Page 1 of 2 12 LastLast
Results 1 to 10 of 17

Thread: IRAY SOFT EDGELESS MATERIALS - volume depht + 3d textures

  1. #1
    Join Date
    Aug 2016
    Posts
    23

    Exclamation IRAY SOFT EDGELESS MATERIALS - Volume Depth + 3d Textures



    HI EVERYONE ! new iray user Here ...

    CAN SOMEONE PLEASE SEND ME SOME MDL CODES FOR A Soft EDGELESS MATERIAL !

    I Want to emulate This Effect in Blender ...

    https://www.youtube.com/watch?v=IYOpKjgU1dM



    Unfortunely i canot find a Way to do this ...

    With Mixing "transparent and non transparent" Materials with Custom curve Material/bsdfs layers or fresnell Layers It just Does not work ...

    As instead of Having Transparency ( the transparent parts become BLACK .. on Edges )


    https://s23.postimg.org/8hv5mck8r/ma...shine_dawy.jpg


    THANKS SO MUCH !


    SOULS STUDIO
    __________________________________________________ ________________________

    A PERSONAL NOTE / REQUESTS TO IRAY MDL CODE DEVELOPERS

    Overall i See that Iray Have quite Some MDL code Problems with

    1) ---------------

    - Conserving Energy with "multiple Volumetric material layers" With Transparency ...
    - Having Subsurface Scatering and Transparency Efects ... Across Many Materials Layers / Bsdfs Sublayers

    This is quite a Big Trouble ...

    And as we dont have a Volumetric DEPHT Ramp Method in Iray there are alot dificulties with Emulatings Softness

    2) ---------------

    - There is Also no Way to Have Geometry Normal Aware Procedural textures Like in blender ...
    So that we use that Geometry Masks outputs as Textures to drive Shader layers MASKS ...

    Expecially Volumetric 3D Masks to Drive SubSurface SCatering And Volumetric Transparency Effects ..

    3) ---------------

    THERE IS THE 3D TEXTURES Available ...
    - But No Documentation on how to USE or Do them in Any 3d aplication ...

    4) ---------------

    There is also no Procedural real 3D Volumetric Textures Like in blender ...
    Only Uv Aware Procedurals, that dont have 3D Depth information ...

    5) ---------------

    AND we also need a BLUR Node ... to blur 2d Textures and even Shaders Results .

    In Some cases to Blur Entire Shader Layers ( maintaining Transparency ) Would be Reaaly needed ...

    6) ---------------

    - We Found out also that After Some like 10 Material Layers on top of Eachothers

    - SubSUrface Scatering / Emission and overall Volumetric Effects They STOP working ! Even With Normalize Layers ....

    8 ) ---------------

    - We Reaaly Need Edgless Fuzzy Soft / Glow Effects to Iray materials ..


    9) ----------------

    AND FINALY WE REAALY NEED! SOME SOFT FUZZY SKIN SHADERS BSDFS ... or SOME SORT OF BETTER SKIN SHADER support in iray !!

    As All base BSDFS Shaders Are For Plastic or Glass, IRAY Skins will always tend to look like Plastic as There is no Proper VOLUMETRIC SOFT SKIN SCATERING SSBSDF Base !


    And as NVIDEA is So further in Realtime SkiN Development ...

    We are Sure is Prety trivial to Apply Same realtime Shader Tecniques in MDL language ...

    We just need the bases or just ( someone at Nvidea trying )


    http://http.developer.nvidia.com/GPU...ems3_ch14.html

    https://renderman.pixar.com/view/imp...-a-skin-bssrdf

    http://graphics.ucsd.edu/papers/egsr...sr2006skin.pdf

    http://www.wikihuman.org/index.php/o...ikihuman-data/

    http://www.anderslanglands.com/alshaders/tut_skin1.html


    Vray Al-Surface Skin Shader is on github as Open Source ...

    https://github.com/ChaosGroup/vray_al_surface/releases

    Maibe Someone could Translate it to IRAY MDL Language !


    THANKS FOR LISTENING !

    ____________________
    ...

    I know that all this Requests are very dificult ...

    But At least the VOLUME DEPH RAMP to make a Volumetric Transparent Soft Edgeless Fluffy / Glow Material ...

    Might be a Thing of Joining Some MDL Ramp Codes Acting on a Inverted Y to Geometry Normal into a Volume Shader ...

    We Tryied and we could not do it anyway ...

    And its Seams so simple Just few Nodes With a Ramp in blender CYCLES ... It Should be as SImple in IRAY !

    https://s23.postimg.org/kigln2rnf/BL...DGLESS_MAT.jpg


    ---



    SO WE WOULDLIKE TO REQUEST / IF AT LEAST ! CAN SOMEONE PLEASE :

    Send us a Code to do a SIMPLE Soft Edgeless Glowing Volume material as in the BLENDER VIDEO ABOVE ..


    We are in a Deadline for a Customer here ... and this is just the last part


    THANK YOU DEEPLY SO MUCH FOR YOUR SUPPORT !


    -------------

    DINIS / SOULS STUDIO
    Last edited by SOULSSAGA; January 30th, 2017 at 18:57.

  2. #2
    Join Date
    Jan 2007
    Posts
    314

    Default

    Hi Dinis,

    in Iray, you need to describe the material in physical terms. For sub surface scattering you should use the scattering and absorption coefficients. Simple hacks (the author in the video calls it a hack himself) will usually not work, there will be side effects a simple (rasterizing/compositing style) renderer as blenders default renderer would ignore. Iray is more similar to cycles here.

