IRAY Photorealism?

1131416181968

Comments

  • MasterstrokeMasterstroke Posts: 1,985
    edited March 2020

    @RashadCarter:
    Thanks for your posts, allways wellcome. Lot's to proccess for me ;-)
    And thank you for mentioning, how much I a am depended on the resources I can get from stores and freebie sections.
    E.G.:: Light or better said HDRI maps. One HDRI environment map can make a character look WOW! and another one makes the same character look flat.
    You are right, I try to avoid stuff baked into the textures. I allways mess around with purchased content. I love iSourcetextures' materials and shaders which I use a lot and still keep trying to customizing it, by removing baked shadows or resizing anatomical texture elements. One important thing is, to keep in mind, that a normal map, a bump map or displacement map should not be just a copy of a diffuse map. Those maps should represent their own properties of the skin.
    I feel quite limited in achieving photo realism. So is this an obstacle, or is it this, that keeps me going?
    It is about squeezing one more drop of realizm out of it. Going beyond (my) limits.
    As Jeff mentioned, I sometimes try to lower quality to cloak the software's or mine icapabilities, by e.g. emulating  a polaroid camera look or sixties black and white photo look or some seventies scrap book photo style.
    Thanks again for your comment, I will keep it in mind. smiley

    Post edited by Masterstroke on
  • CinusCinus Posts: 118
    edited April 2020

    @RashadCarter Thanks for your well thought-out and detailed feedback and your complements regarding the render.

    I do agree that the render is too bright and stylized, but I do not agree that it is the reason the image is not "photo real". You will find plenty of highly stylized very bright and shiny images in magazines and on the web that our eyes will have no problem believing are of actually people / things because they have all the little details we expect to see. 

    I think the "holy grail" is realistic hair. Hair is usually the give away. It's definitely easier to make hair look more real in darker scenes where we can hide the imperfections and have the brain fill in the details.

    You are correct regarding the lips, they are too smooth and a bit too shiny.

    Generally, ALL of the surfaces in your scene have far too much specular shine to appear realistic. ITs nice to see that you've taken the exaggerated specular ideal and applied it uniformly, as this does create a sense of cohesion between the foreground character and the background elements. Its what makes the image appear to me as stylized rather than realistic. Still, if realsim is the goal, you'll need to lose a LOT of the specular from most surfaces. I'll explain. 

    I don't think the issue is "specular shine" or the amount of specular on the surfaces. There's just a lot of light in the scene ;)

    Here's something you probably havent considered: Lack of an atmospheric medium also makes everything look too, well, immediate and somewhat staged. Even the overly strong shiny aspects of the surroundings and skin settings would be muted by a nice atmospheric medium to gray things out and lower the contrast, even if it was subtle. You might even be able to add a filter to soften the image and create the appearance of some atmosphere without having to re-render. Right now it is clear that the woman is living in a virtual vacuum. 

    When you have such a bright outdoor scene, adding an atmospheric volume generally does not work. I have tried it in other bright scenes and the result was not great.

    The top of the building in the background, also too shiny. The overall environment is too clean looking, and also I'm not in love with the way the DOH is operating. Is this derived entirely from Iray or did you paint some of the blurring in post?  

    I did not do any blurring in post. The f-stop is set to a very low value, so only her nose is really in focus. Only thing I did in post was to change the tone and contrast slightly and I added a very small bit of grain. 

    Thanks again for the feedback. I do not mean to sound argumentative, I just have a different opinion on some of the points you made. That said, my next attempt at photo realism will definately be a darker scene with darker hair :)

    Post edited by Cinus on
  • aaráribel caađoaaráribel caađo Posts: 686
    edited April 2020

    @Cinus, here's my couple of bits on this image:

    The first thing I noticed is that the model doesn't appear to be apart of the image. Since I know you used and HDRI as the lighting source (at least I think so), that's a bit odd. My theory is that there's something going on with this set that makes the DOF screw up—it looks to me like DOF added in Photoshop, with extreme settings more than something I've ever seen come out of a camera. Another clue that something is going wrong that has nothing to do with your model is that the metals look wrong to me. They all look a bit flat. Would you be willing to render this again with a different HDRI? hdrihaven.com has scores of free, high quality images you can use. Maybe also try making the DOF wider, to see if that cut back on the blow-outs and excessive bokeh that's making the background look off.

