3Delight Laboratory Thread: tips, questions, experiments

134689100

Comments

  • danieldarknetdanieldarknet Posts: 21
    edited April 2015

    Hello to all someone could help me set up the texture of the teeth, renderizei using Genesis 2 Female Skin + Macro more did not get a nice result with teeth, wanted ficase more realistic, someone would know how to improve?
    I'm using Daz Studio 4.7, Engine 3Delight

    So far I got these results

    Sexy_Lips_Macro_Skin_1.0_-_PS_Edition_+_Logo_.png
    2000 x 1601 - 4M
    Post edited by danieldarknet on
  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    First, look and see which shader is being used on the interior mouth parts...that should be Template 4, under the Default Templates, in the Surfaces tab. The shader is listed up at the top.

  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    Last one for now...

    With displacement working, it's almost as if there's a specular strength map. It does a nice job of breaking up the smoothness of the specular highlight.

    simple_rtsss5.png
    800 x 800 - 671K
  • Mustakettu85Mustakettu85 Posts: 2,933
    edited April 2015

    wowie said:

    I don't know - may be a good time to make any reports or feature requests you've been hovering over.

    Thanks.

    Kettu, didn't you file a report for that? Oren Nayar being a treated as specular brick? Might be worth a shot bumping it up, if you have a link for it.

    I did. But the reports aren't public anymore (they are like any other customer support requests, filed through the same interface), so you won't see it if I link to it. Here's the link anyway: https://helpdaz.zendesk.com/requests/182152 - the 182512 is its ID, I believe.

    Mjc also filed a report on the Shader Builder bug in 4.8 beta (about nothing but surface shaders compiling). I brought that up in the beta thread, but it's Iray Iray Iray there and in the changelog. I'm starting to worry.

    If anyone gives me a link to a good tutorial on how to make new (and easily distributable) bricks for Shader Mixer, I'm willing to give it a shot.

    Soon (tm).

    Even though I dislike Shader Mixer because I can't see the source code.

    Because I don't believe the "RSL" side of Shader Mixer is going to get any updates. Cuz Iray.

    Now, would you folks please listen in. It's in everyone's best interest to make sure Shader _Builder_ works in the 4.8 when it goes public. So if we get a new release and it's still broken, then we all should make it known that we need to have it fixed.

    ASAP.

    Post edited by Mustakettu85 on
  • Mustakettu85Mustakettu85 Posts: 2,933
    edited December 1969

    Hello to all someone could help me set up the texture of the teeth, renderizei using Genesis 2 Female Skin + Macro more did not get a nice result with teeth, wanted ficase more realistic, someone would know how to improve?
    I'm using Daz Studio 4.7, Engine 3Delight

    So far I got these results


    Wowie had some nice renders in the old thread, you should ask him for the settings:
    http://www.daz3d.com/forums/discussion/21611/P420

  • Mustakettu85Mustakettu85 Posts: 2,933
    edited December 1969


    I understand how it works, but I gotta say I don't understand WHY it works. Specifically the colors in the subsurface scattering.


    First of all, with US2 there are presets for SSS, changing scatter and absorption colours.

    Surprise: Those aren't really colours.

    In Jensen's seminal paper on SSS, there are physically measured values of light scatter parameters in several substances. They are distance (in mm) per wavelength (red, green and blue).

    Here's the paper (get the PDF and look at page 5)
    https://graphics.stanford.edu/papers/bssrdf/

    In the shader mixer SSS brick (used in AoA's SSS), these distances are shown as-is.

    In the US2 interface, these distances are represented as colours for ease of use. (some arithmetic is being done on these values in the rendertime script to convert them back to mm)

    Using a specific preset and rendering with a white SSS multiplier and no maps (and diffuse channel off) should give you the surface colour that simulates the object Jensen was measuring.

    When we add maps, we add extra colour. Some of it we don't want, so we start compensating by using various multipliers (like, a cyan tint to the diffuse channel).

    ----------

    Rogerbee said:

    Still don't get it, you take out those lines, then what? Are there other references to that line which can be taken out with Find? I don't want to get lost, but I am.