    If you tell me what iray version you use and in which application, i could help you with setting sss up correctly. Emulation can not work in iray and would need to be done using compositing

    What you try to do will not work as expected for several reasons:
    -hdr: color is not limited to 0-1, but can be arbitrarily high. Even a weight of 0.01 of a diffuse bsdf will lead sometimes to full white in the resulting image and not look transparent
    -light transport needs to be symmetric since iray does not just shoot rays from the eye but also constructs paths from the light. Transmission must be the same for both directions. This means that you will see the inside of the sphere too, and depending on much how area is diffuse reflective, this inside will look pretty dark. In the best case this will expose itself as just a dark ring. It will not behave like a simple composit.
    Only because of this Iray is able to give consistent shadows and reflections.
    Click image for larger version. 

Name:	Untitled.jpg 
Views:	175 
Size:	132.9 KB 
ID:	8794
    -a material layering calculation happens on the microfacet level of the bsdf. Not on the object normal level.
    -a material featuring just a diffuse reflection with a specular transmission does not exists in the real world, so its pushing the envelop for iray and might expose additional artifacts

    in the same spirit a "VOLUME DEPH RAMP" will not work in Iray. Its only making sense for a simple "compositing style" renderer

    >There is also no Procedural real Procedural 3D Volumetric Textures Like in blender ...
    perlin_noise() from the "base" module is a procedural 3d noise texture. You can apply it in world space. You can use it in iray as a texture. Right now Iray does not allow texturing volume coefficients, thats something we are working on.

    >AND we also need a BLUR Node ... to blur 2d Textures and even Shaders Results .
    2d textures you can blur already, writing such a function is possible but for performance reasons i would advice against doing it.
    Bluring shader results again is not possible due to the physical nature of iray. This would be a job for post processing/compositing.

    >We Found out also that After Some like 10 Material Layers on top of Eachothers - SubSUrface Scatering / Emission and overall Volumetric Effects They STOP working ! Even With Normalize Layers
    I would be grateful for an example. In general there are several factors that can lead to SSS beeing inefficient and i would like to discuss this on a concrete example.

    cu
    Jan

  3. #3
    Join Date
    Jan 2007
    Posts
    314

    Default

    since you reformated your post:

    >2) ---------------
    >- There is Also no Way to Have Geometry Normal Aware Procedural textures Like in blender

    you can do geometry normal based shading in iray/MDL. If you want to do view dependent based shading though you will need to use bsdf layering or feed in the camera position as a parameter.


    >9) ----------------
    >AND FINALY WE REAALY NEED! SOME SOFT FUZZY SKIN SHADERS BSDFS ... or SOME SORT OF BETTER SKIN SHADER support in iray !!

    the effects described in the links are all covered with the sss support in Iray. The problem is that it is not easy to tweak compared to a BSSRDF that is focused on dense sss materials like skin. It is also hard to drive from scanned data. We know that and are discussing the addition of dedicated skin BSSRDF.

    >Vray Al-Surface Skin Shader is on github as Open Source
    This does not help. MDL is not a shading language. Light interaction can not be implemented in MDL but is implemented in the BRDF building blocks.

  4. #4
    Join Date
    Aug 2016
    Posts
    23

    Default Shaders mayen ...

    DEAR JAN JORDAN !

    - Thank you Deeply So much for your Awsome Efforted Reply .

    ---------------


    ( I AM VERY GLAD PERLIN IN IRAY IS 3D TEXTURE AWARE ... ) We Did Not Noticed ...

    ... NOW WE JUST NEED a - "3D DEEP BASED RAMP " Kind of Thing ....


    -------

    Maibe exactly as you have there ...




    THAT IS A VERY INTERESTING RESULT You have there ...

    Did you used RAMPS ?

    ...

    CAN YOU PLEASE SEND US THAT MDL SOURCE ?

    Do you have / Work with SUbstance Designer ?

    ...

    We WOuld Like to Investigate that Scene / And try to Improve those Results ...


    ---------------

    ( A Bigger Thing / Issue we See is that ALL EVERY VOLUMETRIC SHaders - Fall Back to THIN WALLED ...

    Once we use TRANSPARENCY / Geometry Opacity of Any kind ....

    We loose The VOLUMETRIC Efects DEEP ...


    - AND THERE SEAMS To not be Any method to Define The THIN WALLED Material THICKNESS ...

    ----------

    In A Way This Does not Allow us to COmbine OPACITY with VOLUMETRICS ...

    This is Reaaly a Big Handycap ...

    ...

    And Volumetrics They are Never FULL OPAQUE ... We Always see The Geometry RIM ... be in Dark or Withe ...

    Unless we Match They Opacity / with the HDRI Enviroment Exposure Externally ...


    ...

    ON A LAST CHANCE ... AFter Your WOUNDERFULL Helpfull Enlightning Post ...

    ----------

    I AM HOPPING HERE we can Hack SOmehow the 3D PERLIN NOISE

    to act as a TRANSPARENCY MASK to Remove the Geometry Boundaries ...

    ...

    But we Still Need the Volume DEEPness Mask Method ...




    ----------------

    ANYWAY !

    ------------------------

    Deeply Excuse-me for Teasing your For Sure Short Time , For "everyone" Support in this forums ...

    .. In our Trying to "communitary find" Some Answers and Solutions to our Crazy Shaders Dilemas ...


    ------------

    FOR BETTER WORKS Betwen us here !

    - NEXT THING : I WILL START BRINGUING SOME SHADER EXAMPLES ! Into this Thread ...
    - - So to nourish it a litle bit Productivly more than just "Talking about"


    --------------------------

    Just For you to know / Better = What we are doing , and How ...



    -» We Are developing some Very Outrageous Complex Mdl Shaders to Destribute Out there Comercially ...

