IRAY Photorealism?

1181921232468

Comments

  • bluejauntebluejaunte Posts: 1,902

    That is insane. Although at that point I really have to wonder, why go through all the trouble? Just make a photo. I mean they probably did and he then replicated through airbrushing? But why? What's the artistic value beyond just the insane show of skill? I don't see any at least in this case.

    Yeah, I get what you are saying.... why reproduce an airbrushed version of a photo?  I think he has done work from actual live models as well.  My reason for posting it was to illustrate whether you are copying a photo, or airbrushing something from your head to produce something so lifelike is amazing.  I would love to reach that level in CG personally and after seeing some of the work posted I have to admit some have reached that level.  

    Artistic value? I guess it is based upon whatever your opinion of what work has artistic value or what does not.    Some would say that if you buy an off the shelf pre-made CG model, add textures, pose and render....what artistic value is that?  As if the only artistic value resides with someone who creates a realistic figure from the top of the head with no so called cheating by using photos, or references or God forbid textures from a real human being.   Of course this is ridiculous as using human reference is not only needed, but the best way to replicate realism in my opinion.    

    Me personally, I am trying to replicate a realistic photo....so I would pick a human reference and use a CG model with similar skin tone, hair and lighting to come as close to that photo as possible.   

    I think your figures are absolutely top notch and have helped bring DAZ figures to a new level of realism I am looking for.   The rest is up to DAZ and hopefully Gen 9 brings us that much closer.  

    If it looks like photo, then its artistic value will be judged by photographic standards and that one looks utterly trivial. Only the husband or maybe mom and dad would give damn about it. I'm sure there are fans of hyperrealism who would see it differently and would love the fact that you could mistake this painting for a photo, but artistic value comes from getting to see something through the eyes of the artist, some interpretation of reality, but if the painting is just a dull photograph then what's the point? The artist replicated what he saw through the camera, to such perfection that the viewer cannot see a difference. There is no more artistic value in that than there would be in the photo. So if it's not an interesting photo, then what's really the point as an art piece? Just as a show of craftsmanship it is amazing but is that enough? 

    I guess it comes down to your opinon.  I actually see and admire the talant in being able to replicate all the details of a dull photo using an airbrush or whatever medium you choose to use.  I am a fan of hyperrealism in any form.   

    Any crappy "Poser" art has more artistic value than this if you ask me. While the craftsmanship is obviously lacking compared to this technically masterful painting, at least I know that someone has put together a scene creatively and not just replicated a dull photograph.

    I'm reminded of what I was thinking a while back. Once we reach absolute photorealism and our renders are not distinguishable anymore from real photos, what will the point of creating CG renders be then? Are we perhaps striving for the wrong thing? Wouldn't it be better to have some clear indicators that what we're looking at is not real? Sure we could make fantastical scenes that can't exist in real life, but what if you just wanted to make a simple portrait? Someone might think yeah this is a nice photo, but the fact that it took a thousand times longer to create it isn't going to matter one bit. Even people who are into CG renders would eventually lose interest as they couldn't tell any more if something is a render at all. We couldn't say "this looks so realistic" any longer, unless we knew beforehand that something is indeed a render. Wouldn't this change everything for the worse?

    Deep thoughts laugh

    Maybe when we reach being able to technically make something absolutely photo real we can put much more concentration on scene content and story telling.  You can do this now, but backed up with absolute photorealism sure makes it a much sweeter deal to me!   As far as making something look extremely real and using it for nefarious reasons....its scary for sure.  Should we just stop trying to shoot for absolute realism?  As long as people want photo realism and willing to either pay for it in either time (learning skills) or money its going to happen.  As they say "the genie is out of the bottle".  

    Very deep thoughts my friend  cheeky

    I admire the talent very much, it just bums me out that he's using that insane talent to create such an overall dull work. I mean you said it right there, you'd use such skills to focus more on content and story telling, aka doing something interesting with it.

    But hey, maybe that was some type of commission work and he had no say in the photo per se. Or the goal was exactly to invoke this kind of reaction. That would make it art right there. I did have a reaction after all. I wouldn't have if I didn't know this was a painting though, that's kinda my point I guess.

  • Siciliano1969Siciliano1969 Posts: 433

    That is insane. Although at that point I really have to wonder, why go through all the trouble? Just make a photo. I mean they probably did and he then replicated through airbrushing? But why? What's the artistic value beyond just the insane show of skill? I don't see any at least in this case.

    Yeah, I get what you are saying.... why reproduce an airbrushed version of a photo?  I think he has done work from actual live models as well.  My reason for posting it was to illustrate whether you are copying a photo, or airbrushing something from your head to produce something so lifelike is amazing.  I would love to reach that level in CG personally and after seeing some of the work posted I have to admit some have reached that level.  

