kuroyume0161 opened this issue on Mar 15, 2005 ยท 30 posts
kuroyume0161 posted Tue, 15 March 2005 at 1:47 PM
Pop on over to CGTalk at this link: The Final Battle This guy isn't rigged yet, but that discussion about Aragorn's facial stubble and "not quite there yet". I think that we've arrived. :)
C makes it easy to shoot yourself in the
foot. C++ makes it harder, but when you do, you blow your whole leg
off.
-- Bjarne
Stroustrup
Contact Me | Kuroyume's DevelopmentZone
Pollee posted Tue, 15 March 2005 at 1:58 PM
Oh my, we've arrived. Thanks for the update. Now I really have something to compare my renders to. He's so real looking.
operaguy posted Tue, 15 March 2005 at 3:22 PM
As an animator, I immediately want to make him move. Even if you could achieve realism of this nature in ANY affordable rendertime budget, which I doubt, now you have a new problem: The eye would demand realism in motions on the same level. But the time you are done.........just hire live actors and film the damn thing! ::::: Opera :::::
face_off posted Tue, 15 March 2005 at 3:26 PM
Wow - awesome image.
Creator of PoserPhysics
Creator
of OctaneRender
for Poser
Blog
Facebook
dlfurman posted Tue, 15 March 2005 at 3:28 PM
Hey, a Poser 6 challenge. Lets try and recreate this image! That is so WOW! It is going in my NOT A PHOTO folder!
"Few are agreeable in conversation, because each thinks more of what he intends to say than that of what others are saying, and listens no more when he himself has a chance to speak." - Francois de la Rochefoucauld
Intel Core i7 920, 24GB RAM, GeForce GTX 1050 4GB video, 6TB HDD
space
Poser 12: Inches (Poser(PC) user since 1 and the floppies/manual to prove it!)
kuroyume0161 posted Tue, 15 March 2005 at 3:33 PM
Yeah, but I don't think any 'live' actors would appreciate being sliced in half with a broad sword for any amount of money! ;) This amount of realism means that cut scenes between the live actor about to be halved and the CG stunt double being halved would be almost imperceivable, even full screen.
C makes it easy to shoot yourself in the
foot. C++ makes it harder, but when you do, you blow your whole leg
off.
-- Bjarne
Stroustrup
Contact Me | Kuroyume's DevelopmentZone
operaguy posted Tue, 15 March 2005 at 3:57 PM
can you imagine the render time? And now much postwork down to the pixel level did the person do? ::::: Opera :::::
bobcat574 posted Tue, 15 March 2005 at 4:50 PM
I just can't believe I smacked my face into my monitor screen trying to get a closer look! Amazing.
kuroyume0161 posted Tue, 15 March 2005 at 5:26 PM
As far as I've read, no postwork at all. All Maya and Mental Ray. The render time was not mentioned, though many have asked. It is certain that this is not a two minute or two hour render. :) Of course, for an animated figure in a movie, the big studios have render farms that make my (real) farm look like a cardboard box. Did you see the one used for LOTR!? 3200 processor farm array (Weta Digital).
C makes it easy to shoot yourself in the
foot. C++ makes it harder, but when you do, you blow your whole leg
off.
-- Bjarne
Stroustrup
Contact Me | Kuroyume's DevelopmentZone
operaguy posted Tue, 15 March 2005 at 5:59 PM
Final Fantasy the Spirit Within = 1000 CPUs running version of RMan, average final render budget = 90 minutes per frame, per an interview with the director. I went to see "Robots" yesterday and noted in the credits that they must have used an HP farm. The render software is their proprietory CGIStudios. ::::: Opera :::::
jeffg3 posted Tue, 15 March 2005 at 6:27 PM
kuroyume0161: "It is certain that this is not a two minute or two hour render." Long time Maya user here. You might be suprised. There's nothing in the image that equals long render times. He said he's using the sss fast skin shader. I'd say (depending on the machine) 10 to 15 minutes. Maybe 20. Though I could be wrong.
