[MeanPi2]

Well turns out animation is hard. No wonder the main focus is on animation not models, backgrounds or really lipsynching. This is a whole 68 frames of that monster .Wav. and it's still not done within those frames. This is simply a progress report in reference to the arms and shoulder area. Doesn't look quite right to me but not sure why- suggestions?

This whole 2 second animation is in XViD codec format, but the CC's were changed in VirtualDud so it WILL play with DiVX 5.1.1 codec and should play in the older DiVX 4 and 3 codecs also-not sure about those though.

Edit: Oh yea, the models changed a little tiny bit- she has invisable to camera shadow casters parented to each arm and leg bone so shadows aren't limited to only shadow-maps now. Not sure why it was so bothering in the first place but whatever.

[kumar]

Cool,,, very interesting approach.

[MeanPi2]

Well found out 1 reason why it looked off, didn't like the left shoulder movement during the sigh. Here's an updated video of the simple change. Nothing like burning up bandwith..

[puppet]

cool, you have a nice little style going on there

[MeanPi2]

Thanks Kumar and Puppet. Got rather frustrated with keyframing the bones and their problems, not to mention the morphs, so took the day off to create an intro to the animation (minus the intro to the intro) to learn how to work the graph editor and scene editor in LW8 plus to test some facial morphs and camera movement. Proved just as much a pain in the kidneys. It's 400 frames long (13.3 secs) and rendering, I'll post it when it's done.

Edit: Oh yea, the turd makes a cameo in this one. or

[mtmckinley]

pretty neat. Looking forward to progress.

It'd be really cool if you could make a couple of different head textures and animate between them to convey lip synching.

[MeanPi2]

Actually, I have morphs set up for the mouth to lipsynch and found some visiual phonetic references and an interesting 'essay' on why not to use pure phonetics alone..unfortuantly little of it does any good since I can't grasp the use of 3d phonetics and shapes on an essentially 2d mouth. Found this out trying to form the word 'Well..'. The currently rendering animation does use, for the most part, all the facial morphs constructed so far, including the mouth (but didn't have time for the eyeball movement this time due to self-imposed time limitations- those aren't morphs, individual objects using a null and the 'Follower' plugin which itself is screwy, can't have the eyes 'rotate' to the null by moving the null on the X and Y axis but only if the null roates on H and P and can't link 2 nulls (one per eye) to a master null linked with follower to those nulls to allow indepenent rotation along with symetrical rotation). So basically body movement is compensating for the lack of phonetics ability. Heh.

As for using different head textures I don't understand this- could you elaberate on it?

[colto]

For the eyes, just make a model that is nothing but a single point and save it as eye_target or something to that effect. Then use channel follower to make the eyes follow it. Or u could clone it and have one per eye and then parent them to the null and control it that way. This way the eyes should work properly and u may animate them nicely!

[MeanPi2]

Colto: That's what I'm doing now. I was trying to use a null per single point eye with both nulls 'parented' via Follower to a single master null- wouldn't work course I only spent about 20 minutes on it.

Stopped the animation at 371 frames or so and was ready to send it. But I'm sorry, even with XViD compression at 1 Quality and resizing the frames from 640x480 at 32bits to 100x80 at 16bits in Virtual Dub (without sound) I can't get the video below 1 meg even using max compression in WinRar. In an actual animation the sound file is aprox 351 frames long, with sound being about 70K in a video file. Either I'm doing something wrong in the compression settings or things aren't really adding up for a file size under 1meg being possible.

Anyways, looking into hosting after work tomorrow, the video will be available then- or sometime, at a good resolution.

Edit: File size on disk is 48K, VirtualDub reads it around 70K.

[MeanPi2]

Maybe..

Nope, file size too large I guess. Oh well time for bed.

[MeanPi2]

Ok, couldn't give up dispite having to go to bed (thank God I only drive big trucks all day- no need for sleep. Heh.). Here's a 160x120 video using Lancoz3(whatever) resizing with no shadows or particles or textures. But it's under a meg and rendered real quick. (Unfortunatly the 300x200 was still over 1024bytes..sigh.)

Edit: Screw the quicky video, replaced it with the same resolution but this has textures, shadows (lame lighting) and particles (quickies which kinda suck and are totally wrong).

[MeanPi2]

Well this was done after work today, it's the continuation of the intro leading right up to the speaking parts. Somewhere along the line I lost, in the animation at least, all wall textures this time around so no annoying pelicans or whatever those birds are.. It's also quite short. Critiques as always are welcome.

[mtmckinley]

looks pretty good, although a big stiff. Not a lot of body movement. I like that the face can animate.

[Fallenswordsman]

I like it - despite the simplicity of the character its moving pretty well!

[Zeeman]

nice expressions. Maybe u need to make the arms flail or make the head jerk a little.