Loading...
> >

Photo-real characters. A different approach.

24567…26

Comments

  • Serene NightSerene Night Posts: 17,680

    It doesn’t look real to me. The shadows are wrong and the skin is too even and her buttocks too stiff on the edge of the fountain. 

  • outrider42outrider42 Posts: 3,679
    edited March 2018
    Divamakeup said:

    I think you're right about the eyes. It's not just the eyes though, I think. Real humans have a TON of asymmetry in just about every part of their faces and bodies. Most 3D characters have too much symmetry and the brain picks up on that even if we aren't aware of it consciously. 

    People have so much asymmetry that most people if you copy half their face and flip it/mirror it, they often look like a different person.

    Original face - without a doubt a real person:

    But once you mirror one side of the face, we can tell something's not quite right.

    Left side mirrored:

    Right side mirrored:

    If you take just about any face, and mirror it, your brain tells you that something's not quite right.

    I think too much symmetry plays a big role in how we're able to spot a render vs a photograph.

    What if I told you I think the first non mirrored photo looks the worst? Her eyes are TOO asymmetrical, possibly one eye is just slightly more closed than the other, and she is not looking perfectly straight at the camera, either, which contributes to it. The problem with all of the photo flip tests is that they do not take into account the width of the face. Just look at Megan in those flips, why is her face wider in the 3rd pic, and more narrow in the 2nd? They are not perfect flips, that is why. She has much more space between her her eyes...again, why? So when people look at these pictures, they reject them not because of symmetry, but because of the abnormal space between the eyes, and the abnormal width of the face. The 2nd pic does the best job, and looks the best because of it. But either way, I have never thought Fox was even remotely attractive, so this isn't a fun test to take.

    Studies have shown repeatedly that humans like symmetry. Our species looks for symmetry. Heh, we not only look for symmetry, we even perceive symmetry when its not really there. You only spot these things when you look real close at someone.

    http://www.bbc.co.uk/bbcthree/article/6adaed45-2f21-4a7b-942a-5c40a7181dd6

    Its part of us. When somebody is far too asymmetrical, they seek out surgery to fix it. Lop sided breasts are considered not ideal.

    Here's another article, the "science of sexiness." Which points out very quickly that symmetry is part of this ideal.

    https://www.telegraph.co.uk/science/2016/03/12/the-science-of-sexiness-why-some-people-are-just-more-attractive/

    https://www.telegraph.co.uk/news/science/science-news/3343640/Symmetrical-human-faces-are-more-beautiful.html

    So it makes logical sense that 3D models would gravitate towards symmetry, because most 3D models are shooting for ideals rather than purely copying reality.

    Now I am not saying that asymmetry is bad. I am providing a counterpoint to the argument in this thread that symmetry is automatically less real. Our species is what it is because we look for symmetry.

    So no, in my conclusion, symmetry is not why a render looks like a render. It is the skin, the eyes, and the hair. Well, that's pretty much everything! Daz Iray has come along decently, but this is still not what human skin looks like. Far from it. All of my gripes with Genesis 8 convey this. The waxy skin in promos, skin that glistens too perfectly. Hair is transmapped. Eyes too glossy like they are glass. All of these add up.

    Skin has pores. You never see real pores in Daz. You see bumps from microbump maps, but not pores. You might see pores painted on base maps, but they have no depth. That is just a limit of tech at this time. Look at your own hand in front of the monitor. Look at the lines in the skin. You don't see these lines in Daz figures. The best ones come close. My fingernails have very subtle ridge in them that I have never seen in a Daz figure.

    Another consideration is how large renders can be. If you keep renders small you can hide some of these things in the lower resolutions. Think about what happened when TVs became HD, and programs first started filming in HD. It was a mess. They had to rethink how to do makeup just for HD programming. We need to think about that for Daz, and as renders grow larger, how to make them look better. It is no coincidence that the image Diva linked is a bit far from the face. Its certainly a great image, but would that character hold up as well with the camera on top of her face? (And I'd write the bump on her back off as being from a thin woman in such an extreme pose.)

