Best format for mask images?

[najork]

Hello,

I was wondering if someone deeply in the know could give me some optimization advice.

I lean heavily on static grayscale mask clips in my comps, mostly using "50 Mask" bend mode within groups.
These are my format options, as far as I know.

JPG
PNG
DXV normal single frame (looks a little compressed to my eye)
DXV normal short loop (what I mostly use)
DXV HQ short loop
DXV HQ single frame
DXV HQ collected sequence (different masks on each frame, to be paused and selected in the deck)

I used to format these clips as jpgs, but sometimes I'd notice a little performance hits when I'd trigger them. So I started using short DXV clips, having read that those make better use of the GPU.

But besides the resource usage of loading a clip when previewing or triggering, I don't know if there is any difference while the clips are in use. Like, is a 1 frame DXV movie less demanding than a 12 frame movie? Is HQ more taxing that normal DXV (especially when simple imagery makes the files tiny either way)?

Is there an ideal format for actively mixing between many different static clips? Is the a different ideal format for a static clip that is perpetually active and unchanging?

Thanks!

Eric

[Zoltán]

I'd go for DXV normal Q, generally, HQ if you need more detail in gradients.
These will have the best triggering performance.

You'd need to have a lot of clips playing to notice a difference between NQ and HQ performance on hardware nowadays.
viewtopic.php?t=11093