    Artistic value? I guess it is based upon whatever your opinion of what work has artistic value or what does not.    Some would say that if you buy an off the shelf pre-made CG model, add textures, pose and render....what artistic value is that?  As if the only artistic value resides with someone who creates a realistic figure from the top of the head with no so called cheating by using photos, or references or God forbid textures from a real human being.   Of course this is ridiculous as using human reference is not only needed, but the best way to replicate realism in my opinion.    

    Me personally, I am trying to replicate a realistic photo....so I would pick a human reference and use a CG model with similar skin tone, hair and lighting to come as close to that photo as possible.   

    I think your figures are absolutely top notch and have helped bring DAZ figures to a new level of realism I am looking for.   The rest is up to DAZ and hopefully Gen 9 brings us that much closer.  

    If it looks like photo, then its artistic value will be judged by photographic standards and that one looks utterly trivial. Only the husband or maybe mom and dad would give damn about it. I'm sure there are fans of hyperrealism who would see it differently and would love the fact that you could mistake this painting for a photo, but artistic value comes from getting to see something through the eyes of the artist, some interpretation of reality, but if the painting is just a dull photograph then what's the point? The artist replicated what he saw through the camera, to such perfection that the viewer cannot see a difference. There is no more artistic value in that than there would be in the photo. So if it's not an interesting photo, then what's really the point as an art piece? Just as a show of craftsmanship it is amazing but is that enough? 

    I guess it comes down to your opinon.  I actually see and admire the talant in being able to replicate all the details of a dull photo using an airbrush or whatever medium you choose to use.  I am a fan of hyperrealism in any form.   

    Any crappy "Poser" art has more artistic value than this if you ask me. While the craftsmanship is obviously lacking compared to this technically masterful painting, at least I know that someone has put together a scene creatively and not just replicated a dull photograph.

    I'm reminded of what I was thinking a while back. Once we reach absolute photorealism and our renders are not distinguishable anymore from real photos, what will the point of creating CG renders be then? Are we perhaps striving for the wrong thing? Wouldn't it be better to have some clear indicators that what we're looking at is not real? Sure we could make fantastical scenes that can't exist in real life, but what if you just wanted to make a simple portrait? Someone might think yeah this is a nice photo, but the fact that it took a thousand times longer to create it isn't going to matter one bit. Even people who are into CG renders would eventually lose interest as they couldn't tell any more if something is a render at all. We couldn't say "this looks so realistic" any longer, unless we knew beforehand that something is indeed a render. Wouldn't this change everything for the worse?

    Deep thoughts laugh

    Maybe when we reach being able to technically make something absolutely photo real we can put much more concentration on scene content and story telling.  You can do this now, but backed up with absolute photorealism sure makes it a much sweeter deal to me!   As far as making something look extremely real and using it for nefarious reasons....its scary for sure.  Should we just stop trying to shoot for absolute realism?  As long as people want photo realism and willing to either pay for it in either time (learning skills) or money its going to happen.  As they say "the genie is out of the bottle".  

    Very deep thoughts my friend  cheeky

    I admire the talent very much, it just bums me out that he's using that insane talent to create such an overall dull work. I mean you said it right there, you'd use such skills to focus more on content and story telling, aka doing something interesting with it.

    But hey, maybe that was some type of commission work and he had no say in the photo per se. Or the goal was exactly to invoke this kind of reaction. That would make it art right there. I did have a reaction after all. I wouldn't have if I didn't know this was a painting though, that's kinda my point I guess.

    Good point, I guess its whatever your opinion is.   I would most certainly love to focus more on content and story telling, but since realism is so super important to me...that is what comes first.  Its my personal choice I guess.  

    It could be a commission work deal or to invoke a reaction as you say...probably is.  I know he does classes all over the world so there must be something to it.  He has done other work with aircraft and machines that looks pretty good.   I think he has done realistic photo stuff using live models as well.  Me personally, I'm sticking to CG.  LOL

     

  • j cadej cade Posts: 2,310
    edited May 2020

     

    A thread talking about skin, photorealism and using filmic in blender? Back into the fray!

    AFAIK filmic is or is similar to a LUT, but is specifically geared towards accurately transforming 32bit render data into something resembling what you would expect to come out of a camera, whereas must LUTs are generally designed to take what is coming out of a camera (either 8 or 16 bit) and turn it into something pretty. Here's a great example image (from blenderguru I think)

    its not just that the shadows are lighted look at the edges of the direct light see how the first has this really weird and ugly falloff?

