Windy Noise

[salvo00786]

Yes, i know the manual, but it can't see the real differences between 2 and 3. I can assume probably that 3 is better? I requested the differences to you, because probably with your big experience in videos you have tried these. It's an assumption.

[jagabo]

I can assume probably that 3 is better?

No, it depends on the particular video. A single rate deinterlacer will often start with one field of an interlaced frame, then fill in the other field by some method (simple interpolation, smarter interpolation, retaining the other field in areas where there is no motion and only interpolating the rest, etc.). In a case like this one field has intact data from the original video, the other is some kind of synthesized data. Cleaning up the video while prioritizing the intact field should give better results. That's the difference between InputType 2 and 3 -- one prioritizes the top field, the other the bottom field. So you want to use the one that works best for your source video.

[salvo00786]

And If the video is progressive, How can I know if I have to prioritize the top or the bottom field?

2 is for top field and 3 for bottom field?

What do you think about Sourcematch=3, Lossless=2 for progressive videos?

If my video in origin is bff, I can assume that if I want a second pass qtgmc I have to specify the inputtype for bottom field or my assumption is wrong?

Is Santiag more destructive to details than qtgmc or is a light antialiaser?

What can I do for moire?