PapaBlueMarlin opened this issue on Aug 18, 2006 · 87 posts
bagginsbill posted Sat, 19 August 2006 at 4:42 PM
Define "easier". I did the whole 6 step tutorial, including writing all the comments, saving multiple versions of the script, and generating and saving a bunch of test renders, in exactly 30 minutes.
If I wasn't making a step-by-step tutorial, it would have taken no more than 5 minutes.
Now how quick can you paint a bruise then test it in a render? Then change your mind about the colors, size, position, several times. I can't do it in 5 minutes.
Another thing is, if you want a bruise on one of the seams, like the back, you're in for a treat trying to paint that. You actually have to paint two separate bruises and get them to line up.
Painting a bruise separately, then blending with nodes, that seems more reasonable. But I still couldn't do it faster than I did.
The thing you have to remember, is that the algorithms are reusable. You really can't judge the effort of creating the algorithm versus painting one bruise. Instead, imagine you are doing a movie. The director tells you to produce 30 different bruises on 5 characters, The story spans enough time that the bruises have to evolve, changing colors and fading as five movie "days" go by. That's 150 different bruise images. But if you have an algorithmic way to make them, then you just call it with new parameters.
I agree that I didn't make it look realistic, but I was only trying to demonstrate the possibilities, as requested. Given enough time, a very realistic reusable bruise generator is totally possible.
Finding the algorithms can be tricky, but once they are defined, using them is really easier than painting ANYTHING. For example, I recently was asked how to make reptile and goldfish scales. I spent two hours on the algorithm. Now people can have scales of any size, any color, any shape, any orientation, in seconds.
Check out these images by me and others using the scales. With a painted image they'd all be the same. With the procedural approach, no two are alike.
http://www.renderosity.com/mod/gallery/index.php?image_id=1274107
http://www.renderosity.com/mod/gallery/index.php?image_id=1266453
http://www.renderosity.com/mod/gallery/index.php?image_id=1254884
http://excalibur.renderosity.com/mod/gallery/index.php?image_id=1264358
http://www.poserclub.jp/ph_view2.htm?ph_no=2010&phid=1154574595
http://www.runtimedna.com/mod/forum/messages.php?forum_id=43&ShowMessage=233907There's lots more, all using the same one-page script, but you get the point.
My point is, I think algorithmic shaders give so much more variety and are much more useful than image based shaders. Sure, there are times when the image based shader is the only way to go, but not always. So I always encourage their use.
I think maybe the "hobbyist" market, which seems to be mostly what Poser caters to, doesn't care that much about saving a lot of time. Maybe I should be peddling my "medicine" to the pros? :) LOL
Renderosity forum reply notifications are wonky. If I read a follow-up in a thread, but I don't myself reply, then notifications no longer happen AT ALL on that thread. So if I seem to be ignoring a question, that's why. (Updated September 23, 2019)