    I actually happen to have a pretty good (if specific) example of where this really adds some realism. Render straight out of iray vs sending the exr over to blender and filmiced

    See those weird blown out hotspots in the bokeh? fugly and something I've never seen when taking a photo

    Also, thank you folks evangelizing spectral rendering, I had dabbled, but lacked some stick-to-itiveness. The SSS is indeed way more accurate, I can do 100% SSS without turning red or a glowing mess! Mind you because 100% SSS doesnt use the difuse map, if theres funkiness with the Translusency map things get weird fast, Although I have some preliminary expiriments that look hopeful, even for darker skin (which Iray is way more fiddly about)

     

    And speaking of the fray, here's my hat in the ring, as it were. No lofi look for me, I love me some dramatic lighting

     

    Skin using 100%SSS Strand based hair, strand based eyebrows, velus hair... so many strands (in retrospect the velus hair might be a bit to strong, but it didnt look so clear until halfway through rendering and i decided I'd done enough re-renders). I'm also not fully satisfied with the hair, mostly trying to get baby hairs round the hairline. Sadly, you can't control the thickness if individual strands with a textuure, which I think would have helped (you cant do this in blender either and it makes me sad) I'm actually quite satisfied with the eyes, though, although this isnt really a shot that shows them off

     

    One thing that was mentioned earlier that I think needs to be repeated, the morph is super important. I've started sculpting morphs (this lady's head is a custom morph for instance) and one thing I've noticed is that If i don't work from reference, a lot of the features end up sort of this is the concept of "nose" rather than an individual's nose. (if that makes any sense at all)

    bokeh.jpg
    874 x 495 - 79K
    victoriafinal.jpg
    1000 x 1500 - 254K
    Post edited by j cade on
  • j cadej cade Posts: 2,310
    edited May 2020

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    victoria filmic compare.jpg
    1333 x 1000 - 265K
    Post edited by j cade on
  • Sven DullahSven Dullah Posts: 7,621
    j cade said:

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    This looks very nice indeed!!

  • PaintboxPaintbox Posts: 1,633

    j cade, excellent stuff, going to try this! 

  • PadonePadone Posts: 3,700

    This is just to let you people know that basic support for HD characters in blender is here. Now you can import and render HD characters with cycles with iray materials preserved fine. Below an example.

    https://www.daz3d.com/mutation-morphs-hd-for-genesis-8-female

    https://www.daz3d.com/forums/discussion/comment/5580371/#Comment_5580371

  • davidtriunedavidtriune Posts: 452
    j cade said:

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    This is amazing. If you dont mind could I also share your other work here? I think the hair you render is some of the best I've ever seen.

    This is strand based hair rendered in blender.

  • shavonnewshavonnew Posts: 34

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

    LunaRedJacketHandsReachingOut_Camera.jpg
    1080 x 1350 - 171K
  • lilweeplilweep Posts: 2,489
    j cade said:

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    This is amazing. If you dont mind could I also share your other work here? I think the hair you render is some of the best I've ever seen.

    This is strand based hair rendered in blender.

    that hair is amazing

  • davidtriunedavidtriune Posts: 452
    edited June 2020
    shavonnew said:

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

    The overall color looks very realistic but the lips are the main giveaway, look a little like wax or painted. Good job.

    edit: edited out wrong advice

    Post edited by davidtriune on
  • PaintboxPaintbox Posts: 1,633
    edited May 2020

    New day, new test.

    This is through the filmic blender setup from DavidTriune. Thanks David!!  I love how it gets rid of the plasticky / waxy look of standard iray.

    - This time I tried to go for a simple lightbox setup, since there is a ton of reference on the net. Used Charlotte 8 skin, Nazar brows and Delphi Hair. The eyes are Babina 8.
    - I also noticed while mucking about in Affinity that a slight high pass filter also helps to further bring out the good kind of detail.

    Still a lot to learn (looking at David and J Cade work)

    What is a good vellus hair product?

    charlotte-test-01.png
    1400 x 1400 - 5M
    Post edited by Paintbox on
  • shavonnewshavonnew Posts: 34
    shavonnew said:

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

    The overall color looks very realistic but the lips are the main giveaway, look a little like wax or painted. Good job.

    Most iray renders aren't using accurate  gloss, namely they don't use fresnel reflections, which is a type of gloss that appears on every object in the real world.

    You can turn on this gloss by turning on "top coat" and setting it to "custom". This setting uses "schlick's approximation" formula to approximate fresnel gloss. (The "fresnel" setting looks different and is probably for metallic materials not dialetric like flesh). Set TC color to white. Leave TC curve 90 and TC Exponent alone. You can leave TC curve 0 alone (unless you want to get very accurate but I don't recommend this. see this link for more info. Subsitute TC curve 0 with F0). Next, I'm not 100% sure this is how you do it but if anyone knows, feel free to correct: Dial in TC weight between 0-1 and TC roughness 1 - TC weight. So if you set TC weight to 0.35, set roughness to 0.65. If you have a roughness map, set it to both the TC Roughness and TC weight.

    This gloss replaces all other gloss, so you can just turn off dual lobe and glossy layered weight.

    in Blender it is very easy  to have physically based gloss. Using one node (principled shader) you pretty much set one value (roughness)  and leave the specular around 0.5.

    Thanks for the feedback David!
    I added the topcoat and made it glossier. It looks better now though I'm not sure if it made her anymore realistic. She still has an anime cartoony vibe to her and I can't place my finger on why..