    As far as the model, the big thing I noticed wasn't the hair but the lips. I had this problem (see my image on page 4 of this thread), and it's unfortunatley something that tends to really stand out. The hair is actually pretty good. You could perhaps horizontally tile the textures on the strands as long as the scalp texture isn't included it to clean it up a bit, but I think the bigger issue is the lighting and the lips.

    Also, I agree with @RashadCarter that the set seems to be too white/clean, but that might be the intense light. Whatever the reason, it doesn't look photoreal to me, especially combined with the DOF artifacts.

    Post edited by aaráribel caađo on
  • CinusCinus Posts: 118
    edited April 2020

    For all you guys that don't like highly stylized overly bright and shiny renders wink:

    Link to full size image: https://www.daz3d.com/galleryimage/image/970781/karina_full.jpg

     

    Post edited by Chohole on
  • CinusCinus Posts: 118
    edited April 2020

    @aaráribelcaađo

    The first thing I noticed is that the model doesn't appear to be apart of the image. Since I know you used and HDRI as the lighting source (at least I think so), that's a bit odd. 

    The environment is "Eastern Sanctuary". It does come with an HDRI, but the environment intensity and environment map values are set fairly low. The HDRI contributes to the light, but the majority of light is from a distant light and a couple of spotlights.

    My theory is that there's something going on with this set that makes the DOF screw up—it looks to me like DOF added in Photoshop, with extreme settings more than something I've ever seen come out of a camera. 

    The exaggerated DOF is intentional and meant to highlight the model and also blur her hair slightly since only her nose is completely in focus. As mentioned in my reply to Rashad, it was not done in post processing. Professional  photographers use exaggerated DOF often.

    We all strive for great lighting in our renders. Light is what makes or breaks a render, but it's not what makes or breaks photo realism. I can take a picture of a real person and completely screw up the lighting, exposure and DOF and you would still find it photo realistic (as long as the main subject is at least somewhat in focus).

    I do think using a very brightly lit render makes it more difficult to pass off as photo real. It's much easier when the image is darker or grainy so our minds can fill in the details that are missing.

    Thanks for the feedback.

    Post edited by Cinus on
  • MasterstrokeMasterstroke Posts: 1,985
    Cinus said:
    The environment is "Eastern Sanctuary". It does come with an HDRI, but the environment intensity and environment map values are set fairly low. The HDRI contributes to the light, but the majority of light is from a distant light and a couple of spotlights.

    Could this be part of the problem?
    I think it was mentioned earlier here, that somehow having to many lights in a scene might make a render look less photoreal.
    Try using your HDRI and maybe only ONE extra light source.

  • nonesuch00nonesuch00 Posts: 18,131
    edited April 2020
    Cinus said:

    For all you guys that don't like highly stylized overly bright and shiny renders wink:

    Link to full size image: https://www.daz3d.com/galleryimage/image/970781/karina_full.jpg

     

    You have to blur to get realistic when the models have insufficient levels of realism. Also such saturated colors, if you've noticed, are usually research for tourism advertising or magazine fashion advertising and the fashion advertising is always highly photoshopped to the point you would never call it realistic.

    Post edited by Chohole on
  • Cinus said:

    The environment is "Eastern Sanctuary". It does come with an HDRI, but the environment intensity and environment map values are set fairly low. The HDRI contributes to the light, but the majority of light is from a distant light and a couple of spotlights.

    Ah! That's the big thing that made it look rendered—the model doesn't appear to have the same lighting at the setting, but the setting doesn't look like a backdrop image. It's a strong visual clue that this isn't something that came out of a camera. People taking pictures with cameras don't have access to distant lights other than the sun :) 

    Cinus said:

    The exaggerated DOF is intentional and meant to highlight the model and also blur her hair slightly since only her nose is completely in focus. As mentioned in my reply to Rashad, it was not done in post processing. Professional  photographers use exaggerated DOF often.