Becco_UK posted Tue, 15 March 2005 at 6:34 PM
A lovely, detailed, modelled character but the render doesn't quite look right to me. The skin seems a little strange.
operaguy posted Tue, 15 March 2005 at 7:02 PM
Attached Link: http://market.renderosity.com/softgood.ez?ViewSoftgood=15063
There's another question I have...and to the general audience.....and this is not to take away from the fantastic render we are all looking at.....Isn't it PhotoRealisticTexture IN, HyperRealisticRender OUT?
I purchased a texture here at Rendo called Golden Girls (link above). This texture is SO realistic, all you have to do is pop it on V2 or EJ and you have grandma, every wrinkle and lump and use-your-imagination.
My first render was so real it scared me. I did no modeling, postwork, morphing or anything. I just hit the make art button.
::::: Opera :::::
Message edited on: 03/15/2005 19:11
jeffg3 posted Tue, 15 March 2005 at 7:12 PM
operaguy: "Isn't it PhotoRealisticTexture IN, HyperRealisticRender OUT?"
Not exactly - but it can get you a long way there.
A figure needs to look good from different angle and under a variety of different lighting conditions.
A lot of what give a character "reality" is the way the specular highlights are handled (and modified by bump / displacement maps) and the subsurface scattering effect of the shade (Poser 5 doesn't really have this out of the box).
There's a lot of "real" looking Poser work out there. But I've never seen Poser 5 produce a full color, well lit image that I had trouble distinguishing from photograph.
Message edited on: 03/15/2005 19:14
Becco_UK posted Tue, 15 March 2005 at 7:45 PM
jeffg3: Perhaps the forthcoming Poser 6 may produce more realistic renders. I've never seen any close up render, from any software, of a person that is 'photographic' quality.
Keeping it in context, Poser 5 is a fraction of the price of Maya Unlimited, so the natural expectation would be for the 'higher' end software to peform better.
What's good for all the 3D community is how an increasing number of 3D software companies are improving their products at a relatively lower price compared to a few years ago.
Message edited on: 03/15/2005 19:48
dlfurman posted Tue, 15 March 2005 at 7:57 PM
Maybe with Poser 6??? :)
"Few are agreeable in conversation, because each thinks more of what he intends to say than that of what others are saying, and listens no more when he himself has a chance to speak." - Francois de la Rochefoucauld
Intel Core i7 920, 24GB RAM, GeForce GTX 1050 4GB video, 6TB HDD
space
Poser 12: Inches (Poser(PC) user since 1 and the floppies/manual to prove it!)
jeffg3 posted Tue, 15 March 2005 at 8:00 PM
Becco_UK: " I've never seen ANY close up render of a person that is 'photographic' quality."
Check this out. It's pretty close
http://www.xmg.ro/mihai/pages/img_pages/box6.html
or this
http://www.raph.com/3dartists/artgallery/6327.jpg
these next two are also pretty good, but I believe they are closer to simple projections than true, light-interactive shaders
http://www.highend3d.com/artists/artist.3d?au=3dcharacters&iid=133
http://www.highend3d.com/artists/artist.3d?au=3dcharacters&iid=151
jeffg3 posted Tue, 15 March 2005 at 8:06 PM
dlfurman: " Maybe with Poser 6??? :)" Well, P6 has SSS and image based lighting. That coupled with a modified face_off shader network should produce some stunning renders. Too bad CL doen't allow 3rd part hair plugins. And the new ambient occlusion will also help a lot!
Dave-So posted Tue, 15 March 2005 at 8:53 PM
I'm really looking forward to P6... My goal, although not too close yet, is to get to that near photorealistic render. Some of the textures from Stephania Zambini, Morris, and Syyd at RDNA are real close.....but the lighting can kill them in a heartbeat. Camera angles and so forth throw them out really fast, so the lights have to flow with the movement of the figure, but then you lose reality of the scene, as the sun or even a room light does not move in real life with the camera movement, but the shadows and light target move.