    LunaRedJacketHandsReachingOut_Camera2.jpg
    1080 x 1350 - 1M
  • notiuswebnotiusweb Posts: 110
    edited May 2020

      

    shavonnew said:

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

    I don't know why yet, but this looks 'real' to me.  I think it is that clear glasses frame that looks real.  And the hair looks photo real, very good.

    But now the figure itself looks fake, but like a fake dummy that would be used in a TV show or movie.  So it looks like a real fake person, like a mannequin, if that makes sense.  Now, for this person, I think the skin looks color wise very mono-color, even if it has a very realistic hue, but the detail is very symetric looking, and then the lids look too flat, less organic if you will.  But a very good render.  

    I think again the glasses tell my brain "that is real" and then it is willing to concede then whatever looks off is a mannequin prop of some type.  The hands DOF may be helping also, like itit adds more 'depth'...Maybe it is obscuration-decoys like the glasses frame and hands that helps push images into the photoreal realm.

     

    Post edited by notiusweb on
  • notiuswebnotiusweb Posts: 110
    edited May 2020
    j cade said:

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    Both looks awesome...but in both cases,, something with the eyes - I immediately looked at the eyes and said, "not real".

    If I put my hand over the eyes, I say "Is that real?", and then I looked at the neck and said good, but then the chest below neck, "no shadow, not real"....I am not sure why I think this, it's what my instinct tells me.  As a render my brain says, "perfect execution", you know, but my brain says "not photo" so I am trying to convey that response.

    I think some shadow, or jewlery (necklace?) on the chest could decoy my brain into thinking real, and then with the eys, maybe the pupil size could be smaller, I don't know....or maybe a varying darkness on the eyes could decoy me.  I am not sharing this as a critique, I am just trying to convey my brain's rsponse so I myself can understand what maybe is a threshold for this phenomenon.  Because as a render it looks sick, but I really want to be tricked into seeing a photo smiley.

    EDIT - I looked again quick for an instinct response and that swirly reflection line on the eyes also looks too perfect, like "rendered".   Now, I do believe such reflection is possible in real life, but it did trip my brain subconsciously into examining the eyes further, if that makes sense.  My brain did not give that reflection shape an immediate "pass", as if a security guard at an airport is checking an ID.  Maybe if that swirly reflection was broken up maybe, or more sharp... twinkly?...it may convey the eyes moisture more effectively, I don't know....just brainstorming!

     

     

    Post edited by notiusweb on
  • davidtriunedavidtriune Posts: 452
    shavonnew said:
    shavonnew said:

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

    The overall color looks very realistic but the lips are the main giveaway, look a little like wax or painted. Good job.

    Most iray renders aren't using accurate  gloss, namely they don't use fresnel reflections, which is a type of gloss that appears on every object in the real world.

    You can turn on this gloss by turning on "top coat" and setting it to "custom". This setting uses "schlick's approximation" formula to approximate fresnel gloss. (The "fresnel" setting looks different and is probably for metallic materials not dialetric like flesh). Set TC color to white. Leave TC curve 90 and TC Exponent alone. You can leave TC curve 0 alone (unless you want to get very accurate but I don't recommend this. see this link for more info. Subsitute TC curve 0 with F0). Next, I'm not 100% sure this is how you do it but if anyone knows, feel free to correct: Dial in TC weight between 0-1 and TC roughness 1 - TC weight. So if you set TC weight to 0.35, set roughness to 0.65. If you have a roughness map, set it to both the TC Roughness and TC weight.

    This gloss replaces all other gloss, so you can just turn off dual lobe and glossy layered weight.

    in Blender it is very easy  to have physically based gloss. Using one node (principled shader) you pretty much set one value (roughness)  and leave the specular around 0.5.

    Thanks for the feedback David!
    I added the topcoat and made it glossier. It looks better now though I'm not sure if it made her anymore realistic. She still has an anime cartoony vibe to her and I can't place my finger on why..

    Looks better. BTW I didn't mean just the lips but the whole face, looks like you glossed just the lips.

    Maybe she has a blank stare look. Maybe have her look at the camera. Select both eyes and set "point to" to camera.

    Also the sweater could probably use some bump mapping.

  • PaintboxPaintbox Posts: 1,633

    For me it is the expression that throws me off a bit probably, could you try a different range of emotions? I notice sometimes some expressions don't click with a certain character. Her right hand also feels a bit unnatural as it is framed in the current render. Maybe less twisted inwards?

    shavonnew said:

    I've been fiddling with my latest render in photoshop but I'm still unable to reach the realism I'm looking for. I've been staring and staring at photographs of real women but  can't figure out what it is that makes her look cartoony. Any ideas?

     

  • davidtriunedavidtriune Posts: 452
    Paintbox said:

    New day, new test.

    This is through the filmic blender setup from DavidTriune. Thanks David!!  I love how it gets rid of the plasticky / waxy look of standard iray.

