Friday 27 February 2009

I thought It'd be wise to go through the process of what I've done so far, seeing as I'm halfway through a night shift with little to do. Also, I'm pretty sure things are going to get busier, and I'll probably won't have time to do reflection later on.
I'm not too crazy about the blog system, I guess I just don't like blogging. I kind of paid for it last project - I've done a crazy amount of work, but didn't get graded for it because I didn't put much effort into documenting it, and much of it didn't show up in the film. 

Should be noted that spelling and grammer are out of the window as it's 3:50 and I'm not that awake. Should also be noted that I'm aware of how boring this section is going to be. In fact I'm aware of how boring the whole blog is. I'm doing this as ass-coverage rather than out of a genuine will to blog or to record my process. So if nothing else at least I have some self-awareness as to the level of excitement going on here.

Modeling: I found modeling environments and buildings quite easy. You just modeled something to look like you want it to look. They weren't alive and they weren't all that appealing. In fact, because we did post apocalyptic, the messier the better. Modeling for the sake of animating is a whole different animal. Modeling people is also really different than modeling objects. A lot more time consuming, a lot more to keep track of, but also a lot more rewarding to see your character coming to life slowly. The final model is about the 3rd attempt or so. The other two attempts were done in a day, so I didn't lose too much time on them. The first one had to be binned because there were too many vertexes too early. I cluttered the geometry because I thought It'd make it easier to model. So as I was building the torso, for example, I'd add edges wherever I thought my character might have to move or twist eventually. After a bit of reflection I realized It's probably better to take the modeling one step at a time - first blocking out the character very roughly, and not worry about the amount of edges and vertexes. Keep it to a minimum, in fact. And so that's what I did on the second attempt. Only then I started pulling the vertexes to make it look smoother. And it did look smoother and nicer. Problem was, I didn't block out the fingers, and adding fingers to a smoothed hand proved to be a pain. The third and final attempt at modeling the body was just really blocking everything out as simply as possibly, and only when everything's blocked out start modeling. 

The head was the biggest challenge. I've followed the book 'Stop Staring' In my free time on holidays, but have only gotten up to the nose bit. So I could do the lower face quite well, up until the nose, and indeed that part I've done quite fast on my model. It was after that that I got stuck. I tried continuing the face with pollys by extruding them, but after about 4 or 5 failed attempts t just didn't feel like I was going anywhere. This is where Jamie came to the rescue. With his help I've redone the upper face, this time with curves again, until finally I got the whole shape. Then with a lattice deformer I've changed my face to fit the original drawing exactly (my model has shrunk a little, so the lattice helped me scale it up in areas that were scrawnier than they should be). One more helpful tool I used was the average vertexes tool. It does something not unlike smoothing, only without adding any extra vertexes. 

Throughout the modeling stage I kept looking at other characters and other models to see how they've coped with not having triangles in areas where they feel needed

One thing I think my model is lacking is more definition on the clothes. I was waiting for a reply from Alex regarding what's the best way to approach clothing and a dress - do it seperately, or model it with the mesh? By the time I've gotten my answer, I've already spent too much time on my model. The clothing ended up not looking very convincing at all. Oh well..

Texturing: Most of what I've done last project was texturing and UV mapping. So in this project texturing has taken me about 2 days of not too hard work. I think I can safely say that I feel very confident about working with UV maps, and enjoy the combination of photoshop. I've introduced to a lot of people on the course the PSD Network option, which is a really good way of doing complex layer-based textures in photoshop and updating it immediately in Maya. It's surprising how really helpful you can discover almost by accident. 
I think last project was a bit difficult for me in terms of artistic fulfillment. I was mainly applying someone else's textures to someone else's models, 90% of which were never even seen anyway. I felt like the uncredited behind the scene guy, but I've also made a conscious decision halfway through to project to just get on with it and be a team member rather than thinking about this project in terms of personal goals. I think maybe this led me to really go to town a little with texturing in this project. I was texturing my own character for my own story, etc, and had no one else to consider. I stuck to the claymation style the brief has set, but only to some extent. I felt that pure claymation look actually compromises the character a bit, and decided to add a bit more colours and shades, and generally more of a complex texture. I do think the result is quite lovely, though not entirely plasticine-like. 

To be continued another time, I'm boring myself here..

Thursday 26 February 2009

Rigging finished

Rigging is finished, which is good. My rig is really basic, but I think it'd do. After I finished rigging I found a better tutorial about it, but it was more advanced. I think I'm happy that I've started wit the basic one, and with the next character I'll try the advanced one. With the old lady at least I don't need too much flexibility. Plus, time limit. Now for the facial rigging, which I think is going to take longer. After that, skinning, painting weights and all that jazz, and if time allows it, animation at the end.

I was about to post another video before I realized I have a problem with the rig. back to drawing table - thank god for saving in multiple files

Rigging the other hand

The other hand has a few different features in it. In one hand, she'll be able to hold something with a handle (a cup for the story) with the help of a set driven key, and she'll be able to pick something up (near the end of the story), whereas I made a holding driven key in other hand for her to hold her walking stick.

