Poser's "Color Math" mystery for me


  • Poser Ambassadors

    As the blush map is a mask wouldn't a Blender node be more appropriate than Colour Math?



  • Hmm... texture strength, perhaps? These are 3 plain bitmaps, one with R=200/G=200 (=yellow), the other with B=200, the other with R=200; the Color Add seems to be doing its stuff correctly:

    0_1530461147172_t11.png

    But then if I mess with Texture Strength in one of the textures, that adds an unexpected white color, which I can then tint to a red for areas that the 3 textures are full black (R=G=B=0):

    0_1530461234625_t12.png

    So I'd suspect that a Texture Strength < 1 may be messing with your color add and adding a bias to the color?



  • @fbs7 yes, definitely another possibility, and certainly will cause a larger variation of dark colours that gamma correction would. It would definitely help to see @Y-Phil 's image node expanded and the image selection dialog, showing the selected Gamma, for completeness.



  • Thanks everybody for your suggestions.
    I've tried to set the blush bitmap's gamma factor back to 1.0 but it hasn't changed anything: same weird effect.
    And in fact, I've discovered that it was because one of the RGB component values was 0. The green in this case...
    Putting a 1.0 as green value was the solution.

    Examples with the Red component:
    With a 0 as value:

    0_1530561759842_2018-07-02_22h02_14.png

    with 1 as value:

    0_1530561796838_2018-07-02_22h01_48.png

    Once again: thank you...
    0_1530561833850_aww.gif



  • @Y-Phil Where can one download Redviper's blush mask? I can't seem to find it, but I may be out of date with his hang-outs.



  • This post is deleted!


  • @englishbob said in Poser's "Color Math" mystery for me:

    @Y-Phil Where can one download Redviper's blush mask? I can't seem to find it, but I may be out of date with his hang-outs.

    Basically, it was for the Far East Girl collection, here
    And the blush mask is in his Easy Shader3
    0_1530608033973_aww.gif



  • @Y-Phil Thanks! That explains why searching for 'blush map' didn't work. I must have passed over those downloads at the time.

    I have his original Easy Shader, which incidentally solves another search I had, which was for tan lines for V4. So double thank you!


  • Poser Ambassadors

    This is a new bug.

    I am able to confirm this behavior in my own system. It was a shock. It's a kind of bug I've never seen before.

    If one of the components of the color chip is set to 0, then whatever is plugged in to that chip is ignored. This happens even with clouds node, tile, brick, etc. - does not have to be an image map. The image map, therefore, has nothing to do with the problem.


  • Poser Ambassadors

    Oh .. my .. God. It looks like the new developers have no idea what they're doing.

    Not only do we have situations where the color chip is used and the plugged in pattern ignored, we also have situation where the plugged in pattern is used and the color chip is ignored.

    I don't know what to say. This demonstrates a total lack of understanding of basic programming, as well as a complete abandonment of QA regression testing.

    0_1530614679747_16be98a4-f302-4d7a-81c2-33409bdf5690-image.png



  • @bagginsbill would that kind of treatment (admittedly faulty) perhaps stem from trying to handle existing nodes the way the SuperFly Cycles nodes treat in-node values being subsequently ignored (and not displayed) when a node is plugged into them?



  • @englishbob said in Poser's "Color Math" mystery for me:

    @Y-Phil Thanks! That explains why searching for 'blush map' didn't work. I must have passed over those downloads at the time.

    I have his original Easy Shader, which incidentally solves another search I had, which was for tan lines for V4. So double thank you!

    You're welcome
    0_1530617236852_aww.gif



  • @bagginsbill said in Poser's "Color Math" mystery for me:

    Oh .. my .. God. It looks like the new developers have no idea what they're doing.
    ...

    Before writing such insulting nonsense you should better make some tests. I tested this issue with Poser 11 SR2 and its already there.


  • Poser Ambassadors

    @nagra_00_ said in Poser's "Color Math" mystery for me:

    @bagginsbill said in Poser's "Color Math" mystery for me:

    Oh .. my .. God. It looks like the new developers have no idea what they're doing.
    ...

    Before writing such insulting nonsense you should better make some tests. I tested this issue with Poser 11 SR2 and its already there.

    It's not nonsense and your response is only adding to the confusion. I am 5 releases PAST sr2. The bug was not FIXED in SR2, it was ADDED after SR2. Not sure which, but probably the 11.1 release (you are on 11.0.2)



  • It's not nonsense and your response is only adding to the confusion. I am 5 releases PAST sr2. The bug was not FIXED in SR2, it was ADDED after SR2. Not sure which, but probably the 11.1 release (you are on 11.0.2)

    I tested it with SR2, SR7 and 11.1 all have the same issue.


  • Poser Ambassadors

    So you think the old developers did it? If so then the old developers didn't know what they're doing. Either way it's a pretty egregious mistake that should have been caught in QA.


  • Poser Ambassadors

    Workstation Urania is Win7Pro-64 and P11Pro version is 11.0.8.34338.

    The bug is present in build 34338 (SR8).

    Interestingly, with the color saturation at 240, the preview engine displays correctly, though Superfly renders with red all over. I used a B&W mask w gamma=1.

    As long as the saturation is <240, it works correctly. With saturation = 238, it's OK. I couldn't check saturation=239 because Poser keeps changing the saturation to 240 if I enter 239. (Enter 239, click OK, re-open color picker, and the saturation shows as 240).



  • SR3 (32055) - bug exists in Firefly Color Math node, not exists in Cycles Mix node. (Well, now I know how I missed it before...)



  • Seems like the bug exists in SF only! i.e. if FF engine is active, both preview and render work as should.



  • And this bug exists since 11.0.0.31047 at least.