    - This time I tried to go for a simple lightbox setup, since there is a ton of reference on the net. Used Charlotte 8 skin, Nazar brows and Delphi Hair. The eyes are Babina 8.
    - I also noticed while mucking about in Affinity that a slight high pass filter also helps to further bring out the good kind of detail.

    Still a lot to learn (looking at David and J Cade work)

    What is a good vellus hair product?

    No problem! Looks like a big improvement, at least the lighting part. In my experience, using filmic blender (or any processing on the image) does not help with realism if your materials are poor. Filmic also brings out more detail when it's probably better to hide details for low-res textures of DAZ characters.

    The only vellus hair I know of are https://www.daz3d.com/peach-fuzz-facial-vellus-hair-for-genesis-3-female-s and https://www.daz3d.com/dforce-vellus-hair-for-babina-8

  • bluejauntebluejaunte Posts: 1,902

    Nice to hear from you j.cade! I honestly thought you were gone for good. smiley

  • shavonnewshavonnew Posts: 34
    edited May 2020
    notiusweb said:

    I don't know why yet, but this looks 'real' to me.  I think it is that clear glasses frame that looks real.  And the hair looks photo real, very good.

    But now the figure itself looks fake, but like a fake dummy that would be used in a TV show or movie.  So it looks like a real fake person, like a mannequin, if that makes sense.  Now, for this person, I think the skin looks color wise very mono-color, even if it has a very realistic hue, but the detail is very symetric looking, and then the lids look too flat, less organic if you will.  But a very good render.  

    I think again the glasses tell my brain "that is real" and then it is willing to concede then whatever looks off is a mannequin prop of some type.  The hands DOF may be helping also, like itit adds more 'depth'...Maybe it is obscuration-decoys like the glasses frame and hands that helps push images into the photoreal realm.

     

    Okay I added more shading to her eyelids and purple/blue/green hues to skin! 

    Looks better. BTW I didn't mean just the lips but the whole face, looks like you glossed just the lips.

    Maybe she has a blank stare look. Maybe have her look at the camera. Select both eyes and set "point to" to camera.

    Also the sweater could probably use some bump mapping.

    Ah yes I misunderstood haha. Okay added more gloss to her face, redid her eyes and gave her sweater a knitted bump map!

    Paintbox said:

    For me it is the expression that throws me off a bit probably, could you try a different range of emotions? I notice sometimes some expressions don't click with a certain character. Her right hand also feels a bit unnatural as it is framed in the current render. Maybe less twisted inwards?

     

    I fiddled with the face a little but this seemed to be the only expression my character can go that isn't the resting bitch face lol.

    Thank you all for the feedback! She's a little closer now, though still not yet completely photorealistic.

    LunaRedJacketHandsReachingOut_Camera3.jpg
    1080 x 1350 - 1M
    Post edited by shavonnew on
  • Rashad CarterRashad Carter Posts: 1,803
    shavonnew said:
    notiusweb said:

    I don't know why yet, but this looks 'real' to me.  I think it is that clear glasses frame that looks real.  And the hair looks photo real, very good.

    But now the figure itself looks fake, but like a fake dummy that would be used in a TV show or movie.  So it looks like a real fake person, like a mannequin, if that makes sense.  Now, for this person, I think the skin looks color wise very mono-color, even if it has a very realistic hue, but the detail is very symetric looking, and then the lids look too flat, less organic if you will.  But a very good render.  

    I think again the glasses tell my brain "that is real" and then it is willing to concede then whatever looks off is a mannequin prop of some type.  The hands DOF may be helping also, like itit adds more 'depth'...Maybe it is obscuration-decoys like the glasses frame and hands that helps push images into the photoreal realm.

     

    Okay I added more shading to her eyelids and purple/blue/green hues to skin! 

    Looks better. BTW I didn't mean just the lips but the whole face, looks like you glossed just the lips.

    Maybe she has a blank stare look. Maybe have her look at the camera. Select both eyes and set "point to" to camera.

    Also the sweater could probably use some bump mapping.

    Ah yes I misunderstood haha. Okay added more gloss to her face, redid her eyes and gave her sweater a knitted bump map!

    Paintbox said:

    For me it is the expression that throws me off a bit probably, could you try a different range of emotions? I notice sometimes some expressions don't click with a certain character. Her right hand also feels a bit unnatural as it is framed in the current render. Maybe less twisted inwards?

     

    I fiddled with the face a little but this seemed to be the only expression my character can go that isn't the resting bitch face lol.

    Thank you all for the feedback! She's a little closer now, though still not yet completely photorealistic.

    Yep, getting closer indeed!

  • Rashad CarterRashad Carter Posts: 1,803
    j cade said:

     

    A thread talking about skin, photorealism and using filmic in blender? Back into the fray!