In terms of the fingers, I've decided a more efficient way would be to attach the actions (curl, lef-right and up-down movement) into the already existing translation values. I've limited their movement, and also locked any option to rotate or scale. The result is the sliding controller, similar to Moon's facial control panel. It's more convinient and intuitive, I find, than the first hand. For the sake of consistency I should change the other hand, but I don't think it's worth the time. Next rig I do, I'll know better

More


Ok, so now I've done whatever it's called to each finger seperately. after a few fingers I've also realized what wasn't really needed, and where I set the values too high or too low. I've done curl and stretch (though decided to drop the stretch halfway through as it wasn't looking good, and not very important anyway), left and right, and up and down. It was important for me to have a difference between the down and the curl, as you can lower your finger, but it won't be quite the same as curling it. I remember Moon had all these options, but I'm not sure if it's standard to have all of these as riggable options, or whether it's something the animator will teak himself. It's actually probably easier to tweat it yourself rather than setting the values. I didn't use most of them when animating moon last year, but then again I wasn't that familiar with the rig, so maybe this time it'll be different. further evaluation will hopefully come once i'm in the animation stage.

setting driven keys

This is actually quite fun. I feel like i'm in a scene from the terminator. The second one of course, I don't think I'd want anything to do with the new one if Christian Bale is going to be so rude to the staff and everything.

I'm looking forward to see how terribly bad this actually goes once I paint the weights and everything, but for now- fun.
One I started with the rig I've noticed that my thumb is too short for this sort of rig. I think had I done a more complex hand rig, I could have gotten away with it. It's just that I've made the thumb directly connected to all the other fingers, whereas in reality the thumb is connected more to the wrist area, and therefore has more manoeuvrability and independence (i.e. opposable thumbs). It's really interesting though, it's kind of a practical lesson on anatomy. Thinking how your bones move, what you can do with them, what's the best structure, etc. The less interesting part is just mindlessly repeating the same process with the other hand, but it's gotta be done.

I've had a few issues with the pole vector constraints - Whenever I attached the hand's IK handle to an object (I use the object as a control to rotate the knee/ankle), the rotation of the bones seem to have flipped around. It's kind of hard for me to explain as I don't really know the technical terms for all this stuff, but basically the hand just flipped over. After a few attempts, I realize that it's actually quite important where you're placing the item that you're going to use as controls - it has to be in a certain angle to the ik handle, otherwise it flips over.
This is dull and boring and not helpful to anymore, but it's for asssesment purposes.

Anyway, these are some hand poses that seem quite important to have. I'm gonna add seperate finger control in a minute too.

Wednesday 25 February 2009

This looks really promising, and it's claymation, too



Meanwhile, texturing and everything is done, so I'm working on the rig right now, we'll see how it goes

Monday 23 February 2009

Thursday 19 February 2009

Finale result

Oh yeah, probably worth putting here what I decided to go with

my bump-map:

colour map:
and test render:

More render tests

The first image is halfway through my attempt of gettingt he right clay-ish texture using aardman picture as a reference

I'm starting to think about compositing a bit, and this is quite a nice example of good compositing to cover for half-assy animation
.

Anyway, some bump-colouring tests, using a psd-network constantly updated (works much better than texturing in layers in maya, imo).
I really enjoy texturing, I'm very comfortable with Photoshop, so it feels like a piece of land in the stormy waters of maya





Friday 13 February 2009





1. Is looking closely at the sort of texturing I more or less aim for. The rest are first attempts at basic texturing. Too light, even though I reduced the light source dramatically, and still generally too smooth. Will keep playing to create desired effect

Wednesday 11 February 2009

Photoshop


I was a little tired of Maya, and wanted to see how the animation option in photoshop work..

Tuesday 10 February 2009

Finished for the day



It's 18:30, I'm the last one in, and I think I'm done for today. Huge progress with the modelling, with the helpful help of helpful Jamie.
I'm pleased.

Another attempt


After quite a few failed attempts at modelling, I think I'm finally on the right track. Mostly so far I've tried extruding the faces to come with the forms I needed. I've done this so that the curves are going to be sequentials. Jamie's help today has been extremely helpful, he helped me come at it from a different angle, and go to curves again. I'm not done yet, but I'm getting there, as far as modelling is concerned.

I've also decided to abandon the 'Stop Staring' book that I've been following so far. It's a great book, and really helpful. But I'm going for modelling and rigging-light, whereas this book goes into great depths. I was trying to skip through most of it and just take the practical stuff out, but they're just too difficult rto seperate. I'm gonna try following 'Digital tutors' videos more.

One more issue I'm flagging up for myself is a scheduel red-flag. A month is left for this project. If I'm lucky, I'm gonna finish modelling ,texturing and rigging within two weeks. Then one week to animate, and then one week for camera-set up, rendering, compositing and burning.
One week to animate. This is extremely frustrating, and is just the thing I feared would happen. I'm looking around me and everyone is more or less in the same boat, so at least it's comforting to know it's nothing wrong with my pace. Still, so far 2 terms without serious character animation. I realize that at my spare time i'm gonna have to do a lot more animation, but it's a real pity that the course doesn't support what I wanna do

Sunday 8 February 2009

Thursday 5 February 2009

More mesh problems.



I've saved the file at college, opened it at home, and again the same problem.
If I'm going to work back and forth, it seems like a problem I'm gonna need to solve. I've joined CGTalk, and reading their faq. I'll try posting there and seeing if anyone can help.

The horror

I was working on my face, which is Nurbs and curves so far, while the body was hidden, presumably safe. When I unhid it, this happeend.



I'm pretty sure my copies at home are fine, but this is one odd problem. Only thing I can think of is version incompatibility.
Meanwhile, lips are almost done except for finale tweaking, and I'm moving on to the nose. I'm thinking and hoping the mouth was the most difficult area, and that everything else will be easier. Still waiting for Alex Hulse to help me with some 3d clothing issues, but he hasn't answered my email.
I'm also thinking i should upload failed attempts of modelling, just to show more of the process.

Tuesday 3 February 2009