rdg Posted August 31, 2007 Share Posted August 31, 2007 http://www.preset.de/2007/0831/hungry_and_supine/ A wanna-be-PBR-movie. -- --- I encountered a lot of new questions during this take. Georg Quote Link to comment Share on other sites More sharing options...
Jason Posted August 31, 2007 Share Posted August 31, 2007 Well done there! This has an interesting look.. tell us more about it! http://www.preset.de/2007/0831/hungry_and_supine/ A wanna-be-PBR-movie. -- --- I encountered a lot of new questions during this take. Georg Quote Link to comment Share on other sites More sharing options...
edward Posted August 31, 2007 Share Posted August 31, 2007 This couldn't be composited with COPs? Quote Link to comment Share on other sites More sharing options...
3__ Posted August 31, 2007 Share Posted August 31, 2007 can never view your work/reel; unknown format. -cpb Quote Link to comment Share on other sites More sharing options...
Jason Posted September 1, 2007 Share Posted September 1, 2007 can never view your work/reel; unknown format.-cpb "xine" on linux works for me. gmplayer and totem doesn't work. I'll bet vlc works too. Quote Link to comment Share on other sites More sharing options...
rdg Posted September 1, 2007 Author Share Posted September 1, 2007 (edited) This couldn't be composited with COPs? What are you referring to as this is composited in COPS ... a: the pink Houdini - Logo? This is since H9 - I guess some quantization settings. b: the fact that this could be easily done with a diffuse and a unnecessary AO-pass, followed by a heavy blur? Why not - but this asumes you know the look before you start. c: the fact that cops crashed every 15 frames and it took me almost an hour to restart houdini and restart the cop process? And all because I didn't want to color-correct the pink Houdini Logo. This has an interesting look.. tell us more about it! I started with the idea to have something swiggelish. I build a setup to create n-legged-sea-star-like things. H9 feature to have spare channels everywhere was extremely handy for this process: Eg: I added a channels to a pointSOP that use $PT and $NPT and a lookupCurve to shape the scaling and weight of the copied metaballs. Also the movement (noise) is controlled by such spares. I first tried to procedural modell this thing with poly - but switched to metaballs for faster results: a 'failure' as metaballs have a nice look but bad habits. Not that I didn't know about them: Sometimes it'S good to experience them n live action. I added a skin material to my sea star. and while experimenting with this I added an environment light and started the PBR thing. The floor is just a grid with some VEX mountanins and a Rope material. I had problems to vary the rope color by an attribute - shadeops yesterday gave me the hint that its just basecolor not ropeColor_basecolor ... This and the SSS and the PBR indtroduced so many unknown variables to the set I decided to blur it away with an extreme F-Stop setting. The alternative would have been spending more time ... which was against my: "do this in a few hours" premise. I realized that "thing" would very likely penetrate the floor ... I had a few ideas how to prevent this, but decided to test them in a separate setup later (how to bones react to copy/stamps and raysops, howto proceduarly generate and animate bones, are there other ways that do not use bones?) and just flip the thing over Particles where missing: I added a few, realized that my quick'n'drity node cluttering was already production unfriendly and decided against thing-particle-collisions. I spend another hour trying to figure out how to implement sprite-displacement-bounds-rendering with h9. I cranked the PBR samples up and what looked like a cooled displaced skin was suddenly a boring dull surface Skins displacement bounds, materials and me ... the documentation told me that such parameters would be better in materials on one page ... another said the the object is better place ... I lowered the samples again. My hope was that the flickering contributed to the underwater - elecontron - microscope look. First test sequences told me that metaballs where not the way to go. A test with metaball-procedurals (pointer by ihab) changed the look in a not acceptable way. A test with subdivision surfaces, too. So I sticked with my metaballs, accepted that rest-positions and UVs are just a waste of time and promised to try something different the next time. The rendering was extremely uneventfully: one machine crashed every 20 frames, the other rendered ~ 90 frames in 8 hours. I did some colorcorrection tests in photoshop and turned the dull brownish image into that blue look ... well that was auto-levels not me. I converted the 0..255 levels setting to Houdinis 0..1 levels added some addition blur, a function (1/asin Edited September 1, 2007 by rdg Quote Link to comment Share on other sites More sharing options...
Andz Posted September 1, 2007 Share Posted September 1, 2007 Thanks for the info on it Georg. I think it looks really nice, only the "blue birds" distract a little Quote Link to comment Share on other sites More sharing options...
rdg Posted September 4, 2007 Author Share Posted September 4, 2007 (edited) I think it looks really nice, only the "blue birds" distract a little well, this are no birds but protozoa (see nutrition facts for details and differences to birds) I started a new version of the creature - without metaballs: It offers two shape-curves for the overall arm shape and adds little suckers (controlled by size of face and normal and random). The arm length is randomized (not that visible in this image) I think this will be more usable. Georg Edited September 4, 2007 by rdg Quote Link to comment Share on other sites More sharing options...
Andz Posted September 4, 2007 Share Posted September 4, 2007 well, this are no birds but protozoa (see nutrition facts for details and differences to birds) sorry it was a joke :bag: The shapes on the new version looks more interesting. Quote Link to comment Share on other sites More sharing options...
rdg Posted September 28, 2007 Author Share Posted September 28, 2007 The thing somehow managed to get back up. I tried to control the IK goals with a raysop. I guess wrong displacement bounds are causing the black spots around the suckers. Georg Quote Link to comment Share on other sites More sharing options...
cellchuk Posted October 19, 2007 Share Posted October 19, 2007 http://www.preset.de/2007/0831/hungry_and_supine/ A wanna-be-PBR-movie. -- --- I encountered a lot of new questions during this take. Georg that looks pretty cool. good job :thumbsup: Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.