    AFAIK filmic is or is similar to a LUT, but is specifically geared towards accurately transforming 32bit render data into something resembling what you would expect to come out of a camera, whereas must LUTs are generally designed to take what is coming out of a camera (either 8 or 16 bit) and turn it into something pretty. Here's a great example image (from blenderguru I think)

    its not just that the shadows are lighted look at the edges of the direct light see how the first has this really weird and ugly falloff?

    I actually happen to have a pretty good (if specific) example of where this really adds some realism. Render straight out of iray vs sending the exr over to blender and filmiced

    See those weird blown out hotspots in the bokeh? fugly and something I've never seen when taking a photo

    Also, thank you folks evangelizing spectral rendering, I had dabbled, but lacked some stick-to-itiveness. The SSS is indeed way more accurate, I can do 100% SSS without turning red or a glowing mess! Mind you because 100% SSS doesnt use the difuse map, if theres funkiness with the Translusency map things get weird fast, Although I have some preliminary expiriments that look hopeful, even for darker skin (which Iray is way more fiddly about)

     

    And speaking of the fray, here's my hat in the ring, as it were. No lofi look for me, I love me some dramatic lighting

     

    Skin using 100%SSS Strand based hair, strand based eyebrows, velus hair... so many strands (in retrospect the velus hair might be a bit to strong, but it didnt look so clear until halfway through rendering and i decided I'd done enough re-renders). I'm also not fully satisfied with the hair, mostly trying to get baby hairs round the hairline. Sadly, you can't control the thickness if individual strands with a textuure, which I think would have helped (you cant do this in blender either and it makes me sad) I'm actually quite satisfied with the eyes, though, although this isnt really a shot that shows them off

     

    One thing that was mentioned earlier that I think needs to be repeated, the morph is super important. I've started sculpting morphs (this lady's head is a custom morph for instance) and one thing I've noticed is that If i don't work from reference, a lot of the features end up sort of this is the concept of "nose" rather than an individual's nose. (if that makes any sense at all)

    Surprised no one mentioned how incredible she looks. If you could put her in an environment it would be sweet to see. Personally I'm not a big fan of skin studies in black environments because in real life a single light in an perfectly dark room or space is physically impossible or at the very least implausible. Also it onluy shows you the way the skin looks as lit by the tight bright specualr type lighting that gives us the refelctions. Basically its easier to pull it off when the only light hitting the model is "direct" lighting. However I think the bigger challenge is getting skin to look as it should with indirect lighting in place as well. I'd LOVE to see you take her and put her in a real "place" and see how she looks. Love what I'm seeing here!

  • Rashad CarterRashad Carter Posts: 1,803

    That is insane. Although at that point I really have to wonder, why go through all the trouble? Just make a photo. I mean they probably did and he then replicated through airbrushing? But why? What's the artistic value beyond just the insane show of skill? I don't see any at least in this case.

    Yeah, I get what you are saying.... why reproduce an airbrushed version of a photo?  I think he has done work from actual live models as well.  My reason for posting it was to illustrate whether you are copying a photo, or airbrushing something from your head to produce something so lifelike is amazing.  I would love to reach that level in CG personally and after seeing some of the work posted I have to admit some have reached that level.  

    Artistic value? I guess it is based upon whatever your opinion of what work has artistic value or what does not.    Some would say that if you buy an off the shelf pre-made CG model, add textures, pose and render....what artistic value is that?  As if the only artistic value resides with someone who creates a realistic figure from the top of the head with no so called cheating by using photos, or references or God forbid textures from a real human being.   Of course this is ridiculous as using human reference is not only needed, but the best way to replicate realism in my opinion.    

    Me personally, I am trying to replicate a realistic photo....so I would pick a human reference and use a CG model with similar skin tone, hair and lighting to come as close to that photo as possible.   

    I think your figures are absolutely top notch and have helped bring DAZ figures to a new level of realism I am looking for.   The rest is up to DAZ and hopefully Gen 9 brings us that much closer.  

    If it looks like photo, then its artistic value will be judged by photographic standards and that one looks utterly trivial. Only the husband or maybe mom and dad would give damn about it. I'm sure there are fans of hyperrealism who would see it differently and would love the fact that you could mistake this painting for a photo, but artistic value comes from getting to see something through the eyes of the artist, some interpretation of reality, but if the painting is just a dull photograph then what's the point? The artist replicated what he saw through the camera, to such perfection that the viewer cannot see a difference. There is no more artistic value in that than there would be in the photo. So if it's not an interesting photo, then what's really the point as an art piece? Just as a show of craftsmanship it is amazing but is that enough? 

    Any crappy "Poser" art has more artistic value than this if you ask me. While the craftsmanship is obviously lacking compared to this technically masterful painting, at least I know that someone has put together a scene creatively and not just replicated a dull photograph.

