2

I am trying to make a reasonable decision in an automated logic flow if a video is interlaced, and then deinterlace it.
I am using ffmpeg and idet, and I would like some guidance on how to best use the information produced by idet.

I know interlace / telecine is a much more complicated topic than just idet, but please humor me and help with how to interpret idet. (also open to better but as simple solutions using ffmpeg CLI)

I am aware of the following resources on the topic:
Detecting interlaced video with ffmpeg
How to know if a video is interlaced or progressive?
idet.sh

In my testing I found lots of media where the field flags are not set, so using e.g. ffprobe or mediainfo alone is insufficient, so I do use ffprobe and mediainfo and idet.

The no longer maintained idet.sh script from MPV Player (if I interpret the script correctly) only uses multi frame output, and makes a determination based on interlaced being 5% or more of known frames.

When I look at synthetically generated content, it is very easy to make a decision, but when I look at in the wild content, I am not so sure if using 5% and only multi frame is a good approach.

E.g. synthetic interlaced sample:

ffmpeg -an -sn -dn -i "interlaced.mkv" -filter:v 'idet' -f null - Repeated Fields: Neither: 8991 Top: 0 Bottom: 0 Single frame detection: TFF: 8991 BFF: 0 Progressive: 0 Undetermined: 0 Multi frame detection: TFF: 8991 BFF: 0 Progressive: 0 Undetermined: 0 

E.g. synthetic progressive sample:

ffmpeg -an -sn -dn -i "progressive.mkv" -filter:v 'idet' -f null - Repeated Fields: Neither: 8992 Top: 0 Bottom: 0 Single frame detection: TFF: 0 BFF: 0 Progressive: 8992 Undetermined: 0 Multi frame detection: TFF: 0 BFF: 0 Progressive: 8992 Undetermined: 0 

With real media it is not so clear:

Repeated Fields: Neither: 76434 Top: 0 Bottom: 0 Single frame detection: TFF: 560 BFF: 6353 Progressive: 64750 Undetermined: 4771 Multi frame detection: TFF: 610 BFF: 6459 Progressive: 69231 Undetermined: 134 

When should I use single frame vs. multi frame vs. both for decision making?
What are reasonable ratios to use for decision making?
Are there other maintained sources of code/script I can refer to that use idet?

2
  • Came across this page when thinking about similar solutions, will follow in case there's any updates. Out of curiosity, why do you believe the mpv script is unmaintained? The repo itself is maintained, and I can't find any issues/comments/docs in the repo stating that script is no longer to be used or is not fit for purpose. Did you settle on a certain solution since creating this question? Thanks! Commented May 25 at 9:46
  • "This script is partially successful, but is considered a legacy / unsupported script by the MPV Team." (community.topazlabs.com/t/…) Commented Jul 3 at 17:46

0

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.