Open oyvindln opened 3 years ago
Current code only looks for a standards-conforming line of eq/vsync/eq pulses. Dropons on a field is usually okay due to guessing based on previous vsync, but we get issues with samples where the vblank area is abnormal like "240p/288p" type signals.
Maybe a first fallback could be to do something based on pulses longer than some set length since it will have to be present in some form for a CRT to display it.
--fallback_vsync
option does this for vhs-decode now. Causes some sync issues on the crafters sample that at least needs to be figured out before making it standard.
Issue for both cvbs-decode and vhs-decode.