I'm trying to get some of the montage done to close the scene. I don't think it will be done and polished by the deadline, but I definitely have enough done already to fulfill the assignment requirement, so worst case scenario I get the montage roughly done for the deadline, then I return to it later on my own time and refine everything for my animation reel.
Disappointed day today. After doing A LOT of rendering and then editing of image sequences I discovered -the hard way- that setting the raytrace shadow rays to 5 and ray depth limit to 2, is just not high enough and I've ended up with noise in the shadow that doesn't look bad in a single frame but is pretty distracting on the video playback.
Ugh I spent so much time already testing out these lights in Mental Ray still to come up short in the quality results :( Doing more tests today to see what I can do to improve. Unfortunately I doubt I will be able to re-render all the shots I've done for the MPS deadline, especially since that renders times would be doubled with higher quality settings and the grain will be certainly reduced, but may still be present in the playback even with ray trace setting 4x higher. I have more animating and more rendering to do yet, so I must forge ahead! I'm finding something of a compromise by only increasing the levels of the practical light in the living room, which produces the most noticeable shadow of Malcolm on the couch.
(The first image is using 20 shadow rays and a depth limit of 5 with a render time or 4mins 11sec, while the second uses the settings described in the paragraph above with a render time of 1m 58sec. In the last image only raytrace/depth limit is increased (20/5) on the practical living room ceiling light with a render time of 2m 26sec.)
The set ended up being much more time-costly than I would have liked. It takes a lot to make a set like an apartment appear lived-in and so more and more effort went in to providing those details in order to sell it. Probably my biggest blunder in the production of this MSP assignment was splitting the animation up between two scenes. Because of this I was forced to move back and forth between the two scenes every time I made a change to the set in one or the other. Perhaps it's possible to have instead exported the Malcolm character animation into an updated scene at the very end, but I've had trouble with the export feature in the past in regards to preserving constraints, etc. -something I plan on figuring out on my own time for sure, in the meantime, lesson learned.
Another major time-consumer was lighting. I really wanted to look at creating a lighting scheme that gave the scene a more natural, cinematic feel than what I've done in the past. Not only did this mean researching what lights and corresponding settings to employ to achieve this look, but also which to use in order to cut down on the rendering time, as that factor is quickly affecting by how complicated -and consequently realistic- your lighting setup is. This was probably one of the hardest things to weigh out: render time vs. quality, especially because of the length of the project and my further ambitions for it. Consequently, I feel both the frame size (HD720) and lighting/shadow attributes will have to be reconsidered for my final showreel, but for this project, I think the focus is really on how much I do with the character animation, so I'm trying to produce as much of my concept as possible before deadline :/
Ultimately both of these consideration was worthwhile, I think, in terms of advancing in Maya. The set and the props demonstrate how such choices can contribute to characterization, and in the world of cinema these qualities are often better shown than told. Not only that but it has been good practice in materials and textures. The lighting I see as the first real stepping stone to a greater understanding of how to achieve a more polished, naturalistic and ultimately cinematic look in my animation production.
The Picker Tool
One of my key goals in working with the Malcolm rig was to introduce myself to the picker tool, a visual, comprehensive interface for controlling a character rig quickly and easily. Malcolm's picker options are quite detailed, which gives the animator a lot more freedom of expression when posing the character. This picker tool is of course provided by the fine people at www.Animschool.com
The Tween Machine
When I was at VFX festival a few months ago, I went a presentation on character animation by an animation director, who shared with us a couple of ways of refining character animation without ever using the graph editor. One method was simply middle clicking on your key in the timeline and moving to a new position on said timeline in order to hold a key. The other means was by employing a plugin called the tween machine, created by Justin Barrett and free at: http://justinsbarrett.com/tweenmachine/, which essentially creates breakdown poses using a percentage of influence from the key before or after the key you wish to create. Saves a lot time when you trying to apply the principles of animation to your character!
Create Cluster from Soft Selection
Another goal I had for this particular project was to create/animate impressions in the sofa as Malcolm distributed his weight on the cushions. In Maya you can create and animate a cluster of vertices with the deformer option but there's no way to make a soft selection with the feature. Alternatively I could have create a blend shade to animate the indents, but it just didn't seem like the most intuitive way to go about it. This Python plugin, written by Brian Escribano and available free at:
http://www.mediafire.com/file/545fkr217ag549t/ld_createSoftCluster.zip, solved that problem beautifully.
I decided on changing Malcolm's appearance, for a couple of basic reasons: I wanted his outfit to reflect his environment better, i.e. lounging about his apartment. Originally I had planned for Malcolm to be outside, hence the warmer clothes. The other main different between option 1 and 2 is the hair. The longer hair required a little more animation work and was intersecting with the couch mesh a lot and I thought with this relatively short time budget, shorter hair may be more realistic to work with.
Richard Cunningham - IHM Founder