    You just do what the tutorial says. Everything you need to find/kill/insert is referenced in the "code" blocks (the ones with a green line on the left and numbered lines).
    You don't really have to _think_ about anything. Save this for later =)

    ----------

    Ah, too bad about the physical sky. Well, if you ever tinker with EnvLight again, I always wondered if it is possible to cobble up a solution like the UberSoft Light kit. It has some controls for sky/horizon, plus there's also light nodes you can place around the scene. I'm thinking since you know have a point and vector with those nodes, is it even possible to use those nodes as base to compute specular light?

    There are hints in the 3Delight threads that they are doing something interesting to the trace() shadeop again, which looks like it could be used for envlights casting specular. Although I'm not sure. It's not even in the changelog yet.

    As for general skies, I need to sit down with the new Maya skylight and see what's being done to the sky colour once it's calculated. I am not giving any timeframes for this, though.

  • Mustakettu85Mustakettu85 Posts: 2,933
    edited December 1969

    mjc1016 said:
    Last one for now...

    With displacement working, it's almost as if there's a specular strength map. It does a nice job of breaking up the smoothness of the specular highlight.

    You could try more contrast in the lighting: a brighter key and also a rim light, that will bring out the interesting surface features more, I'd say.

  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    Rogerbee said:

    Still don't get it, you take out those lines, then what? Are there other references to that line which can be taken out with Find? I don't want to get lost, but I am.

    You just do what the tutorial says. Everything you need to find/kill/insert is referenced in the "code" blocks (the ones with a green line on the left and numbered lines).
    You don't really have to _think_ about anything. Save this for later =)

    Do you mean take the line out and then close the gap it left? I think I'll just copy and paste the finished bits, then I might get somewhere. You have to remember that I have the sum total of zero experience when it comes to editing.

    CHEERS!

  • mjc1016mjc1016 Posts: 15,001
    edited April 2015

    mjc1016 said:
    Last one for now...

    With displacement working, it's almost as if there's a specular strength map. It does a nice job of breaking up the smoothness of the specular highlight.

    You could try more contrast in the lighting: a brighter key and also a rim light, that will bring out the interesting surface features more, I'd say.

    That was just straight environmental lighting with added specular. Just one of the tests...and the others were before adding the displacement.

    I've been playing a lot with it.

    I've got to say Grosjean SSS has got to be the easiest there is. And damn near impossible to "over do' it...it' takes work to go all waxy...slick/oily...easy to do since I tweaked the input controls...overdone/waxy SSS very difficult. I suppose with the proper wax values, then it would be easy. And I've been trying this out in many different lighting set ups...it responds very well to light...much more 'naturally' than the other implementations in DS.

    Also found out that this particular shader doesn't like geometry that hasn't been UV mapped. I had an item I was uing for testing, that no matter what I did, texture/no texture map, 'regular', pointcloud, full rthider, area, spot, environmental (any env light shader) or anything, with this shader applied, it would render black. Swap out shaders...it would render with color/texture. So I looked at the UV mapping...non-existent. I unwrapped it in Blender and repeated...worked perfectly.

    I should probably drop back in at the 3DL forums and give pberto and Keith a big thanks for this one...it wasn't all that hard to get it working in DS, which is very, very nice.

    Post edited by mjc1016 on
  • Mustakettu85Mustakettu85 Posts: 2,933
    edited December 1969

    Rogerbee said:

    Do you mean take the line out and then close the gap it left? I think I'll just copy and paste the finished bits, then I might get somewhere. You have to remember that I have the sum total of zero experience when it comes to editing.


    Yes. If you read me telling you "kill this", just kill the indicated one and clean up the tracks. Basically I think there should be "before" and "after" versions in the code blocks, are there none?

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


    mjc1016 said:

    That was just straight environmental lighting with added specular.

    I found out HDR intensity isn't a constant between various maps. Some have to be driven up to 11.

    I've got to say Grosjean SSS has got to be the easiest there is. And damn near impossible to "over do' it.

    Also found out that this particular shader doesn't like geometry that hasn't been UV mapped.

    There's a reason why the DNA guys bothered to code the model in, right? =)

    I wonder if it's RT SSS not liking the lack of UV. I haven't tested with non-UVmapped meshes, I think.

  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    Gotcha! Now I know what you're on about!

    CHEERS!

