharply. There are hardly any smoothing, so that's a bit disappointing.
Last, (but not least) the limbs.
The older model had a fairly decent amount of control points, but except the most important ones (the 3 points that control his limbs) they weren't very useful, and just complicated the rig further. It was also kinda hard to pose his arms and legs with the control points, cause they weren't perfectly aligned to the model. Distortions are also occurred pretty often because of that.
That's all!
Well, these are the major changes, but because I haven’t tested the new rig really much, that’s all I can say about it for now. (There will be surprises later for sure, but I gonna write them in a future post, when I fully studied it).
That’s pretty much wraps up this review. I’m really excited to finally make some decent animations with my all-time favorite character! It won’t be easy, but that won’t scare me. I LOVE Toothless too much, to such a thing would bother me! I’m planning to do some comparison animation videos with the movies, and some other which are based on my ideas. It will be an exciting (and challenging) ride! :D (Prepare Tarch, I'm coming! :))
Wow! I just realized that this post is almost 6 pages long now! (in MS Word) Well, I could write a couple of more, but that would bring up some more topics, that would require more pages to explain, so I have to stop here. Don’t worry, there will be posts about Toothless again soon, but if I keep writing, the post will never end. Argh!
Thanks for reading my analysis! I hope, I didn’t bored you with the technobabble. Sadly, 3D is full with math concepts and scientific terms, so I can’t really avoid it. Ignoring it also would be foolish, cause then I would show you a completely false picture about it, making look like it’s trivial. And if you didn’t do anything like this previously and you got encouraged to try it for yourself, you will struggle when you face the technical stuff. (And you could definitely have the right to blame me, why I didn’t told you, that it’s full with this crap). So that’s the reason why I got into more detail, how these models are working under the hood. Now you got an insight how they are really like.
I’m planning to do another post today, cause the trailer for the 3rd movie is coming today, and I can’t explain to you how excited/scared/shocked I am right now. Definitely making a post about it, and also a review. (This movie will be the end of me! D:)
See you soon! (If I won’t get a heart attack while I watch the trailer!)
No comments:
Post a Comment