Development > Artwork
Human Head Model Proves A Headache...
Destructavator:
Well, I managed to get the eyebrows and lashes to look less fake, different technique, but I'm not too confident yet in the way it looks on the low-poly version - I don't think the low poly copy (second screenshot) came out as good as it could have.
I did some research on different ways of mapping high-detail textures onto low-poly objects, I think I might try a different method of doing it later, but for now I need a break!
Destructavator:
Figured it out - at least partially - I accidentally left out some lighting that resulted in un-even lighting in different areas, which made the low-poly copy's texture look so awful.
I fixed that, figured out some more hair styles, made some other fixes too - Yes, I know on the side views the eyelashes and eyebrows look like they stick out a bit, but the final product won't have that.
PhilRoi:
lol, the low poly looks good actually! I like it better then the high poly version.
the problem is your running into what psychologists and artists call the "the uncanny valley". The human brain has developed very sensitive facial feature recognition centers. You hit a point where as things get more detailed your brain stops noticing what is right about the face and starts focusing more on what is wrong... making it more detailed can make it worse, sometimes simpler is better! A simple face our brains fill in the gaps and fools itself into thinking its a a friendly/good face. the more realistic it gets the less room for your brain to fill in the blanks and it ends up tripping over the wrongness more.
sometimes suggestive is better then representational! The smaller faces for the low poly version lose detail and move out of "the uncanny valley".
cheers!
Destructavator:
--- Quote from: PhilRoi on July 19, 2009, 12:08:36 pm ---lol, the low poly looks good actually! I like it better then the high poly version.
the problem is your running into what psychologists and artists call the "the uncanny valley". The human brain has developed very sensitive facial feature recognition centers. You hit a point where as things get more detailed your brain stops noticing what is right about the face and starts focusing more on what is wrong... making it more detailed can make it worse, sometimes simpler is better! A simple face our brains fill in the gaps and fools itself into thinking its a a friendly/good face. the more realistic it gets the less room for your brain to fill in the blanks and it ends up tripping over the wrongness more.
sometimes suggestive is better then representational! The smaller faces for the low poly version lose detail and move out of "the uncanny valley".
cheers!
--- End quote ---
Thanks! I'm glad you told me about that - it would sure explain some things I've noticed when looking at the model in Blender from various angles and distances. I'll keep this in mind as I work on the soldier heads.
Destructavator:
Did It!
- Finally got the texture put on a working low-poly copy without any serious screw-ups!
In the screenshots the low-poly version is on the left, the high-poly (which can be used for cinematics and such) is on the right.
I forget how many polys are in the lower version, I think its around ~300 or so, the high-detail one has upwards of ~1600 or something.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version