    -»» Our Main Platform of Work is Nodeflex / Substance Designer / Hand Coding ...


    ----

    And From There We Are Developing Shaders for DazStudio Market / and All Other Iray / Maya / Max / Cinema4D ... platforms Hopefully Also MDL to OCTANE and VRAY ... Latter ...



    -----

    - There are 2 Main Shaders We Are developing


    » A Soft Volume Efects Shader That Go much on Scy-fy / Planetary Hallos / Space Effects ...





    That one Reaaly needs Volumetric Fusiness ...

    ------------



    » And A Quite Crazy ( Literaly Hundred ) BSDF / MAterial Layers Soft SubSurface Skin Shader ... That Took 8 months to Develop ..

    (Removed by moderator due to nudity)


    WE ARE STILL VERY FAR From the results we WISH ...




    -----

    - But We Deeply Believe that MDL LANGUAGE Have Already ( ALMOST Everything There ) ...


    - ONLY 2 Things That We lacked was the Lack of Proper VOLUMETRICS

    - And SOme SKIN Soft BLURING as the Preintegrated Techniques ...



    http://http.developer.nvidia.com/GPU...ems3_ch14.html

    It Should not be So Dificult to RECODE those techniques to IRAY ...

    Languages are so similar ...

    PHP Code:
      float4 finalSkinShader(float3 position POSITION,  
      
    float2 texCoord TEXCOORD0,  
      
    float3 normal TEXCOORD1,  
      
    // Shadow map coords for the modified translucent shadow map  
       
    float4 TSM_coord TEXCOORD2,  
      
    // Blurred irradiance textures  
       
    uniform texobj2D irrad1Tex,  
      . . .  
      
    uniform texobj2D irrad6Tex,  
      
    // RGB Gaussian weights that define skin profiles  
       
    uniform float3 gauss1w,  
      . . .  
      
    uniform float3 gauss6w,  
      
    uniform float mix// Determines pre-/post-scatter texturing  
       
    uniform texobj2D TSMTex,  
      
    uniform texobj2D rhodTex )  
    {  
      
    // The total diffuse light exiting the surface  
       
    float3 diffuseLight 0;  
      
    float4 irrad1tap f4tex2Dirrad1TextexCoord );  
      . . .  
      
    float4 irrad6tap f4tex2Dirrad6TextexCoord );  
      
    diffuseLight += gauss1w irrad1tap.xyz;  
      . . .  
      
    diffuseLight += gauss6w irrad6tap.xyz;  
      
    // Renormalize diffusion profiles to white  
       
    float3 normConst gauss1w guass2w + . . . + gauss6w;  
      
    diffuseLight /= normConst// Renormalize to white diffuse light  
       // Compute global scatter from modified TSM  
       // TSMtap = (distance to light, u, v)  
       
    float3 TSMtap f3tex2DTSMTexTSM_coord.xy TSM_coord.);  
    // Four average thicknesses through the object (in mm)  
       
    float4 thickness_mm 1.0 * -(1.0 0.2) *  
                          
    logfloat4irrad2tap.wirrad3tap.w,  
                                       
    irrad4tap.wirrad5tap.));  
    float2 stretchTap f2tex2Dstretch32TextexCoord );  
    float stretchval 0.5 * ( stretchTap.stretchTap.);  
    float4 a_values float40.4330.7531.4122.722 );  
    float4 inv_a = -1.0 / ( 2.0 a_values a_values );  
    float4 fades expthickness_mm thickness_mm inv_a );  
    float textureScale 1024.0 0.1 stretchval;  
    float blendFactor4 saturate(textureScale *  
                                  
    lengthv2f.c_texCoord.xy TSMtap.yz ) /  
                                  ( 
    a_values.6.0 ) );  
    float blendFactor5 saturate(textureScale *  
                                  
    lengthv2f.c_texCoord.xy TSMtap.yz ) /  
                                  ( 
    a_values.6.0 ) );  
    float blendFactor6 saturate(textureScale *  
                                  
    lengthv2f.c_texCoord.xy TSMtap.yz ) /  
                                  ( 
    a_values.6.0 ) );  
    diffuseLight += gauss4w normConst fades.blendFactor4 *  
                    
    f3tex2Dirrad4TexTSMtap.yz ).xyz;  
    diffuseLight += gauss5w normConst fades.blendFactor5 *  
                    
    f3tex2Dirrad5TexTSMtap.yz ).xyz;  
    diffuseLight += gauss6w normConst fades.blendFactor6 *  
                    
    f3tex2Dirrad6TexTSMtap.yz ).xyz;  
    // Determine skin color from a diffuseColor map  
    diffuseLight *= pow(f3tex2DdiffuseColorTextexCoord ), 1.0–mix);  
    // Energy conservation (optional) – rho_s and m can be painted  
       // in a texture  
       
    float finalScale – rho_s*f1tex2D(rhodTex, float2(dot(N, V), m);  
    diffuseLight *= finalScale;  
    float3 specularLight 0;  
      
    // Compute specular for each light  
       
    for (<em>each light</em>)  
        
    specularLight += lightColor[i] * lightShadow[i] *  
                    
    KS_Skin_SpecularNL[i], Vmrho_sbeckmannTex );  
      return 
    float4diffuseLight specularLight1.0 );  

    --

    FOR EXAMPLE ! In our Skin SHader ...



    We Reaaly lack some sort of Vevety Soft Skin Fuzzy BLURED VOLUMETRIC ...

    BSDF Layers To Put on Top of SKIN ...


    To make it MORE Real ... And that is So Much Essencial ...


    ( If you please have Some Codes to BLUR TEXTURES ... Shader codes ... )

    We Dont mind of Trying it ...

