What do you need from MCTD? I've ported the denoising stage (only 1-pass) with MDegrain. No pp at all, no deblock. FFT3DFilter doesn't support >8bit so I used dfttest. Code:
--------------- i ahve big hair i cant spell or type,, get over it
To me MCTD is more precise. KNLMeans blurs too much on fades. MCTD have some limiting (which makes it less efficient than SMDegrain, but preserves more details). Does KNLMeans replace MDegrain?
I too would love to see MCTD ported, I remember the results to be very comparable to commercial software back when I used it in AVS and would love to see how it stacks up against today's iterations of commercial denoisers. :thanks: If anyone is going to port it, please consider adding frequency separation as discussed .
--------------- "Bye bye" - Admin Thanks for editing my sig Admin. and thanks for removing my avatar Mods Edit by Administrator - Now thank us for putting you in timeout... next step is the door...
I noticed MCTemporalDenoise got ported to VS now. :) Very good to see! Typically, MCTemporalDenoise wasn't very stable, though, and had to run inside at least SetMTMode(3) (on AviSynth, of course). So, how has VapourSynth solved this matter?
--------------- SR740il (It is truly a sickness with no cure!)