    Post edited by outrider42 on
  • notiuswebnotiusweb Posts: 110

    Another technique as opposed to adding in false textures on a real photo to see what triggers the "AH, it's all fake..." moment, could be to take a real photo, and render your own figure in the exact same pose and size, and then start adding in the real things via layering the 2 images in photoshop/Gimp.  Then you can see what makes the image "look real".

    Ie- I render same pose as the Meagan Fox face, put my character on top as a PNG, and erase features off my character to see how the eyes and skin being brought in affaect my perception of the 'else' of the fake character.

    Like a process of reduction to see what is holding up your render from looking truly real.

    I know 'they' always say, "Well, your pose has to look real..."  But, this is not exactly a science, because you could take a real photo of someone, in an unatural stiff pose, and have it be in black and white, at low resolution, and you would still believe it to be real.   What the heck....Maybe...It might be the proportions of the actual mesh itself, or the lighting, with respect to the total light space in the actual photo.

    Right?  I mean, high detail pores are not present in an old blurry old picture that looks real.  Maybe renders themselves are flawed becauyse there is no distortion in the cam lens, you know?

    Pictures from the 1800s sometimes look like renders, flat lighting, stiff poses and expressions, limited color, but they still look real....why?....a lighting balance?

    (Is it known why the Mona Lisa comes off as looking more 'real', say vs some other ancient art?)

    WHY????!!!!  cheeky

     

    PS - THIS THREAD IS AWESOME!!!!

  • DrNewcensteinDrNewcenstein Posts: 816
    Gregorius said:
    DrNewcenstein said:
    The bump mapping on the lips is not different from the rest of the face. The freckles are nice, but the bump mapping of the lips throws it off. As well, the whites of the eyes are too white, almost luminous.

    In the second photo, the hands are the problem. There's no life to them. Hands are the dead giveaway for all CGI to date. Squaresoft (Square-Enix, etc) have always had the best video game CGI, and even their Final Fantasy movies have been great, except for the hands. The hands always ruin it because no one can do realistic hands. Even the latest Marvel movies fell victim to the hands for the CGI stuff.

    If you tell a model to pose for a shot, they hold a pose. If you snap a shot of someone performing an action, the result is drastically different.

     

    Thanks for your comments!  I anticipated a few possible critiques, but that first one about the lips wasn't one of them, perhaps because they stand out at least somewhat better on the bump map itself (attached in case you're curious).

    After reading your comments, I Iooked up a real photo of human eyes with reasonably similar lighting, sampled the eyewhite in a well-lit spot, and looked at the HSB values.  I then sampled the eyewhite in a well-lit spot on the above render and looked at its HSB values.  Strangely enough, the brightness of my rendered eye actually seemed to be lower than that of the photographed eye.  The former hovered on average around the high 60s or low 70s, while the latter was consistently in the low 80s.  This appeard to be the trend after taking a few samples from a couple different eye photos.

    Of course, there's no guarantee that the lighting was sufficiently equivalent, but it seems that one or both of us have skewed expectations for sclera color.  To me, the eyes of many 3D characters look a bit too dark, perhaps in a misguided attempt to bring out the reflections better.  I also wonder if the contrast with my character's rather tanned skin might be contributing to the discrepancy in our perception.

    Although I'm not sure I see what you see with the hands, I don't think there's much I can do about them yet.

    Interesting notes on the sclera colors between a photo and the render. Perhaps it's the falloff where it curves?

    For the hands, it's not the texture, it's the pose. Her arms are reaching out, but her hands are not. They're not "doing anything", or at least they're not doing what the arms are doing. Like I said, there's a difference between posing and performing an action, and in 3D, that's one of the "Big Things" that hurt realism.