    As We can put These as ALTERNATIVE user Chosing Inputs ... Trough Shader Conditionals ...


    ....

    ANYWAY ! We are Working on a "Not caring for limits at all" Much less the "Phisical Plausive" ones ...

    We are ARTISTS - The Type of Results we need is More Near Artistic Fx Shading ...

    And For that We are Mixing and Taking All Available MDL Resources to the Limits ...

    ....

    THE TYPE OF SHADER LAYERING COMPLEXITY ! We Are Putting in is Something Deeply ABSURD !



    You cant Even Imagine how many "Interface troubles" we had so far ...

    Because our MDL codes just BREAK INTERFACES ! ...

    ...

    BUT WE ARE TRYING TO TAKE IRAY TO ITS LIMITS ...

    In our Soon to be Products ... At daz3d / Renderosity / Substance / Adobe / MARKETS ...

    And for that We are Taking a ( Less Tecnical / Artistical / "Put in it all" ) Special Way of Doing This ...

    ---


    And as far as we can See ( Diferent From other Engines / Less Arnold Render ) ...


    We can EVen Overlay Up to Large Hundreds Material layers ...

    AND IRAY STAYS STILL STABLE ... And Doesnt Even Takes much of Render Speed ...

    No matter how Much Complexity we put In Shaders .

    And we are taking That Advantage to its limits .

    ...


    MOST OF ALL We Hope our Far From The Common user Edge Experiences Can take IRAY as a Engine ...

    More Near to Engines as OCTANE / ARNOLD Render / CYCLES ...

    ...

    In Wich they Give us All the Power of "Physical based Realism

    But Still Allow us ShaderFx WOrkflows ... That Allow us to Hack All kind of Effects In its Shaders ...


    - It is Already Possible in those Other Mature Unbiased Engines

    ...

    And we Believe That kind of NO LIMITS - ARTIST FRIENDLY MDL Authoring ...

    ... it Should be a Priority to IRAY Team , So that More Tecnical Artist Chose IRAY as a Full Featured, No Much Compositing Needed for Cinematic FX Kind of Results Renderer ...

    SO LETS SEE HOW MUCH WE CAN HACK IRAY MDL ! To Make it So Versatil as CYCLES Renderer .


    ...

    THANKS SO MUCH FOR YOUR INTEREST IN OUR MDL ADVENTURES !


    Ruben Dinis / SOULS STUDIO


    ----

    PS: For More Productive Endeaveours Than Just Talk : Next posts i will Send here some Codes/MDL nodes SHaders ...

    THANKS SO DEEPLY MUCH !
    Last edited by Phillip Miller; February 1st, 2017 at 15:06.

  5. #5
    Join Date
    Jan 2007
    Posts
    314

    Default

    Hi Ruben,

    neither the blender example nor my material use a depth ramp but rather directional blending based on the normal. the ramp is just used to modify the directional behavior. In MDL, the closest equivalent would be measured curve layering:

    Code:
    export material softedge(
        uniform color color_0 = color(0.),
        uniform color color_1 = color(1.),
        uniform color color_2 = color(.5),
        uniform color color_3 = color(.1),
        uniform color color_4 = color(0.)
    )
     =  material(
        surface: material_surface(
            scattering: df::measured_curve_layer(
                curve_values: color[](color_0,color_1,color_2,color_3,color_4),
                layer: df::diffuse_reflection_bsdf(tint: color(.8,.5,.9)),
                base: df::specular_bsdf(tint: color(1.0), mode: df::scatter_transmit) 
            )
        ),
        ior: 1.0,
        thin_walled: false
    );
    But i strongly urge to abandon that route, hacking around physics will not get you anywhere.

    the correct way to do this is using SSS. You *can* create fluffy things with sss and they will automatically behave correct in all lighting situations. You can also stack them into each other like the 3 sss "fog" spheres in the rendering:

    Click image for larger version. 

Name:	3_spheres.jpg 
Views:	167 
Size:	230.7 KB 
ID:	8804

    Code:
    export material true_softedge(
        uniform float base_thickness = 1.0, //in m
        uniform color volume_color = color(.9,.9,.9), // avoid 1 and 0 values since those are unrealistic
        uniform float scattering_ratio = .9
        )
     =  material(
        surface: material_surface(
            scattering:  df::specular_bsdf(tint: color(1.0), mode: df::scatter_transmit) 
        ),
        volume: material_volume(
            scattering_coefficient: (base_thickness <= 0)? color(0): math::log(color(1.-scattering_ratio)) / - base_thickness,
            absorption_coefficient: (base_thickness <= 0)? color(0): math::log(volume_color) / -base_thickness
        ),
        ior: 1.0,
        thin_walled: false
    );
    A Bigger Thing / Issue we See is that ALL EVERY VOLUMETRIC SHaders - Fall Back to THIN WALLED
    Hmmm there is a misinterpretation i believe. The only mdl feature that requires "thin_walled" is double sided materials, i.e. different reflective behavior on front and backside.

    I do use Substance designer for larger work, but for forum example materials i often handwrite the MDL code in order to keep them small and simple.

    And Volumetrics They are Never FULL OPAQUE ... We Always see The Geometry RIM ... be in Dark or Withe ...

    Unless we Match They Opacity / with the HDRI Enviroment Exposure Externally ...
    I do not fully understand that comment. I would argue that in the next rendering the material is essentially opaque and does not show any rim:

    Click image for larger version. 