    I'm not trying to be critical for sake of critical, but to analyze why your image didn't strike me as photoreal (or come close enough that I had any doubts that it was a render). The disconnect of the lighting and the unnatural (or unphotographic) DOF seem to be the things that are most likely responsible. I often use very shallow DOF (it's especially good for recreating old, large format camera photos), but the way your render came out, it didn't look like any camera-produced effect I've ever seen. I don't know what went wrong—maybe it's the set is too white; maybe the software struggled with some other element of the background, perhaps it's something in the toning settings that went awry—but it just doesn't look photoreal to me. 

    I brought this into Photoshop and applied a number of different black and white LUTs and it looked much more convincing with all of them but one. A really solid render.

    I've found Daz/Iray is much better at producing black and white images that pass for photoreal, but this one really jumped out as seeing a huge improvement. I'm not sure if it's just that the skintones are too even with this skin or if there's some ohter element that's missing. Maybe it's the nose. In black and white, the lack of SSS in the nose is less apparent.  

  • SnugginsSnuggins Posts: 52

    Wow this is all very surprising, lighting-wise. I'd given up on anything but environmental lighting, and the occasional emissive surface. Point light! Who woulda thought.

  • Snuggins said:

    Wow this is all very surprising, lighting-wise. I'd given up on anything but environmental lighting, and the occasional emissive surface. Point light! Who woulda thought.

    Im my experience, environmental lighting does it make it hard to achieve photorealism... that said, I tried recently, and used PaperTiger's HDRI X-Ray Cam... it helps, see below.

     

  • @Jeff_someone, what do your renders look like if you render them out to be mostly noise free and don't give them the lo-res treatment?

  • Snuggins said:

    Wow this is all very surprising, lighting-wise. I'd given up on anything but environmental lighting, and the occasional emissive surface. Point light! Who woulda thought.

    A lot of HDRs don't have sufficient dynamic range. That could be something that impacts what people get vs. a single point light. That said, I do simple light set-ups tend to produce better results most of the time.

  • @Jeff_someone, what do your renders look like if you render them out to be mostly noise free and don't give them the lo-res treatment?

    As requested, here's the same image in original form.  The render fully converged and has no post-processing on it.  

  • CinusCinus Posts: 118

    You have to blur to get realistic when the models have insufficient levels of realism. Also such saturated colors, if you've noticed, are usually research for tourism advertising or magazine fashion advertising and the fashion advertising is always highly photoshopped to the point you would never call it realistic.

    You don't actually get more realism when you blur the image. You are just cheating the brain into filling in the details that are not there. 

    I was not trying to render an amature photography type image, the intent was a magazine fashion style image ;) 

  • CinusCinus Posts: 118

    Ah! That's the big thing that made it look rendered—the model doesn't appear to have the same lighting at the setting, but the setting doesn't look like a backdrop image. It's a strong visual clue that this isn't something that came out of a camera. People taking pictures with cameras don't have access to distant lights other than the sun :) 

    Are you sure about that? The distant light in this render is the substitute for the sun. The spot lights are filler lights. Even cheap phone cameras have flashes and photographers use filler lights all the time when you have a subject with a bright background.

    I'm not trying to be critical for sake of critical, but to analyze why your image didn't strike me as photoreal (or come close enough that I had any doubts that it was a render). The disconnect of the lighting and the unnatural (or unphotographic) DOF seem to be the things that are most likely responsible. I often use very shallow DOF (it's especially good for recreating old, large format camera photos), but the way your render came out, it didn't look like any camera-produced effect I've ever seen. I don't know what went wrong—maybe it's the set is too white; maybe the software struggled with some other element of the background, perhaps it's something in the toning settings that went awry—but it just doesn't look photoreal to me. 

    I get it, you don't like my use of DOF, but I can assure you that it's not that hard to achieve with a real camera :)

    I brought this into Photoshop and applied a number of different black and white LUTs and it looked much more convincing with all of them but one. A really solid render.

    I've found Daz/Iray is much better at producing black and white images that pass for photoreal, but this one really jumped out as seeing a huge improvement. I'm not sure if it's just that the skintones are too even with this skin or if there's some ohter element that's missing. Maybe it's the nose. In black and white, the lack of SSS in the nose is less apparent.  

