Doing fullscreen colormaps in postprocessing?

[dpJudas]

Okay. Does DrawBlend do the flashes? Eruanna asked me a while back if those flashes could happen after the tonemap. The reason being that if the palette mode is active the screen flash affects which colors match in the palette, while original doom just changed the hardware palette.

[dpJudas]

In that case I suggest a two-stage approach: Don't ever check gl_renderbuffers directly but instead copy its value to another variable when a frame starts, do with the framebuffers what is necessary if it changes and use the copy for the internal checks.

That's exactly what I had in mind. PR with this coming up soon.

[dpJudas]

Pull request fixing gl_renderbuffers on/off: https://github.com/coelckers/gzdoom/pull/93

[dpJudas]

I think that DrawBlend is currently being called twice for a frame if conditions are right for it (due to a missing mBuffers->BindCurrentFB() call in RenderViewpoint).

Want me to do a pull request with the following changes: https://github.com/dpjudas/zdoom/commit ... 55f389cc3e ? It makes the tonemap no longer see the flashes, which is perfect for the palette mode.

[Graf Zahl]

I have to say that this was worth it. Finally the invulnerability colormap effect looks correct. It never did until now.

[Graf Zahl]

I think that DrawBlend is currently being called twice for a frame if conditions are right for it (due to a missing mBuffers->BindCurrentFB() call in RenderViewpoint).

Want me to do a pull request with the following changes: https://github.com/dpjudas/zdoom/commit ... 55f389cc3e ? It makes the tonemap no longer see the flashes, which is perfect for the palette mode.

I'd rather fix the bug. The call inside EndDrawScene is for when renderbuffers are off. It's just missing a '!' and the same condition needs to be used for the one in WriteSavePic.

[dpJudas]

Oh. I actually first assumed that was the case, but when I remarked the one in EndDrawScene I noticed that the flashes stopped. So I guess the correct fix is to add the !'s and then also add the mBuffers->BindCurrentFB() line. It is needed in the case where no postprocessing effects are on while multisampling is active.