    I'm reminded of what I was thinking a while back. Once we reach absolute photorealism and our renders are not distinguishable anymore from real photos, what will the point of creating CG renders be then? Are we perhaps striving for the wrong thing? Wouldn't it be better to have some clear indicators that what we're looking at is not real? Sure we could make fantastical scenes that can't exist in real life, but what if you just wanted to make a simple portrait? Someone might think yeah this is a nice photo, but the fact that it took a thousand times longer to create it isn't going to matter one bit. Even people who are into CG renders would eventually lose interest as they couldn't tell any more if something is a render at all. We couldn't say "this looks so realistic" any longer, unless we knew beforehand that something is indeed a render. Wouldn't this change everything for the worse?

    Deep thoughts laugh

    Clever observations!

    The Elton John movie demosntrates for me exactly WHY we often need to strive for photorealism in CG....because its useful from a commercial standpoint. Basically, if you CAN take a photo of what you need, then by all means take a photo. However there will be many instances when an actual photo is simply not feasible, and it is in these instances that photorealistic renders are necessary.

    Photorealstic environments are how architects convince investors that a project is worthwhile. People can't just picture it for themselves, their imaginations often need a little "helper" in the right direction.. welcome to the archvis world.

    We could use realistic looking humans to sell things too. As a career Actor I'm worried by this, but then, I'm also concerend for all those stunt men and woman who loose their lives pulling off stunts that probably could have been created in CG if the technology was truly up to the challenge of making it look real. I'm sure the stuntperson with a newborn disagrees...or does she?

    I often consider Magic shows and magicians... David Copperfield, David Blaine, Pen and Teller etc....

    Why do we watch magic?  Because the magician manages to make it look "real." That combination of "looks real, but can't be" is what gets the butterflies going in our stomachs. Realism can be used for mundane depictions, or for very dramatic ones. The more dramatic and still realistic, the better!

  • CinusCinus Posts: 118
    edited May 2020
    j cade said:

     

    Skin using 100%SSS Strand based hair, strand based eyebrows, velus hair... so many strands (in retrospect the velus hair might be a bit to strong, but it didnt look so clear until halfway through rendering and i decided I'd done enough re-renders). I'm also not fully satisfied with the hair, mostly trying to get baby hairs round the hairline. Sadly, you can't control the thickness if individual strands with a textuure, which I think would have helped (you cant do this in blender either and it makes me sad) I'm actually quite satisfied with the eyes, though, although this isnt really a shot that shows them off

     

    One thing that was mentioned earlier that I think needs to be repeated, the morph is super important. I've started sculpting morphs (this lady's head is a custom morph for instance) and one thing I've noticed is that If i don't work from reference, a lot of the features end up sort of this is the concept of "nose" rather than an individual's nose. (if that makes any sense at all)

    @jcade I agree. The eyes look fantastic! Would you mind sharing the secret?

    Post edited by Cinus on
  • nonesuch00nonesuch00 Posts: 18,131

    I saw David Blaine one in a big clear plastic box sleeping over the Thames in midday sun when I was walking over a bridge once. It was not exactly a magical moment. laugh

  • shavonnewshavonnew Posts: 34

    Does she look photo real? Somehow I think she looks a little more real here in the studio setting...but still not real enough!
    Attached a half-assed render cause I got bored of waiting.

    LunaRedJacketHandsReachingOut_Camera3.jpg
    1080 x 1350 - 1M
    test.jpg
    1350 x 1080 - 1M
  • j cadej cade Posts: 2,310
    j cade said:

     

    Surprised no one mentioned how incredible she looks. If you could put her in an environment it would be sweet to see. Personally I'm not a big fan of skin studies in black environments because in real life a single light in an perfectly dark room or space is physically impossible or at the very least implausible. Also it onluy shows you the way the skin looks as lit by the tight bright specualr type lighting that gives us the refelctions. Basically its easier to pull it off when the only light hitting the model is "direct" lighting. However I think the bigger challenge is getting skin to look as it should with indirect lighting in place as well. I'd LOVE to see you take her and put her in a real "place" and see how she looks. Love what I'm seeing here!

    Technically she's in a dark room (a dark grey cube to be precise) I try to avoid rendering in voids because you end up losing some bounce light (there would be less on the undrside of her neck for instance)

    I was going for studio portrait with a ring light (here's an internet example)

    But my setup is still is a bit abstracted I definitely want to stick her in more of a scene, although even in scenes my aestetics tend more photoshoot than candid (or if not photoshoot, dramatic even in the realworld most of the photos I take are of dramatically backlit trees)

     

     

    Cinus said:
    j cade said:

     

    @jcade I agree. The eyes look fantastic! Would you mind sharing the secret?