Name:	skin.jpg 
Views:	169 
Size:	172.8 KB 
ID:	8809

    You have not answered which software you are primarily using From the copied conversations i assume DAZ and skin seems a primary concern. The current DAZ skin material is based on some sample materials we provided. By now Iray can do better. In the DAZ skin for example there was relatively high amount of diffuse reflection to mimic small radius sss. There was also just a single specular lobe. Both can be improved. Current Iray can now support colored scattering which allows better modelling of the scattering behavior of skin. Also work like http://www.iryoku.com/stare-into-the-future illustrate that the a simple glossy lobe is not enough for skin.

    The above image is a testrendering of an updated skin material example we are developing. It features dual specular lobes, cavity fadeout towards edges as well as retweaked sss settings using a colored scattering coefficient (which Iray only supports since last fall). The above rendering has 0 diffuse reflection contribution, just specular and sss!
    The skin prototype is *work in progress*, we still want to improve "tweakability", but i attach all materials used in those renderings for you to play with.

    In general it is advisable to keep a material as simple as possible. As a general rule, a material should not have more then ~10 parameters and you should have a strong reason to have more then ~4 BSDF layers. To many parameters will just reduce the tweakability of a material. To many BSDF will just make it harder to understand why a given material does not work as expected. This is especially true with SSS.

    cu
    Jan
    Attached Files Attached Files
    Last edited by JanJordan; February 1st, 2017 at 15:44.

  6. #6
    Join Date
    Aug 2016
    Posts
    23

    Default

    DEAR JAN JORDAN !!

    - WOW !! Wow ! Wow ! MILLION THANKS FOR YOUR AMAZING SUPPORT !

    Thank you deeply so much for your Time !






    ( Here some of the Kind of Cinematic Images YOUR Planet Hallos Soft Fuzzy Shaders Will end up in ^^ )




    - Our Primary Iray Softwares are SDESIGNER / DAZ3D .


    - We Are Urging to Develop Complex Iray Materials Setups Primary For DAZ3D Market and Also All others 3D Aplications / Cinema4D / 3dsmax / maya ...


    But DazStudio Is our FIRST GOAL ...







    We Will Try To Adapt your HALLO Shader Method to Both PLANETS And SKIN ... to have a better BLURED SOFT Skin Emulation ...

    I Will bring Some Results Here Latter ( I am the CODE ARTIST And Only working with a Simple Laptop At the Moment ... )


    Only my DAZ3D Products Partner DIMENSION-z .. he is the one With the BETTER WORKSTATION for Rendering ,,,

    http://www.daz3d.com/dimension-z


    We Already have the very best UBBER SKIN SHADER At the Market ...




    But now for Next Versions we want to take the Skin Shader to a NEXT level of Realism Near ARNOLD / VRAY ...



    We Actually Started our Skins by Trying to EMULATE / Arnold Render XSI Skin Shader Nodes Trees ...

    Then Some of Blender Cycles ...


    WE ARE SO SO GLAD You Trowed at us the NVIDEA NextGen Skin Solution You Guys Are working on !

    Because we SURE going to take it to new Limits ...


    http://as-dimension-z.deviantart.com...Foot-650515897


    ( I WILL REVIEW NEXT DAYS YOUR SETUPS AND CODES ! ) ...

    I am basically Working 24/7 day And night in this ...

    Already for 8 months straight ...

    I am so exausted ..


    ---

    PS: Do you mind if we Mention Your Name in The Products Credits ?

    You Were of a EXTREME Help and we for sure would not Make it Better Without you !


    We See us Soon latter in this Tread with More news of our Development .

    THANK YOU DEEPLY SO MUCH !
    With my very Best Regards


    RUBEN ISMAEL DINIS


    3DD SOULS STUDIO

    http://www.3dd.us
    http://soulsshine.deviantart.com
    http://facebook.com/soulsshine
    Last edited by SOULSSAGA; February 1st, 2017 at 17:41.

  7. #7
    Join Date
    Jan 2007
    Posts
    314

    Default

    I noticed that the skin example i attached was actually missing part of the functionality. But as i said its wip.
    So, here is one that does what i described and a quick test using some DAZ assets. Again, no diffuse reflection, just sss and just 3 maps: specularity, bump and transmission color

    Click image for larger version. 

Name:	female_tweaked_sss2a.jpg 
Views:	173 
Size:	124.0 KB 
ID:	8840
    Attached Files Attached Files

  8. #8
    Join Date
    Aug 2016
    Posts
    23

    Thumbs up

    Dear JAN JORDAN !

    Friend Thanks you deeply so much ! For Your post !

    AND WOW WHAT A RESULT !




    THAT IS ONE OF THE MOST PERFECT IRAY "FAST SKIN" EXAMPLES ! Out there ! I am Excited !


    I Like how IN SUCH A SIMPLISTIC OPTIMIZED WAY ! You Even tought of the Thick powder Make up Side And the large Scale Features Pores and wrincles !

    ITS ALSO AMAZING HOW SUCH SHORT CODE AND SO FEW PARAMETERS DRIVE A SO PERFECT RESULT ! Congratulations !

    THANK YOU DEEPLY SO MUCH FOR YOUR AMAZING EFFORTS FOR THE COMMUNITY !

    With my very best regards !

    RUBEN.


    ....


    Going Further and i dont Resist ...
    ON CODE WISE ...

    I Sow The Changes you did were ...

    _______________

    You removed IOR

    PHP Code:
       //uniform float ior = 1.5, 

    ______

    PHP Code:
    weightdiffuse_weight*diffuse_weight
    Squaring ... the parameter to behave more natural !

    _________

    And that is Cool ! I Usually Do that ...