  • Serene NightSerene Night Posts: 17,680
    j cade said:
    Serene Night said:
    CG Female faces often almost seem symbolic. Rather like someone took a slightly stylized very 2d illustration of a face and then tried to reconstruct a 3d face from it. Like the nose, say, can't stray to far from the platonic ideal of what a nose is or it won't be recognizable.

    You add that to the maket pressure for the female characters to be pretty, and...

    Of couse the weirdest part from my perspective, is I personally find that there's way more variation that women can have and still be aestetically appealing... I have a way more specific type for men.

    You put it very succinctly. I tend to agre. I tend to like a different sort of male face than what trends on DAZ. So I make my own.

  • fastbike1fastbike1 Posts: 4,078

    @notiusweb "(Is it known why the Mona Lisa comes off as looking more 'real', say vs some other ancient art?)"

    Yes it is known. (of course the answer also depends on which ancient art is your reference. FWIW, DaVinci isn't considered "ancient")

  • notiuswebnotiusweb Posts: 110
    fastbike1 said:

    @notiusweb "(Is it known why the Mona Lisa comes off as looking more 'real', say vs some other ancient art?)"

    Yes it is known. (of course the answer also depends on which ancient art is your reference. FWIW, DaVinci isn't considered "ancient")

    Right, you give me food for thought!....

    Alot of the fanfare of an image is based on timing...

    In the past, Mona Lisa probably seemed 'more real' than other images.  But why, no one can say with definite 100% metrically repeatable scientific accuracy why.  You can say what you feel like it is, but not in a metric way.  ie - the expression is subtle, thus looking more real.  But this type of thing doesn't make Mona Lisa look 'photo-real', as in, "that was a living person captured in an image format".  Nor does it capture a specific metric which can be reproduced.  And, in fact, any variation of the subtle expression by DaVinci would not have made it appear any more photo-real.  So, this reveals there is often confusion when discussing photo-real:  (1) How it connected with you on an element of reality vs. (2) Did it trick your brain into thinking it was photo-real...

     

    ie:

    Q - Do the character images in this thread convey interesting and thought provoking elements of the reality? 

    A - YES

    Q - Under scrutiny, do the images look like they were photos?

    A - NO

    Thus they are not the same thing at all.  An image can convey a perspective on reality, and/or convey emotion that is so real.  But the image itself is easily picked by the brain as not photo-real, under scrutiny.

     

    The only image that passes is Meagan Fox.  Yet, this doesn't have to mean it is not possible to render "photo-real characters", which is in the title of this thread.  It just means our attempts have yet to be proven successful, under scrutiny.  Interesting to imagine ways of metrifying, in a predictable and measureable way, that can be reproduced on-demand, an adjustment to a character render to make it look photo-real.  

     

     

  • HavosHavos Posts: 5,404
    Divamakeup said:

    Here's one of the few images that made me look twice and had to really LOOK at it before I realized it was a 3D render. On first glance though, it had me fooled for a minute. 

    The body folds help. And I think the fact that it's kind of an unnatural awkard pose actually helps "sell it" as well - as most real people trying to sit in that pose will look a bit stiff or unnatural; so the fact that it's not really a normal pose that someone would sit in, helps - as we don't often see people sitting like that so we don't have a "quick reference" file in our memories to instantly compare it to other images we've seen with real people in that pose. The fact that the face is partially inverted also helps "sell it", I think, as we don't often see faces that are up-side-down. So our brain doesn't immediately recognise artificialness that would be easier to spot with a face that's vertical or in a position we normally see faces in. 

    I think the lump on the back definitely doesn't look right to me, and the hair is definite give away. Still, just browsing past it among other "photos" it wouldn't jump out immediately (at least to me) as 3D. 

    That is a nice image, and one of the few where the human looks more convincing than the surrounding environment. The razor sharp line of the falling water looks particularly odd.