  • wowiewowie Posts: 2,029
    edited December 1969


    I did. But the reports aren't public anymore (they are like any other customer support requests, filed through the same interface), so you won't see it if I link to it. Here's the link anyway: https://helpdaz.zendesk.com/requests/182152 - the 182512 is its ID, I believe.

    Well, I filed a report. just in case.

  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    I filed one to get US2 recompiled and I'm hoping it's on someone's to do list.

    CHEERS!

  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    There are two ways to 'fix' US2, while waiting for them to do it...

    Find the sdl file for it and run the shaderdl recompile option on it. shaderdl.exe -recompile-sdl name_of_sdl.sdl

    
      -o               :  Override default output shader name
      -d               :  Specifies an output directory for compiled shader(s)
      --strict               :  Strict SL syntax
      --int                  :  Generate `.sdl' file as byte-code (default)
      --embed-source         :  Embeds source code in `.sdl' file. Such a file
                                may be used as a source for future recompilation
      --recompile-sdl        :  Use this option to update a sdl file to a newer version
      --messagelog           :  Send warnings and errors to .log file
      -O0                    :  No optimizations
      -O1                    :  Small optimizations
      -O2 or -O              :  Agressive optimizations (default)
      -O3                    :  Very agressive optimisations
      -w or -w0              :  No warnings
      -w1                    :  Log important warnings only (default)
      -w2                    :  Log all warnings
      --allerrors            :  Log all errors, disabling error and warning filter
      --no-dso               :  Do not search into DSO directories for undefined functions
      -I                :  Add directory for searching include files
      -D               :  Define a preprocessor symbol
      -E                     :  Run only the preprocessor
      -q                     :  Don't print out the source file being compiled
      -s                     :  Overrides the source name in error messages. Useful with standard in
      --no-array-check       :  Don't generate code that checks array bounds at run-time
      --[dont-]keep-cpp-file :  [Don't] keep file generated by the C++ preprocessor
                               (NOT kept by default)
      --[dont-]use-shadeops  :  Enable [disable] use of shadeops (enable by default)
      --searchpath-debug     :  Displays a trace of where shadeops are searched
      --stdin                :  Reads shader code from standard in
      --stdout               :  Outputs target code to standard output
      -v or --version        :  Print compiler's version number
      -h or --help           :  Print this help text
    

    The other way is to edit the header of the sdl in a text editor.

    I don't hold much hope for an update recompile any time soon, because it is just a 'performance' warning and not any kind of error that will prevent successful use of it.

  • Takeo.KenseiTakeo.Kensei Posts: 1,303
    edited December 1969

    Rogerbee said:

    I filed one to get US2 recompiled and I'm hoping it's on someone's to do list.

    CHEERS!

    I posted a script to recompile all the DS Shaders once http://www.daz3d.com/forums/discussion/23867/

  • RogerbeeRogerbee Posts: 4,460
    edited April 2015

    Rogerbee said:

    I filed one to get US2 recompiled and I'm hoping it's on someone's to do list.

    CHEERS!

    I posted a script to recompile all the DS Shaders once http://www.daz3d.com/forums/discussion/23867/


    Yep, that did it! Thank you so much for that. I daresay it'll want running again when 4.8 comes out. You need to get that thread stickied!

    CHEERS!

    Post edited by Rogerbee on
  • NoName99NoName99 Posts: 322
    edited April 2015

    Now, something I just don't understand...
    People will spend days selecting the perfect brush, paints, prepping the canvas and think nothing of it...but when faced with 'tweaking' a shader preset or adjusting materials they'll go off the deep end and complain about how it is 'not artistic' to do that stuff. Isn't understanding the tools, their limitations and how to use them part of all art? Isn't knowing how to get the most out them what separates the good from the not so good? So what's the difference between knowing 10 different shades of crimson and the IOR of plain everyday glass and fine lead crystal?

    Wow, that was really well said, and practically poetic.

    I was having a spirited debate with a fine arts teacher at UCLA who has the "Computers are not art" mentality. What's amazing is that this professor is not an artist himself(which is fine) but he loves art, particularly oil paintings. I don't understand it either.

    Post edited by NoName99 on
  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    I wonder, if Da Vinci were alive today, would the Mona Lisa be a 3D character on an OLED screen!?

  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    Rogerbee said:

    I wonder, if Da Vinci were alive today, would the Mona Lisa be a 3D character on an OLED screen!?