    Alowing Also Extra Tweeks to user Control COLOR / GAMA , OFFSET and INTENSITY ... On Each Texture Bitmap ...
    Its Kind of Usefull to give users That Degree of Texture Control, so they dont have to do to many Changes on Textures Externally ...
    ----

    PHP Code:
    uniform texture_2d dermal_map =texture_2d("white.png"),
    To ..
    uniform texture_2d skin_color_map =texture_2d("white.png"), 
    ___________________

    You Changed Specular from TINT to MONO ! ... Squaring on its weight ...

    PHP Code:
    color specular_weight_2 =    base::file_texture(texturespecular_mapmono_sourcebase::mono_maximum).tint  ;
    to 
    float specular_weight_2 
    =    1.-base::file_texture(texturespecular_mapcolor_scalecolor(.04)*specular_weight ,mono_sourcebase::mono_maximum).mono  
    ----------

    Its Kinda i dont actually like Mono SPeculars ... hehe ...

    EVERYTHING For me Should be TINTED ...
    - So to allow users to control Individually RGB degrees Tweeks ...

    And then Allow User to Control how much Tint to GrayScalling Happens ...

    ____________

    And in exchange you Added Difuse BSD Tint Instead of a Monocromatic color ...
    PHP Code:
    tintcolor(1.0),
    To ...
     
    tintdiffuse_bsdf_tint
    ____________________________


    THAT IS AWSOME COOL !

    I LEARNED ALOT WITH YOUR SHADER !

    Expecially to Give Bitmap Parameters to things like i tought was just possible with FLOATS ...

    -------------

    STILL YOU SAID that latest Iray ... We Could Put BITMAPS on VOLUMETRIC Absortion / Scatering Coeficients ...

    PHP Code:
       volume:  material_volume (
           
    scatteringdf::anisotropic_vdf(),
           
    absorption_coefficient: (subsurface_absorb_distance <= 0)? color(0): math::log(subdermal_color) / -(subsurface_absorb_distance*unit_conversion),
           
    scattering_coefficient: (subsurface_scatter_distance <= 0)? color(0): math::log(subdermal_scatter_color) / -(subsurface_scatter_distance*unit_conversion)
       ), 
    YOU KNOW SINCERELY !

    I see few options there ...

    Were That is the Most important part ...

    Skin SubSurface ScaterinG SHould be Mixed by layers ! And All layers Weights And Colors ! Driven by Bitmaps !

    And that

    _________

    PHP Code:
        uniform texture_2d diffuse_map =texture_2d("white.png"), //mainly for makeup that also blocks light. thick powder layers, thick lipstick
        
    uniform texture_2d skin_color_map =texture_2d("white.png"), //the basic color of the skin
        
    uniform texture_2d specular_map =texture_2d("white.png"), //large scale features, pores and wrinkles 
    ___

    They Should be TRANSLUCENT ! To Reveil in their Transparency the Bellow MultiScatter Layers...

    Thats how Skin Should be ...


    Maibe Allowing BLEND LAYERS ! On Volumetric Absortion Scatering Floats ?

    PHP Code:
      float Volumetric_bsdf_weight base::blend_color_layers(
          
    basecolor(0.0),
          
    layersbase::color_layer[](
              
    base::color_layer(  //color*weight
                  
    layer_colordiffuse_color,
                  
    weightdiffuse_weight*diffuse_weight//square since diffuse will enter the equation when entering and leaving the volume. should make the parameter behave more natural
                  
    modebase::color_layer_blend // User Swapable
                  
    ), 
              
    base::color_layer(  //color*weight
                  
    layer_colordiffuse_color,
                  
    weightdiffuse_weight*diffuse_weight//square since diffuse will enter the equation when entering and leaving the volume. should make the parameter behave more natural
                  
    modebase::color_layer_blend // User Swapable
                  
    ), 
              )
          ).
    tint
    ______________

    And again we hit the OVERLAYERING TRANSLUCENT SOFT MATERIALS SCATERING of this Topic Subject ....



    http://www.3dscanstore.com/image/dat...al/Renders.jpg

    http://www.3dscanstore.com/index.php...&product_id=63

    Is QUite Important ... To drive VOLUME / SCATERING / ABSORTION / TRANSLUCENCY / EMISSIONs by bitmaps ...

    As The INDUSTRY Standart for SKIN SHAders We Need to COntrol Those with PHOTO Bitmaps ...

    PHP Code:
       uniform color subdermal_scatter_color color(.95,.45,.55
       
    uniform color subdermal_color color(.95,.2,.1 ), 
    And the UNIFORM Color Floats ... as its simplisticly there ...

    Dont Allow to Conect 2D_Textures ? ...

    Or they Dont do Just the "Texture Lookup" as Color Average not as Uv Bitmaped Per pixel ?

    SO WERE ARE THOSE NEW IRAY FEATURES SHOWCASED ?

    ____


    https://s-media-cache-ak0.pinimg.com...48f61cd142.jpg


    If we do that Simplistic Coloring Scattering on VOLUMEs Part just with COLORS ... and not with Textures ...

    Results are LESS REALISTIC... WAXYEd ... The Usuall DAZ/POSER thing ...


    http://images.thinkdrawart.com/wp-co..._Tutorial6.jpg

    I think we should get it NEXT GEN there ...

    Because that Same SubsSurface Scatering Codes Are THERE and are THE SAME as the Older days of IRAY / MENTAL RAY MILA maya shaders ...

    That Way of Dealing with Scatering is JUST way too old ... ^^

    ______________


    ANYWAY !

    Thinkering on Why the UNIFORMs thing ...
    I also Sow You changed From Uniform to non Uniform Some Floats ...

    ________________

    PHP Code:
    uniform float specular_weight 1.,
    float specular_roughness_ratio .85,
    To 
    float specular_weight 
    1.,
    uniform float specular_roughness_ratio .85
    __________

    Basically you swaped the UNIFORMS ...

