Advantage of decimate mode 3 over mode 0?

[kikker]

When ripping/encoding from a DVD that contains a minimal amount of non-FILM material, it is best not to "force film" it and better to use some sort of IVTC filter, i.e. Decomb.

Given that there may be some non-FILM material, under what circumstance would it be better to use decimate(mode=0), which simply drops every 5th frame with no sort of blending to the non-FILM stuff, than decimate(mode=3), which treats the FILM material in the same way as mode 0, but uses a "smart" blending effect to minimize the jerky effects of dropping frames in the non-FILM based material?

It seems to me at least that the advent of mode=3 has rendered the mode=0 almost useless.

[Guest]

You are suggesting always using mode=3 because there may be some non-film sections. That's a reasonable approach. But with mode=3 detection of non-film sections is not perfect so there is the risk of getting some blends in film sections. If you know that you have pure FILM, then mode=0 is preferred.

[kikker]

Well, I was under the assumption that if the source were actually pure film, then "force film"-ing it would be appropriate, foregoing the need for the decomb filter, and the presence of any non-film material should necessitate the need for the decomb filter.

I suppose it really depends on the source, and it is up to the encoder to determine the better outcome. Thanks...

[Guest]

You forgot that a stream can be hard telecined. Force film works only with soft telecined streams (ones that use MPEG2 pulldown flags).