    No...because he would have invented an amazing alternative to the OLED screen to put it on.

  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    Probably true, perhaps she'd be a hologram.

    CHEERS!

  • NoName99NoName99 Posts: 322
    edited December 1969

    Hey everyone,

    Does Kettus EnviLight 2 IBL method work for animation?

    I downloaded the .pdf from sharecg this morning but haven't gotten a chance to try it yet.

    I've been making solid progress with UE2, but EnviLight looks pretty solid.

  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    dinopt said:
    Hey everyone,

    Does Kettus EnviLight 2 IBL method work for animation?

    I downloaded the .pdf from sharecg this morning but haven't gotten a chance to try it yet.

    I've been making solid progress with UE2, but EnviLight looks pretty solid.

    Don't see why it wouldn't, it's just a light like UE2. Envlight2 tends to be 'faster' than UE2...different sample algorithms and such.

  • Mustakettu85Mustakettu85 Posts: 2,933
    edited December 1969

    dinopt said:
    Hey everyone,

    Does Kettus EnviLight 2 IBL method work for animation?

    Yes it should. The tutorial is for making the shader work in DS, and then you just use it as any other light shader.

    Unlike UE2, though, Envlight2 will only do IBL - no colour-bleeding. But it's very good at IBL, particularly in getting nice shadows out of the map.

    The rotated coordinate system of DS might pose a challenge, though, if you want to align a specular light to the map. Basically, I reckon that if you set up your scene with the Z axis pointing "up" (rotate everything manually), it should also help.

  • ZarconDeeGrissomZarconDeeGrissom Posts: 5,412
    edited April 2015

    I'm noticing an odd behavior, tho I remember having it in the past with earlier versions of Daz Studio (4.6 and 4.7, I don't know about before that)

    This is a simple 'Displacement' map (attached) set to 100%, +/- 0.25 min/max. The rest of the tiles for the floor is the 256x256 pixel maps here.
    http://www.daz3d.com/forums/viewreply/809649/
    (floor settings excluding the 'Displacement' map is third post down that page)

    In normal render mode, it works fine. In Progressive mode, the floor tiles appear to be torn or ripped or something. These two renders are using exactly the same settings, except the Progressive On/Off setting.

    Switching 'Smoothing' off or on in the surface tab, did nothing to effect this.

    Is that a Daz Studio interface glitch, or something in 3delight?

    NewFloor_RenderSettings_001.png
    523 x 972 - 37K
    NewFloor_ProgressiveRender_Render_2.jpg
    1600 x 900 - 924K
    NewFloor_FullRender_Render_1.jpg
    1600 x 900 - 903K
    Tile256_001001Bump1_002.png
    256 x 256 - 798B
    Post edited by ZarconDeeGrissom on
  • mjc1016mjc1016 Posts: 15,001
    edited December 1969

    No, they aren't the same...

    Progressive rendering is 'locked' to the Box filter method. I don't know if it automatically flips the filter width, but your setting of 8 for the Sinc filter is way too high...and it's outrageously high for the box filter. The recommended width for Sinc is 4 to 6...higher than 6 and you start degrading quality...like introducing 'ringing' artifacts. For Box...filter width should be 1 or 2.

    Pixel samples...no need to go higher than defaults, unless you have DoF enabled.

    Now, on to what could be causing your displacement problems...the first is the Box filter with an outrageous width of 8.

    Next, you may not think it matters, but Bucket order does...when dealing with displacement, but not really anything else. Very fine, long distance items..like displaced grass and fur, can suffer from severe clipping if the buckets cut across the 'axis of growth'. The other area...long, straight lines...they can suffer 'clipping', offsets, etc. when a bucket cuts across them.

    Third...bucket size. Whole buckets are less likely to suffer from clipping than partial ones.

    Let's say you have a height of a certain size (900 pixels). You want your bucket size to divide into that evenly...whole buckets....if you are using a horizontal bucket order. You use the width figure for a vertical bucket order...and any other order...the size needs to divide into BOTH dimensions evenly.

    Adding all three items together, I'm surprised that there isn't MORE of that 'tearing'. If you look closely, it follows a roughly spiral pattern...

