In any case, this was done in Maya. I used a basic particle system for the blood, with a rotating erythrocyte instanced onto the particles - with a uniform field, a turbulence field, and a radial field also affecting the particles. They still have some self collision issues that need to be worked out, obviously. I attempted to use the work around of having each particle generate a radial field, but while that partially works... it also causes the blood to be very jumpy in parts.
I also feel that the turbulence field may be turned up too high, and the rotation of each individual erythrocyte needs to be slowed down a bit. I had to use a trap function of the turbulence to stop the particles from intersecting with the blood vessel wall, despite the "make collide" feature being activated. The GeoConnectors may have some issues associated with them right now that require further research.
The vessel distorts with the use of a blend shape. Obviously, this is not the final texture. However, it did show me that the texture may stretch when I apply it to the vessel wall. I'm wondering if a texture morph may work better, if Maya even does that. I know 3DsMax does, but I haven't tried it in Maya yet.
The background was just a quick fill so it wasn't just floating in space. It was added in AfterEffects without a lot of tweaking, so it doesn't quite match the camera. When the final animation is finished, the actual heart will be composited into the background with the correct camera moves. White blood cells and platelets will also be added, hopefully with a 'liquidy' effect as well. I plan on doing some research to ensure the scale is correct between the vessel wall and the blood elements, as it is a coronary vessel and a bit smaller than say.. the aorta.
I would appreciate any thoughts or comments as I work this video out.
2 comments:
Sarah, I like the ebb and flow of your RBCs. I'm just wondering if the RBCs would be colliding with the vessel wall as well. Lastly, I really like your camera work, the only thing that I might tweak is your camera angle when you start the vessel wall starting to pinch in. I almost would like to see the more aerial view that you have at the end when the vessel starts to pinch in - but this is definitely a personal preference. Take what you will and run with it - cool animation can't wait to see after tweaked the things you want to. - Jess
Thanks for the input, Jess! Technically, yes the RBCs should be colliding with the vessel wall. Unfortunately, they didn't seem to want to adhere to that after I put the instancer on, so they went through the vessel wall. So, no collisions looked much nicer than having them penetrate the wall. Unless I stick some cholesterol in there, then penetration into the lumen is a must. I'm going to ask my Maya professor about that on Tuesday.
Thanks for letting me know your thoughts!
Post a Comment