    The eyes are Parris' macro eyes with some tweaked settings, but the main key is that I switch the sclera material to eyemoisture, because I fund the gen3 way of dealling with eyes so much less fiddly its not even funny. I can use "thinwalled off" without getting weird shading on the sclera. Morph wise there's the cornea buldge morph and an iris depth morph. Looking at further reference I think I need to turn the specular strength down which is a new one for me

    j cade said:

    Aaaaaaaaad here is a full filmic comparison. Left is straight out of Iray, and right is run through filmic (with the look set to high contrast) Filmic just looks more photolike to my eyes

     

     

    If you also render out a depth pass you can also do some nifty compositing tricks (as long as you're okay with some nodes) like using the defocus node to do depth of feild in post, I do this all the time when rendering buildings. Its so much faster its not even funny. (you can also use it sparingly to ad a tiny bit of dof to portraits, but because of the quality of the depth maps Iray outputs, it doesn't play great with transmaps>hair)

    This is amazing. If you dont mind could I also share your other work here? I think the hair you render is some of the best I've ever seen.

    This is strand based hair rendered in blender.

    I still find blender better for long hair, as I have never been able to get clumping to work in Garibaldi on long hair without clumping through the skull (does someone here know how to fix this? I will love you forever) Its also easier to combine multiple hair groups in blender. some of those small strands were actually a seperate hair object and you can edit both while seeing the other one and all the objects in the scene, which you can't in Studio. On the other hand, Studio is great for short hair, and while you can't have multiple interacting hair objects you do have selection sets, which blender doesnt and it makes me sad

  • bleach2290bleach2290 Posts: 0

    Stumbled onto this thread and figured i'd drop one of mine in here for some feed back myself.

    First one is Ungraded with no post work done inside daz with nothing but the iray shader with a hdr for lighting ,the others are just Luts that i thoguht would look half decent.

    Let me know what you think and any feedbackw ould be nice always looking to improve.

    Base Skin Shader Main Character 1.png
    2560 x 1440 - 3M
    Base Skin Shader Main Character 1 Lut 001.png
    2560 x 1440 - 3M
    Base Skin Shader Main Character 1 Lut 002.png
    2560 x 1440 - 3M
    Base Skin Shader Main Character 1 Lut 003.png
    2560 x 1440 - 3M
  • CinusCinus Posts: 118

     

     

    j cade said:
    Cinus said:

    @jcade I agree. The eyes look fantastic! Would you mind sharing the secret?

    The eyes are Parris' macro eyes with some tweaked settings, but the main key is that I switch the sclera material to eyemoisture, because I fund the gen3 way of dealling with eyes so much less fiddly its not even funny. I can use "thinwalled off" without getting weird shading on the sclera. Morph wise there's the cornea buldge morph and an iris depth morph. Looking at further reference I think I need to turn the specular strength down which is a new one for me

    @jcade Thanks for the info! I have tried setting the sclera to the eye moisture material, but I do not think I have tried switching off "thin walled". I have also not tried using an iris depth morph. Will definitely give this a try.

  • NylonGirlNylonGirl Posts: 1,817

    I remember some time ago, in another thread, there were people posting images said to have a high degree of realism. And there were other people who would list all of the flaws in the images they could think of. The person who created the thread, which I can no longer find, observed that for all of the things they listed as obvious reasons images were fake (poor lighting, perfect lighting, lack of emotion, etc), one could go onto the internet and easily find an example of an actual person who was poorly lit, perfectly lit, or lacked emotion, or any number of other things, yet was clearly a real person. So it became clear that whatever was the key to realism, was not obvious.

    I also noticed a recurring theme was people saying there was something wrong with the eyes of all DAZ figures. There was also a meme on the social networks that stated the Disney Movie, Polar Express, had CG characters with “lifeless” eyes. I also tried showing images to my dad, who has no interest in computer generated images as a hobby, and nothing to do with DAZ or Poser. With the very first image I showed him, he seemed surprised when I told him it was computer generated. It was a small image so it would be hard to look closely at it, and it was not the main focus of our conversation. But with every other image I showed him, he said it was obvious it wasn’t a real person. When I told him many people would say the images were not realistic enough yet nobody could seem to say exactly why, he said the eyes were not right.

    I am not one to try to create my own characters and textures and whatnot. I leave that to people like @jeff_someone and @Bluejaunte. I use products that I think produce good results. And I don’t have any special tricks. But I do have two working theories based on the things I mentioned above. One is that a key factor in people believing an image is a photograph of a real person, is to not tip them off that the image is a computer generated person. They could find flaws in a person standing in front of them if they were convinced it wasn't a real person. And the other theory is the eyes have issues. But I don’t know what kind of issues they have. I think if I want to convince people my image is a photograph then the figure had better be wearing glasses or not looking at the camera.

    @Soldato mentioned a desire to create images like jeff_someone except with a “2020 look”. Also, there was some conversation about HDRI backgrounds not matching the figures in the image. I noticed the current trend in high end phone cameras is a blur effect, and some large lenses on SLR camera have an inherent “bokeh” effect. So it seems the trend in the high end cameras of 2019 and 2020 is to have a background that is unnaturally more blurred than the foreground image. Perhaps we’ve reached a point that the mismatched HDRI backgrounds will reasonably simulate the blur/bokeh effects of the current high-end cameras and produce the desired “2020 look”. I am not going to spellcheck all of this.

Sign In or Register to comment.