    Yes, converting the image to B&W definitely makes it appear more realistic, but I don't think it's because Iray is better at producing B&W images that pass for photo real. It's because of the way our eyes and brain work. A B&W image has a lot less information in it and because we see in color a B&W image does not actually portray reality as we see it every day, so our brains are happy to fill in the details (or ignore the missing ones).

    I attached a version of the render converted to B&W (in post). It does look more real, but I consider it "cheating" :)

    Karina Portrait 6-BW.jpg
    2560 x 1440 - 2M
  • Wow, this actually more convincing—the hair is better and the skin remains strong. 

  • SoldatoSoldato Posts: 7
    edited April 2020

    Hello guys,

    I'm new to this forum and Daz Studio, I started using literally yesterday. (I've already presented myself in the right section wink)

    I've watched various YouTube videos, (there aren't many), and searched this forum for advices, but decided to ask here after seeing the amazing renders of @jeff_someone. Man, your art is amazing and is the exact results that I would like to achieve. Of course, I know I can't learn in a couple of days what you pros have learned in years, I respect you.

    Especially @jeff_someone, what you think are the key elements to obtain photorealism like yours? I read about skin textures mostly, lights and render parameters. Attached you find my first expriment, I bought Romina HD from the shop, I think she's amazing and the skin texture looks realistic too. For the hair I bought Wild Wind Hair. Used a plane with a wall texture for the background, one camera and Sun-Sky only for the Environment Mode in the Render settings. I've also messed around with Romina's skin settings to try to not look glossy or "wet". But of course I miss something, I would like to create something that can be confused with a real person in a real photo.

    The first foto is the final render (with those strange white points, what are they and why are there?) and second after a little re-touch in Photoshop.

    Hope that with that model I can create something photorealistic, because I like her very much.

    Thank you!

    P.S.: I would like to use photos instead of 3D objects for the background (for more realism), is it possible to have good results?

    1.png
    1080 x 1080 - 2M
    2.png
    1080 x 1080 - 2M
    Post edited by Soldato on
  • cpashiacpashia Posts: 1
    edited April 2020

    So it seems like the foundation of photorealism is having extremely good (custom) skin textures. What does everyone's workflow look like for texturing? I found a tutorial on Youtube from renderhub that outlined things pretty well using photography packages of high resolution refrence images and it seems to produce good results. I've started using this method and like what I'm getting out of it so far. For those of you that are getting good results does your process look anything like this? Jeff mentioned adding colors to the skin to give it more realism, does taking skin from refrence images eliminate the need to do this? How do you blend between the different texture maps (torso, legs, head, etc.)?

     

    Lots of great stuff on here everyone, this is a fantastic thread!

    Post edited by cpashia on
  • I did a test to see how a pointlight looked vs. a single softbox set up, based on a video on Youtube about single light portrait lighting. I'd be interested in seeing how if isolating parts of the render process tells us anything about getting a more photoreal look.

    In the video, Felix Kunze shoots in a fairly well-lit white room. To emulate that, I used a small .PNG filled with 90% black as an HDRI. (The first image shows how much light is created by that environmental light). Then I added a single spotlight, set to a 150 cm tall and wide disk at 150000 lumens. That's set up as Kunze suggests for the softest lighting, just in front of the model, but not pointed at her, then tilted slightly in. From the light vew port, the modl is just outside of the larger light circle. The rest of the set contains a 25 cm disk emitting 250 lumens off to the camera's left so the eyes have somethign specific to reflect. There's a backdrop with a typical photography cloudy background, and a dark gray sweep on the floor, as well as a non-glossy white plane to the models left to fill in the shadow. 

    For the pointlight, to get closer to @jeff_someone's set-up, I cut the HDRI/environmental light 25% and used a single point light about 10 cm to the left of the camera, also using 150000 lumens (which turned out to be a bit hot). 

    The skin texture is based on Mel Lin 7, with some custom additions (the scars) and a customized dual-lobe shader set-up. 

     

    Single Light Lighting Tests.jpg
    2700 x 900 - 425K
    Screen Shot 2020-04-10 at 5.22.00 PM.png
    1740 x 1742 - 520K
  • i53570ki53570k Posts: 212

    @jeff_someone - Which Paper Tiger's HDRI do you use, or recommend?  There are so many sets and each set has 20 HDRI lights.

    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

  • I did a test to see how a pointlight looked vs. a single softbox set up, based on a video on Youtube about single light portrait lighting. I'd be interested in seeing how if isolating parts of the render process tells us anything about getting a more photoreal look.

    In the video, Felix Kunze shoots in a fairly well-lit white room. To emulate that, I used a small .PNG filled with 90% black as an HDRI. (The first image shows how much light is created by that environmental light). Then I added a single spotlight, set to a 150 cm tall and wide disk at 150000 lumens. That's set up as Kunze suggests for the softest lighting, just in front of the model, but not pointed at her, then tilted slightly in. From the light vew port, the modl is just outside of the larger light circle. The rest of the set contains a 25 cm disk emitting 250 lumens off to the camera's left so the eyes have somethign specific to reflect. There's a backdrop with a typical photography cloudy background, and a dark gray sweep on the floor, as well as a non-glossy white plane to the models left to fill in the shadow. 

    For the pointlight, to get closer to @jeff_someone's set-up, I cut the HDRI/environmental light 25% and used a single point light about 10 cm to the left of the camera, also using 150000 lumens (which turned out to be a bit hot). 

    The skin texture is based on Mel Lin 7, with some custom additions (the scars) and a customized dual-lobe shader set-up. 

     

    You may want to try a spotlight instead of a point light...I've had more success with that.  Position it exactly the same way as the camera, but just pulled back a little.  Also try to position it such that it causes a reflection in the eye to remove the unrealistic reflections that HDRI lighting causes.  Also, the lacrimal a bit too red perhaps?  Lastly, may want to make the Iris darker, as it's uncommon to see such light eyes in a character of this ethnicity.  

    i53570k said:

    @jeff_someone - Which Paper Tiger's HDRI do you use, or recommend?  There are so many sets and each set has 20 HDRI lights.

    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

    Sorry when I said Paper Tiger's HDRI... I meant Paper Tiger's Precision HDRI X-Ray Camera...  it is configured to enable HDRI lighting to pass through any geometry (walls, etc) such that you can get the effects of HDRI lighting in interior scenes.  

     

  • bluejauntebluejaunte Posts: 1,902
    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

    Sorry when I said Paper Tiger's HDRI... I meant Paper Tiger's Precision HDRI X-Ray Camera...  it is configured to enable HDRI lighting to pass through any geometry (walls, etc) such that you can get the effects of HDRI lighting in interior scenes.

    Huh, how does that work? Is that just a camera setting?

  • GordigGordig Posts: 10,060
    edited April 2020

     

    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

    Sorry when I said Paper Tiger's HDRI... I meant Paper Tiger's Precision HDRI X-Ray Camera...  it is configured to enable HDRI lighting to pass through any geometry (walls, etc) such that you can get the effects of HDRI lighting in interior scenes.

    Huh, how does that work? Is that just a camera setting?

    Based on discussions of this topic in previous threads, it works by using Iray Section Nodes, if I’m remembering the name correctly.  Section planes?

    Post edited by Gordig on
  • SoldatoSoldato Posts: 7

    My latest attempts, I'm using @jeff_someone advices for the surface and rendering, still not convincing. Hair too. Any advices for realistic hair?

    I've seen bluejaunte models, they are great, maybe I'll buy one of those.

    Comments? One point light used. I really don't understand what are the datails that make this fake.

    Thanks

    1.png
    1017 x 639 - 509K
    15.png
    1080 x 1080 - 2M
    Final.png
    1080 x 1080 - 1M
  • bluejauntebluejaunte Posts: 1,902
    Gordig said:

     

    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

    Sorry when I said Paper Tiger's HDRI... I meant Paper Tiger's Precision HDRI X-Ray Camera...  it is configured to enable HDRI lighting to pass through any geometry (walls, etc) such that you can get the effects of HDRI lighting in interior scenes.

    Huh, how does that work? Is that just a camera setting?

    Based on discussions of this topic in previous threads, it works by using Iray Section Nodes, if I’m remembering the name correctly.  Section planes?

    This?

    https://www.deviantart.com/slimmckenzie/journal/Tutorial-Iray-Section-Plane-Node-beginner-661286658

    Literally never heard of it before. And I do this for a living laugh

  • Gordig said:

     

    Also, how do I learn about lighting from HDRI?  I have zero background in photography so I want to cheat using HDRI instead of learning how to set up and configure mutiple lights (also HDRI is much faster to render).  I have been playing with downloaded HDRI, some look good and some look bad, but I have no idea why.

    Sorry when I said Paper Tiger's HDRI... I meant Paper Tiger's Precision HDRI X-Ray Camera...  it is configured to enable HDRI lighting to pass through any geometry (walls, etc) such that you can get the effects of HDRI lighting in interior scenes.

    Huh, how does that work? Is that just a camera setting?

    Based on discussions of this topic in previous threads, it works by using Iray Section Nodes, if I’m remembering the name correctly.  Section planes?

    This?

    https://www.deviantart.com/slimmckenzie/journal/Tutorial-Iray-Section-Plane-Node-beginner-661286658

    Literally never heard of it before. And I do this for a living laugh

    Ha yeh that's it... I never heard of it before I saw someone mention his HDRI X-Ray cam too.  We can't all know everything... that said, you're characters are THE BEST in the store - so keep it up!!!!

     

     

  • Soldato said:

    My latest attempts, I'm using @jeff_someone advices for the surface and rendering, still not convincing. Hair too. Any advices for realistic hair?

    I've seen bluejaunte models, they are great, maybe I'll buy one of those.

    Comments? One point light used. I really don't understand what are the datails that make this fake.

    Thanks

    So a few things I'd point out (and Im not trying to be mean...just helpful):

    - The first pic is pretty good, but the pose you have her in is all twisted/deformed so no matter how realistic the texture, it's still going to look odd.  Realism required 'all' aspects to be realistic, posing included.

    - Not sure what texture you used for the eyes in the first pic... but they look off... the iris in particular -- its too saturated.  Have you tried any of bluejaunte's texture sets?  Its the best for eye realism.

    - Recommend you use fibermesh eyebrows always vice texture painted eyebrows. There are lots of good ones in the store... I prefer the ones from the Lilly character.

    - As for hair, for me, it's a limited selection... for the most part, only those from Windfield make the cut.  The Taylor Hair is #1...    

    - Also recommend you invest in Autoface Enhancer product by D. Master I think.. it'll take your face realism to the next level.

    - In general, i recommend you keep the pose a basic casual standing pose (though even finding one of those in the Store that is realistic is hard) and work on making that realistic before you start posing your characters.

    Let me know how it goes!  My latest is below... those are the Lilly eyebrows... 

     

     

  • SoldatoSoldato Posts: 7

    Thank you for your advices @jeff_someone, your render is amazing as usual!

    Eyes and eyebrows are from Romina, I don't have added anything... Got to buy those stuff, cause they look really good smiley

    I'll work on your advices, thank you very much.

     

     

     

     

  • SadRobotSadRobot Posts: 116

    Do you do posing manually, or do you have any favorite pose sets?

  • You may want to try a spotlight instead of a point light...I've had more success with that.  Position it exactly the same way as the camera, but just pulled back a little.  Also try to position it such that it causes a reflection in the eye to remove the unrealistic reflections that HDRI lighting causes.  Also, the lacrimal a bit too red perhaps?  Lastly, may want to make the Iris darker, as it's uncommon to see such light eyes in a character of this ethnicity.  

    @jeff_someone, thanks. I thought you used a point light instead of a spot, so I was trying to emulate that. Do you set the spotlight to have the default disk size (10cm x 10cm), or go bigger? It looks bigger in the latest, since the shadows on her face are soft. You're definitely right about the eyes—it's the weakest part of my texture set-up, I think (that and the neck, which is too shiny around the creases—they look off in a a number of renders where the rest of the model looks good). 

    The big thing I was trying to explore was the question of how much the front light matters to realism. I don't feel either version looks more photoreal, so I don't know if I came close to answering that question. If you're game for it, I'd love to post my scene, saved with the base G8F in place of my character, and see how your character works with the softbox light. Also, I'd love to know how you arrived at your approach. Did you want to make photos that look like dorm room selfies (or Polaroids), or did you start from just trying to make renders that passed for photos and eventually found that style works best for you?

Sign In or Register to comment.