Mask Tool-Brightness Curve Subtract mode, more...

[den]

...
Ref: PWP 8.0.exe 180417

Subtract-ing a BrightnessCurve mask from a previously Apply Add BrightnessCurve mask will result in no mask, ie, a 100% black mask image... ...regardless of their channel selections.

Also believe there are anomalies with the Overlap and Invert modes for subsequent BrightnessCurve masks applied to previously Apply Add BrightnessCurve masks as well.

Modes seem to work properly if the first mask is made using one of the other Mask Tools [Line, Polygon, ColorRange, Brush, etc] and the subsequest mask is a BrightnessCurve, any channel... ...however if an Undo is selected, this results in the Mask transform completely re-setting rather than Undo-ing the BrightnessCurve mask and creates a Mask image of the first mask.

Also observed that if the first mask is a BrightnessCurve, any channel, that Line, Rectangle, Oval, Polygon, Spline, Color Range, etc tools do not seem to be accessible to make a second mask.

Hopefully I am not getting confused... ...this Mask Tool and Masks have some incredible possibilities!

...den...

[jsachs]

I think it's working. If the second curve lies above the first one, subtracting leaves all black (subtract is actually computed as the original mask minus the new mask) since the new mask is whiter than the original one everywhere. If the second curve lies at least partially below the first one, then you do get a non-black mask.

In some cases you may have to click on the mask dialog to get control of the cursor back to use the other tools.

[den]

...
I was trying to re-create in PWP8 the PWP7 asymmetrical mask of non-sky HSV-V image areas shown at the top of this web page:
http://www.l7.org/IMAGEs_by_DEN/misc/PWP8_AsymMask/

...den...

Addendum: Could it be that the curve start and end Control Points need to be explicit rather than implied?