  • ZarconDeeGrissomZarconDeeGrissom Posts: 5,412
    edited April 2015

    mjc1016 said:
    No, they aren't the same...

    Progressive rendering is 'locked' to the Box filter method. I don't know if it automatically flips the filter width, but your setting of 8 for the Sinc filter is way too high...and it's outrageously high for the box filter. The recommended width for Sinc is 4 to 6...higher than 6 and you start degrading quality...like introducing 'ringing' artifacts. For Box...filter width should be 1 or 2.

    Pixel samples...no need to go higher than defaults, unless you have DoF enabled.

    Now, on to what could be causing your displacement problems...the first is the Box filter with an outrageous width of 8.

    Next, you may not think it matters, but Bucket order does...when dealing with displacement, but not really anything else. Very fine, long distance items..like displaced grass and fur, can suffer from severe clipping if the buckets cut across the 'axis of growth'. The other area...long, straight lines...they can suffer 'clipping', offsets, etc. when a bucket cuts across them.

    Third...bucket size. Whole buckets are less likely to suffer from clipping than partial ones.

    Let's say you have a height of a certain size (900 pixels). You want your bucket size to divide into that evenly...whole buckets....if you are using a horizontal bucket order. You use the width figure for a vertical bucket order...and any other order...the size needs to divide into BOTH dimensions evenly.

    Adding all three items together, I'm surprised that there isn't MORE of that 'tearing'. If you look closely, it follows a roughly spiral pattern...

    I think you lost me in translation when I turn on Progressive render mode. I chose that screen-cap, as it showed all the settings, including ones that disappear when Progressive is on.

    In any case, trying to translate your description to what's on my screen. the LCD of 1600 and 900 is 100. That is slow as all hell to get any results going!!!! 24 was a good balance between quick 'First Pixels' for setting stuff up, and giving the CPU enough work to keep it at full load.

    Now for that filter size, as I think you suggested regarding the box filter pixel samples, well that kind of goes away with Progressive render???

    Now why are you suggesting I roll back the Anti-aliasing (Pixel Samples) from 8x8 to 4x4??? The attached render looks worse, because of the edges of the angles are all jagged.

    oh, and the ripped floor is still there, lol.
    (EDIT)
    Using the attached settings, I've just done 'Circle, and Zigzag, with the same exact results.

    ZdgBox2dot2_002_AAfilter4x4_Bucket100_ZigZag_Render_8.jpg
    1600 x 900 - 1M
    ZdgBox2dot2_002_AAfilter4x4_Bucket100_Circle_Render_7.jpg
    1600 x 900 - 1M
    ZdgBox2dot2_002_AAfilter4x4_Bucket100_settings001.png
    511 x 860 - 30K
    ZdgBox2dot2_002_Filter4x4_001_Render_5.jpg
    1600 x 900 - 1M
    Post edited by ZarconDeeGrissom on
  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    Progressive isn't tied to 'box', I have Progressive on and my filtering is Catmull Rom and can be adjusted to whatever I choose.

    CHEERS!

  • ZarconDeeGrissomZarconDeeGrissom Posts: 5,412
    edited December 1969

    Rogerbee said:

    Progressive isn't tied to 'box', I have Progressive on and my filtering is Catmull Rom and can be adjusted to whatever I choose.

    CHEERS!

    For me it is in 4.6 and 4.7.
    with 4.8 I have the settings available as shown in my former post.
    Clipboard16_lbl1.png
    356 x 349 - 9K
  • RogerbeeRogerbee Posts: 4,460
    edited December 1969

    I have had that come up a couple of times, but, not recently and I am using 4.7. When did you last update the drivers for your graphics card?

    CHEERS!

  • ZarconDeeGrissomZarconDeeGrissom Posts: 5,412
    edited April 2015

    Rogerbee said:

    I have had that come up a couple of times, but, not recently and I am using 4.7. When did you last update the drivers for your graphics card?

    CHEERS!

    ah, last week for some stupid game driver thing, kind of silly as I didn't have that game, lol.

    The bigger question is, looking at my render settings with Progressive render on, (see screen-cap above), how do I set that filter??? That 'Bucket Order' (Spiral/Circle/Zigzag/Vertical/Horizontal) setting, is just for the order that the thing renders the image in.

    Post edited by ZarconDeeGrissom on
Sign In or Register to comment.