  • OstadanOstadan Posts: 1,128
    Wonderland said:

    I think that we notice things others don't. Sometimes I show my more realistic renders to people who know nothing about 3D rendering and they just can't get it through their head that it's not a real person. When I tell them the girl doesn't exist, she's just pixels and polygons, their brain kind of explodes. They think I'm a photographer and I have to keep explaining that no, there never was a real girl, she just doesn't exist. Often making the image black and white or using Instagram type filters make the images look more real than a straight render. 

    Also, sometimes people don't really look hard at a picture.  Not long ago, I did a 'fan art' of a friend's story, which he posted to his blog.  One of the comments was from someone whose first reaction was, 'Wow, someone went to the trouble of getting actors to stage this scene?'  When that person really looked at the picture, they realized it was art.  But at first glance, for people who aren't in this hobby/field, I think quite a lot of renders would pass the 'first glance' test.

    Thinking a little more about my picture in question, I think the furniture (lots of glass and reflections) and lighting helped a lot.  To quote Pooh-Bah, merely corroborative detail intended to give artistic verisimilitude to an otherwise bald and unconvincing narrative.

  • aaráribel caađoaaráribel caađo Posts: 690
    CypherFOX said:

    Greetings,

    I've more than once considered postworking my images to have the flaws that a cellphone camera in sketchy light adds in, just to see if it increased the versimilitude of the images.  Making the white balance worse, and stuff like that.

    Also almost no human holds a camera perfectly flat; in studio shots you can get away with that, but everywhere else there's a hint of tilt or angle.  Also some hair is (again) too perfect, typically...not enough flyaways, arches or bridges, details.

    But it's all part of the same thing; asymmetry and imperfection, eyes that seem real, picture quality, subtle inconsistencies everywhere that cue us to what's real and what's not.  It's kinda cool that we're at the point where that's what it's going to take to get there, having gotten past the body/face form itself, the intricacies of sub-surface details, bumps, hairs (both body and atop the head) and more.  Now we're down to the really subtle, hard to pin down things.

    That's one of the reasons a _lower_ quality image can seem realer; its own imperfections hide the too-perfect rendering.

    --  Morgan

     

    One of my more successful images happened when I set depth of field, forgot abou it, then shifted the pose, so the focal point was just behind the character. That made for a very spontaneous look. Depth of field is great for hiding flaws in a photorealistic way, as well as drawing focus to the parts of the image you want featured. 

  • FSMCDesignsFSMCDesigns Posts: 12,783
    Wonderland said:

    I think that we notice things others don't. Sometimes I show my more realistic renders to people who know nothing about 3D rendering and they just can't get it through their head that it's not a real person. When I tell them the girl doesn't exist, she's just pixels and polygons, their brain kind of explodes. They think I'm a photographer and I have to keep explaining that no, there never was a real girl, she just doesn't exist. Often making the image black and white or using Instagram type filters make the images look more real than a straight render. 

    I love when this happens. I show co workers digital imagery all the time and this usually happens, LOL

    Diva, that is a great image. For me, the background and props take away from the realism, but she looks awesome!

  • aaráribel caađoaaráribel caađo Posts: 690
    Divamakeup said:

    Here's one of the few images that made me look twice and had to really LOOK at it before I realized it was a 3D render. On first glance though, it had me fooled for a minute. 

    The body folds help. And I think the fact that it's kind of an unnatural awkard pose actually helps "sell it" as well - as most real people trying to sit in that pose will look a bit stiff or unnatural; so the fact that it's not really a normal pose that someone would sit in, helps - as we don't often see people sitting like that so we don't have a "quick reference" file in our memories to instantly compare it to other images we've seen with real people in that pose. The fact that the face is partially inverted also helps "sell it", I think, as we don't often see faces that are up-side-down. So our brain doesn't immediately recognise artificialness that would be easier to spot with a face that's vertical or in a position we normally see faces in. 

    I think the lump on the back definitely doesn't look right to me, and the hair is definite give away. Still, just browsing past it among other "photos" it wouldn't jump out immediately (at least to me) as 3D. 

    That's a great example of how we spot flaws when we're looking for them that we'd never notice otherwise. For me, her hair looks okay, but the lack of indent where her behind hits the fountain, and the lack of weight on her toes give it away. But I had to look for that. 

  • TaozTaoz Posts: 9,979

    .

    Havos said:
    Divamakeup said:

    Here's one of the few images that made me look twice and had to really LOOK at it before I realized it was a 3D render. On first glance though, it had me fooled for a minute. 

    The body folds help. And I think the fact that it's kind of an unnatural awkard pose actually helps "sell it" as well - as most real people trying to sit in that pose will look a bit stiff or unnatural; so the fact that it's not really a normal pose that someone would sit in, helps - as we don't often see people sitting like that so we don't have a "quick reference" file in our memories to instantly compare it to other images we've seen with real people in that pose. The fact that the face is partially inverted also helps "sell it", I think, as we don't often see faces that are up-side-down. So our brain doesn't immediately recognise artificialness that would be easier to spot with a face that's vertical or in a position we normally see faces in. 

    I think the lump on the back definitely doesn't look right to me, and the hair is definite give away. Still, just browsing past it among other "photos" it wouldn't jump out immediately (at least to me) as 3D. 

    That is a nice image, and one of the few where the human looks more convincing than the surrounding environment. The razor sharp line of the falling water looks particularly odd.

    Razor sharp edges is one of the things that can reveal that it's CGI. Unfortunately many artists still creates environments where you see such edges, I usually don't buy them for I think it ruins the whole thing and make it look amateurish even if the product in general is of high quality.

  • Ooze3dOoze3d Posts: 10
    edited March 2018

    Hi all!

     

    Wow! This thread really exploded! I'll answer to some very good points I've read after this post. I'm sorry I didn't add more in these past few days, but I was doing some research and trying new things.

     

    This is the first time I'm posting the character I'm working on and it's a very special moment for me because I've been tweaking her for a very long time.

     

    But first of all I want to talk a bit more about what makes a human render believable. As I said, I've only seen a few examples (most done with Arnold and lots and lots of maps and layers) which really made me stop and say "wow... that looks like a real person". Most of them took the route of playing with the symmetry and the flaws of an average everyday person to pump the realism (apart from amazingly complex skin, eye and hair shaders), but as I stated in my first post, I wanted to try a different path. Why? Because a very large chunk of the whole Daz/Poser community is focused on the creation of attractive women. Yes, there's another big chunk dedicated to males, but then again, it's usually good looking males, very close to what we can see in advertising and movies. So, not everyone, but a great number of people trying to reach for the look of a "perfect human".

     

    That being said, let's take a look at some pictures meant for advertising purposes. As we all can see, they're heavily photoshopped, with corrected symmetries, corrected colours, softened skins, erased marks, wrinkles and even pores, added highlights... Basically the full range of details used to make a render more lifelike is totally gone in these images, yet we can still identify these photos as real photos showing real human beings. Why?? I have to go back to my point about the eyes. If you take a look, the only thing we can't really mess much with in photoshop is the eyes. You can emphasise the colours, the contrast, even remove veins and details in the sclera, but the moment you touch the iris and the reflections, you're done. It doesn't look real anymore.

     

    These are examples of heavily treated faces, all with various levels of photoshopping which somehow still retain their humanity. The second one is the most obvious, yet still looks like a photo of a real woman.

     

    And here's an example of a poorly done eye treatment which renders the image totally useless because it doesn't look real and alive anymore, but more like a digital painting.

     

    So, I tried to take the same approach with my character. Yes, she has a nice and somewhat photorealistic skin shader, but I didn't add lots of freckles and moles and marks and bumps. Just the right amount. She also has a very detailed displacement map, but again not too obvious. The rest is done with a perfect looking, beautiful model in mind. And what specific area has a ton of work to make it look as real as possible? The whole eyebrows/lashes/eyes group. In fact that's what I've been working on for a few weeks now. As soon as I can I'll make another render with her previous eyes and you'll see the difference, but I wanted to show you how it's going so far. By the way, this is straight out of Daz Studio. No post work of any kind.

    You can click on it to see a 1080p version of the image.

    Now, the eyes are not absolutely perfect, but I think I'm getting close to a point where they're closer to the look of real eyes than many other examples. NOTE: Don't mind the teeth and specially the BRA, which has a terribly low poly count, but I wanted to add a bit of clothing while showing as much skin as possible. Also I just added the physical eyebrows this morning and they need some extra work as well. Also, no DOF, which would greatly improve the feeling of a real photo.

     

    Now, what do you think?

    Post edited by Chohole on
  • Ooze3dOoze3d Posts: 10
    edited March 2018

    Here's the version I just uploaded to DeviantArt (first one ever!). DrFronkonstinMD is my name there. Heavy filtering with NIK but it really pumps up the realism!

    Post edited by Chohole on
  • gederixgederix Posts: 390

    Agree the second version is far more realistic.

  • dreamfarmerdreamfarmer Posts: 2,128

    The fountain-girl: it's funny. The lack of weight on her toes and the lack of indent on her bottom both made me think: this is a contrived pose the model is holding, so of course her butt is tight as she lifts the weight off her toes. It looks, to me, like a real but photoshopped photograph. Sure, her skin tone is even and her hair is smooth and the fountain water has been painted in, but I'm just so used to that in actual photographs of attractive women!

    I like pretty people as much as the next person but I think, partially because of our trained expectations (as modern humans, not 3d artists), it's a lot easier to 'sell' somebody lovely than it is somebody really ordinary looking. And I've noticed that the PAs who do the most ordinary and/or unusual models are also the ones who focus really heavily on the skin. Like Saiyaness...

    And of course we all see different things. I'm not nearly so picky on eyes as some. I can sometimes see what somebody complaining about eyes looks like, but often... I just don't notice. On the other hand, I'm _extremely_ sensitive to posing. Posing is the difference between 'pretty mannequin' and 'genuine personality' to me. I'm always really _aware_ of gravity. Like, in the picture above, that tendril of hair is clinging to her cheek and it would be hanging down if that was a held pose, and moving in other ways if it's an in-motion capture.

  • Ooze3dOoze3d Posts: 10

     

     

    dreamfarmer said:

    Like, in the picture above, that tendril of hair is clinging to her cheek and it would be hanging down if that was a held pose, and moving in other ways if it's an in-motion capture.

    Now I can't unsee!! I'll fix it!

  • nonesuch00nonesuch00 Posts: 18,333

    Easiest way to get more photorealistic looking renders is to use various photo filters of NIK Collection on your renders - the eye notices the stylistic similarity of the render to past photos made that used those type of photo technologies.

  • WonderlandWonderland Posts: 7,053
    edited March 2018

    I sometimes think unfinished renders with their graininess look more real than completed renders and I often add back in grainiest in post. When people don't believe my realistic renders aren't photos, I show them this screenshot, and then they sort of get it. This is unfinished, obviously, but I think looks more realistic than the finished render (which isn't on my iPad that I'm using now.) 

    What really creeps people out though is that uncanny valley when you have realistic looking unrealistic characters LOL.

    g8 willa plus.jpg
    597 x 584 - 70K
    g8 alien marigold hair cu.png
    2473 x 3200 - 2M
    Post edited by Wonderland on
  • nonesuch00nonesuch00 Posts: 18,333
    Wonderland said:

    I sometimes think unfinished renders with their graininess look more real than completed renders and I often add back in grainiest in post. When people don't believe my realistic renders aren't photos, I show them this screenshot, and then they sort of get it. This is unfinished, obviously, but I think looks more realistic than the finished render (which isn't on my iPad that I'm using now.) 

    I agree with you there.

  • GregoriusGregorius Posts: 397
    edited March 2018

    According to some Blender users at least, a decent dynamic range is essential for photo-realism.  I can't speak for DS, but Poser uses sRGB (at least when exporting to JPG and other low-dynamic-range formats), which is a very basic color space.  Emulating the effects of a more nuanced color space in exported renders is accomplished by tone mapping, which is mysteriously available in the FireFly render engine but not in its generally more advanced successor SuperFly.  A bit of research led me to discover two ways of tone mapping.  The first was devised by the guru Bagginsbill.  His method is essentially to export the target SuperFly render in an HDR format and then plug it as a texture map into a shader that implements a tone mapping equation.  That shader is applied to a flat "billboard" prop and rendered in FireFly.  The second method also exports the render in an HDR format, but it then uses Photoshop's built-in tone mapping functionality in order to convert it it to a JPG (or other low-dynamic-range) format that imitates the contrast of the high-dynamic-range version.

    I've been experimenting with both techniques lately, and I'd be really curious to know which one you think produces better (i.e. more photo-realistic) results.  If it helps, I'm attaching a smaller combined image that should make side-by-side comparison easier.

    ORIGINAL RENDER EXPORTED AS JPG

    HDR EXPORT RE-RENDERED IN FIREFLY WITH EXPONENTIAL TONE MAPPING

    HDR EXPORT TONE MAPPED IN PHOTOSHOP AND CONVERTED TO JPG

    CompareJuanita2.jpg
    900 x 300 - 154K
    Post edited by Gregorius on
  • bluejauntebluejaunte Posts: 1,923
    Gregorius said:

    According to some Blender users at least, a decent dynamic range is essential for photo-realism.  I can't speak for DS, but Poser uses sRGB (at least when exporting to JPG and other low-dynamic-range formats), which is a very basic color space.  Emulating the effects of a more nuanced color space in exported renders is accomplished by tone mapping, which is mysteriously available in the FireFly render engine but not in its generally more advanced successor SuperFly.  A bit of research led me to discover two ways of tone mapping.  The first was devised by the guru Bagginsbill.  His method is essentially to export the target SuperFly render in an HDR format and then plug it as a texture map into a shader that implements a tone mapping equation.  That shader is applied to a flat "billboard" prop and rendered in FireFly.  The second method also exports the render in an HDR format, but it then uses Photoshop's built-in tone mapping functionality in order to convert it it to a JPG (or other low-dynamic-range) format that imitates the contrast of the high-dynamic-range version.

    I've been experimenting with both techniques lately, and I'd be really curious to know which one you think produces better (i.e. more photo-realistic) results.

    ORIGINAL RENDER EXPORTED AS JPG

    HDR EXPORT RE-RENDERED IN FIREFLY WITH EXPONENTIAL TONE MAPPING

    HDR EXPORT TONE MAPPED IN PHOTOSHOP AND CONVERTED TO JPG

    Hardly notice a difference. But I think you'd need better source material. Textures would have to be exr or somesuch and actually include all the dynamic range in the first place. Overall I'd say though dynamic range has fairly little to do with realism in the grand scheme of things. You can take a crappy jpg photo with some crappy camera and it'll still look realistic. In fact, all photos other than those HDR ones who try to mimic how our eyes see the world by combining the best parts of various exposures would be low dynamic range. Or rather, even those HDR ones are actually low dynamic range are just faking a higher dynamic range but aren't really. Until all monitors and TV's are HDR, this isn't even a thing.

    Granted, if you take a crappy photo, what you photographed was HDR. It's just the end result isn't going to be because people don't have that sorf of equipment. Maybe that was the point. The VFX industry does usually work in linear color space and preserving as much dynamic range as possible. Is that what makes a human look realistic though? Very much doubt it. Skin, hair, anatomy, shaders, lighting, pose etc. seems way more important.

  • 3Diva3Diva Posts: 11,749
    edited March 2018

    I think skin plays a huge role in realism. I've made some skin shaders and settings that are being looked at by Daz that hopefully can help with that, and hopefully will be in the store soon if all goes well. :) Here's a character with one of the skin shaders with Olympia 7's skin textures:

    Post edited by 3Diva on
  • kaotkblisskaotkbliss Posts: 2,914

    That is very close to fooling me, Diva :)

    I think the biggest give-a-way in that render is the hair.

    I bet with different hair and a photo background with lighting to match and you would have nailed it.

  • 3Diva3Diva Posts: 11,749
    kaotkbliss said:

    That is very close to fooling me, Diva :)

    I think the biggest give-a-way in that render is the hair.

    I bet with different hair and a photo background with lighting to match and you would have nailed it.

    REALLY!? Wow! Thank you for the sweet comment, kaotkbliss! That put a big smile on my face! :) I'm not sure how close to realism it is, but I think it's an improvement at least. laugh Thank you for making my day! 

  • fastbike1fastbike1 Posts: 4,078

    Different strokes for different folks, I guess. The face doesn't look realistic at all. The heavy filtering doesn't per se add to realism either unless realism is defined as a grunge photo. 

    gederix said:

    Agree the second version is far more realistic.

     

  • fastbike1fastbike1 Posts: 4,078

    @Divamakeup

    The hair in the forum image has no/little visible texture. I suspect your monitor image images the hair significantly.The skin looks great, although I still get a greenish gray skin tone around the eyes and chest. The arms look great. I'll double check my monitor calibration. 

    Divamakeup said:
    kaotkbliss said:

    That is very close to fooling me, Diva :)

    I think the biggest give-a-way in that render is the hair.

    I bet with different hair and a photo background with lighting to match and you would have nailed it.

    REALLY!? Wow! Thank you for the sweet comment, kaotkbliss! That put a big smile on my face! :) I'm not sure how close to realism it is, but I think it's an improvement at least. laugh Thank you for making my day! 

     

  • kaotkblisskaotkbliss Posts: 2,914

    I too see the greenish grey around the eyes and  alittle bit in the chest, but I see that a lot on live people (not photos but in person)

    I believe if the image was rendered over a real image and lighting to match what was in the image along with a hair change that it would be very difficult to tell if it was a rea person or not.

  • outrider42outrider42 Posts: 3,679

    I think what most Daz models get wrong is how glossy their skin is. A good many characters have these nice highly detailed gloss maps...but that is the problem. Skin is not that shiny. Sure, some people have oily skin, and some may sweat, but not universally the same way over their whole body and face like many Daz models are. Daz models have shiny ears. Why are the ear lobes as shiny as the rest of the face??? We have tiny hairs that mask much of this shine. It is this high gloss that to me signals CGI most of the time. Which speaking of...I'd sure love to see a new Peach Fuzz product to cover this. One that is fully featured with body hairs for all sections of the body. Such a product would go a long way to tackling some of these issues  Peach fuzz would also take that sharp edge off Daz models, which is another hint smething is off. No matter how high the resolution is, skin does not look that sharp.

    I think gloss maps need to focus on the areas that may be somewhat oily, like nose or forehead, rather than making a noise map that is fairly even and universal. Such maps should try to be more dynamic, and tone down the areas that don't gloss so much. The key is to get this balance right without them looking chalky.

    I know that some people like glossy skin, and it can look really cool for CGI, but its not real. Look at this stock photo. There are few skiny spots on her skin, and the ones that can be seen are extremely subtle, even though the photo is well lit. Perhaps a PA could include 2 sets of gloss maps to cover this, if they have time.

    Here is a shinier face of a woman who is really surprised to be in a Daz Studio thread. But notice this shine is not on her whole face. Her cheeks are rather matte, perhaps due to makeup. Which is common, most blush makeups do not shine. Eye makeup might shine, but not often blush.

    There are exceptions, of course, like this extremely happy man who just found out RareStone released Minto. But even he does not have the gloss factor many Daz models do.

    BTW, gazing at random stock photos can be abnormally entertaining for reasons I cannot explain.

     

Sign In or Register to comment.