Humankind has not
woven the web of life. We are but one thread within it.
Whatever we do to the web, we do to ourselves. All things are bound
together.
All things connect......Chief Seattle,
1854
PapaBlueMarlin posted Tue, 15 March 2005 at 8:57 PM
I suspect that realskin and hyperreal will still work in P6 so combined with ambient occlusion, you'll get a great effect. I don't know what additional nodes added to the materials room or not so how exactly SSS is done in P6 hasn't been revealed yet. I'm getting anxious just thinking about it :)
tlaloc321 posted Tue, 15 March 2005 at 9:14 PM
I am perplexed by this term Photorealistic. I know what they mean but I never considered Photographs Realistic. The eye handles everything differently than a camera, we just take photographs for granted. I think if you showed a photograph to a person who for some reason had never seen one, they would wonder about his strange 2 dimentional approximation of what their eye sees. Just some thoughts.
kuroyume0161 posted Tue, 15 March 2005 at 10:11 PM
jeffq3: I don't use Maya so it's all guesswork. :) Cinema4D could do a render like this is less than an hour - minus the hair. ;) Hair is almost always (unless it's a really good transmapped model) render intensive. But again, he said he used Maya Fur with Mental Ray which seems to indicate optimization between them. Anyway, this takes my mind off of the P6 order problem and helps me resist purchasing MilDragon 2.0. ;0)
C makes it easy to shoot yourself in the
foot. C++ makes it harder, but when you do, you blow your whole leg
off.
-- Bjarne
Stroustrup
Contact Me | Kuroyume's DevelopmentZone
linkdink posted Wed, 16 March 2005 at 12:23 AM
Here's the closest I've come to "real" using Poser 4 and some basic postwork. It's nothing to compared to those Maya images, but 6 months ago when I started playing with Poser, I wouldn't have thought this was possible for me: http://www.renderosity.com/viewed.ez?galleryid=905537&Start=1&Artist=linkdink&ByArtist=Yes (sorry for the shameless plug, but I'm kinda proud of this, and it's only got 130 hits.... )
Dave-So posted Wed, 16 March 2005 at 6:11 AM
looks good... 130 hits...no breasts, no hits
Humankind has not
woven the web of life. We are but one thread within it.
Whatever we do to the web, we do to ourselves. All things are bound
together.
All things connect......Chief Seattle,
1854
Hawke posted Wed, 16 March 2005 at 6:48 AM
Good god - thats amazing, thanks for the link.
Barryw posted Wed, 16 March 2005 at 1:33 PM
I wouldn't even want to try to model that chain mail.
kuroyume0161 posted Wed, 16 March 2005 at 2:55 PM
If you were able to find and read the creator's responses (which I'll admit is not easy with over 16 pages), he did not model the chain mail. I think he modeled one link then used a fractal instancer over another object to simulate the chain mail.
C makes it easy to shoot yourself in the
foot. C++ makes it harder, but when you do, you blow your whole leg
off.
-- Bjarne
Stroustrup
Contact Me | Kuroyume's DevelopmentZone
wolf359 posted Wed, 16 March 2005 at 3:19 PM
you could get close to that chainmail with displacement mapping. something poser has had for years now. too bad hardly anyone bothers to use it for detail work.
Becco_UK posted Wed, 16 March 2005 at 3:45 PM
I bought Poser 5 in 2003 and it had displacement mapping that rendered straight out the box. Throw in a material room that can match many 'high' end programs, plus a cloth/hair room and you have one of the most underated programs made.
The picture is a rough example of Poser 5's displacement, using one of Syyd's Industrial Woman textures.
Message edited on: 03/16/2005 15:47
smantha posted Fri, 15 April 2005 at 1:57 AM
wow, I've been checking out some of the work posted on some of those other links on this thread, and the work is amazing! However the people critizing the work are frigging brutal! At least I know here if I post something that's not very good people will be nice to me and offer constructive criticism. lol