    --------

    SO WHAT IS KIND THE BIG DEAL WITH UNIFORM AND NON UNIFORM FLOATS ? Both on Floats / colors / textures Imputs ?

    I See overall that Putting Uniform on Floats it Gives us LESS options on how we can Adapt and conect Extra Blocks parameters to those Floats latter ... So i Usually Turn them OFF to NO UNIFORM on code ...

    But is there some kind of Phisical based Property Thing Happening ? ...

    Like when Mixing Layers With "NORMALIZE and NON NORMALIZE" ?

    ...

    LOL ... Its such a Litle thing but ...

    I Searched everywere and didnt find any Specific INFO about that ...
    THANK YOU DEEPLY SO MUCH FOR YOUR AMAZING ADVANCED DEVELOPMENT ENTUSIASM IN THIS TOPIC !


    Ruben Dinis
    3DD SOULS SHINE STUDIO

    http://3dd.us
    http://soulsshine.deviantart.com

  9. #9
    Join Date
    Aug 2016
    Posts
    23

    Default


    BY THE WAY ! I have found out some Iray LIMITATIONS ( In Iray Programmers Guide ) That Explain some of our Biggest Troubles ...


    Materials

    Iray Photoreal has the following limitations on its material model, which is based on the Material Definition Language (MDL):

    1) The number of elemental BDSFs per material are currently restricted to a maximum of 16.

    2) The number of (procedural) texture functions per material are restricted to a maximum of currently 32.

    3) Spectral dispersion resulting from a spectral index of refraction (e.g. specified by using base::abbe_number_ior) is currently only supported for df::specular_bsdf and not for df::simple_glossy_bsdf.

    4) The material's volume scattering coefficient is a color in MDL, however, Iray Photoreal only supports a scalar scattering coefficient and hence uses the luminance of that color for this.

    5) The material's volume coefficients are varying in MDL, however, Iray Photoreal only supports uniform coefficients.

    http://www.migenius.com/doc/realitys...photoreal.html

    Is There some Estimation when this Limitations will be Adressed ?

    Or any Way to WOrkaround Those ?...

    ----

    - Our Trouble in Layering more than 16 materials / or bsdfs loosing Volumetric Properties ... Comes from Limitation 1)

    - We Had Some troubles witH Spectral Index of refraction acting Strange with Some Volumetric settings ... on the Halos etc , Limitation 3)

    - No matter how much we workaround , We Canot Add anyway TEXTURES or Layers to VOLUME Scatering & Absortion ... Wille is possible in Mdl ... Limitation 4)

    - THERE is SOmething i StilL Didnt Understanded on VARYING and UNIFORM Coeficients ... But it Seams THAT bug / Limitation in IRAY Not Able to Acept VARYING Coeficients ... Afects Volumetric Possibilities in iray ...

    ...


    HO WELL ( THAT Explains Why The SKIN Shader WIP EXAMPLE is Using MORE Transmission than VOLUMETRICS to drive the Shader Efect ... )

    ... it Seams Iray Curent 2016.3 Versions ... Canot Do Proper Textured Volumetric SCatering / Absortion ...



    IS THAT 2016.3 Version the LATEST iray ? ...

    Or there is a Latest Newest Version that Adresses this Limitations ?

    .....


    PS: I tryed to add Bitmaps to This Volumetric Function ... But i just Could not ...

    PHP Code:
    scatteringdf::anisotropic_vdf(Main_Biasing___L_0_0_2),
     
    absorption_coefficient: (subsurface_absorb_distance <= 0)? color(0): math::log(subdermal_color) / -(subsurface_absorb_distance*unit_conversion),
            
    scattering_coefficient: (subsurface_scatter_distance <= 0)? color(0): math::log(subdermal_scatter_color) / -(subsurface_scatter_distance*unit_conversion)
      
        )    
    ); 

    Is Possible to Translate the SKIN Example Volumetrics Part to BITMAPED Weight Values ?!

    THANKS SO MUCH FOR THE SUPPORT !

    Best Regards .


    Ruben Dinis
    3DD SOULS SHINE STUDIO

    http://3dd.us
    http://soulsshine.deviantart.com
    Last edited by SOULSSAGA; June 25th, 2017 at 13:15.

  10. #10
    Join Date
    Aug 2016
    Posts
    23

    Lightbulb 8 needed things in iray shading

    DEAR FRIENDS !

    DEAR IRAY CODERS DEVELOPERS & PROGRAMERS !


    ANY NEWS ON DEVELOPMENTS ON THIS FRONTIER ?

    There At NVIDEA IRAY heartquarters ...
    HOW IS THE IRAY ROAD MAP RIGHT NOW ? ... 10 Months Passed since my first push up post here ...

    - SO ? Anything Changed on SHADERS and MDL Language Development Side ?

    Any Improvments on Those SHaders WIPs Tests ?

    ...

    Or are We Still Beating a Dead Horse ...
    Dealing With low End SHader Tecniques from 10 Years ago in Iray = ?

    Just Because we dont have Acess Just in Mdl code to more and better Features that Iray could gives ...

    IRAY is on the 5º Most Used Renderer On Planet Earth ... Right now ...
    So It Reaaly SHould Evolve to Get the Competition in Features ...

    This Days EVen GAME Engines ( As Unreal Engine And even Unity ) Have More SHader Features ^^

    ___
    And Speaking of UNITY ... In Few Months they Will Release OCTANE RENDERER in Unity ... For FREE .
    https://www.youtube.com/watch?v=RxoH_Cwvwe0
    https://www.youtube.com/watch?v=8ecfZF-IuSI

    That Will have for SURE a Big hit on IRAY Usage ...

    So i Deeply Believe IRAY Developers And NVIDEA Overall ... SHould Listen More To Users Needs ..
    And Push Up Their Achievments ... to Have a More Constant Updated Product ...


    ________________



    On Our Side ... We Are STill EVolving Things .. on the Layered SUbSUrface Volume Materials For SKin ...



    But we are Still not exactly were we want ...
    We were not Able to Solve ALot of Problems ... And Lack of Features ...
    That Volume Scatering in Curent Iray MDL Give us ...

    STILL After Months Of Experiments & Deep Hard Work to Have it Right ... On our Side ... We are ALmost Ready ...




    But we are Missing 2 to 8 Things in IRAY !

    That We Like to Leave for IRAY DEVELOPERS TO NOTICE ...

    8 NEEDED THINGS IN IRAY SHADING

    1) Ability to have Volume Scatering Modulated PER BSDF Driven WIth Colored Bitmaps ...
    ( Not by Material But by Bsdf ) As Doing it Per Material Just to drive 1 Bsdf is Impratical ...

    2) Procedural Geometry Ambient Oclusion ...
    ( Not by Bitmap ) As There is always Seams and takes to long ..

    3) Ability to BLUR Textures "Procedurally" ... Again to Avoid Seams ...

    4) Geometry Aware ANISOTROPY Not Uv Dependable as today is ... as we just need to dial Irregular Uv Anysotropy levels and seams Pop up instantly ... But SKIN has a Irregular Anysotropy ...
    Maibe doing this with Camera or WorldSPace Anysotropi ...
    http://manao.inria.fr/perso/~pac/microsculpt_us.php

    5) TRUE RADIANCE TRANSFER ! On Bsdfs ...
    ( Reason why all Iray Shaders Look so plastic For Loosing Scatering Energy Across Layers ) As we need the Advancements on bidirectional surface scattering distribution functions (BSSRDF) used for radiance transfer and on new diffusion profiles to model multiple scattering. It requires no pre-computation (like old mental ray misss) while the light transport is not computed in image space (like MILA Scatter) performing importance sampling of the diffusion profile directly and so allowing for complete freedom in reflections/refractions and lens shaders.
    http://www.rombo.tools/romboskin/
    LIKE MR ROMBO SKIN But in IRAY PLEASE !

    We can Layer Many Volumetric Layers to Have SoftScatering Efects ...
    But if we DONT HAVE a Proper (BSSRDF) used for radiance transfer and on new diffusion profiles to model multiple scattering.
    We will not have SOFT SKIN Emulation as Light will not Scatter Across Volume Layers

    We need to do Proper SUM OF GAUSIANS Multiple Scatering Material Across Severall BSDF / Material layers ...

    https://www.youtube.com/watch?v=rfRlagHgJ-o



    But We Canot Control Again Volumetric Deep of each layer by BITMAPS as the Indutry Standart...

    To COntrol that With directional blending based on the normal. WIth a measured curve layering method ... is kinda hard to author results that end up just like WAX Not Skin ... Because Skin is More Translucent in some places than other ... So Volume Depth In The SUm Of Gausian Scaterings it reaaly needs to be BITMAPPED ...

    We can Only do that Mixing Severall Materials ...
    But we reaaly need to do that in each bsdf on a SIngle Material for SPEED purposes ...

    So Each BSDF Layer Should have its Independent "Volume Scatering / Deep / ThickNess and Color ... " Properties to be tweeked All Bitmap driven As We can Do in ARNORLD , Cycles , Renderman ...
    Wich is Impossible on Iray ...

    6) BETTER and Proper SENSOR RESPONSE Functions in CAMERAS ... With Severall comon cameras Profiles ...
    http://www.rombo.tools/2016/04/18/se...onse-functions

    7) Better Hdr Filmic Post Management ... WIth Post Setup BY MATERIAL Id Masks Directly in Viewport !
    WIth Curves and Popular Camera Responses Templates as Blender Maxwell etc ... Check COLIMO http://www.motivacg.com/en/colimo/

    8 ) WE REAALY NEED Advanced BSDFS Like DIsney Pixar ones That Read Better in IRAY Lightning ...
    Also That Give us Better Light Wraping For Surface Enhancements
    http://manao.inria.fr/perso/~pac/

    Because of Lacking of More Advanced Efects on BSDFS ...
    No Matter WHat we do All Shaders in Iray End Up Washed and dull ...



    So We Need Acess to Things like Radiance Scaling etc ..

    Anyway ...

    We Are Going to Try to HACK things Around on our Way ...

    BUT SERIOUSLY IRAY DEVELOPERS Need to see THE LIGHT on Giving us More Advanced MODERN Shaders Possibilities ...
    That Are there in CYCLES / ARNOLD / VRAY / OCTANE ... And Seriously IRAY needs it ...

    Iray Today is Not used just for ARCHITECTURE ...
    There is Alot of CINEMATIC Rendering Addicts Out there ...
    Wanting to squeeze the best of IRAY PHOTOREALISM ...

    That Without Proper Advanced SHaders
    Its quite dificult for us ...

    __

    MAIBE As its Near Same Language ...
    Porting What ROMBO Developer ANd Others Been Doing in Mental Ray To IRAY would be a Good STart ...

    http://www.rombo.tools/


    THANKS FOR LISTENING .


    Bellow i Leave some PROPER Cg SKin Render EXample Made in Autodesk ARNOLD Renderer Version 5 :
    - Industry STandart Renderer With QUite Unlimited Amazing SHaders Possibilities ...
    - - Renderer That is Nearly WEEKLY Updated ... Not Yearly ... like NVIDEA Mental Ray


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •