[FFmpeg-user] decomb versus deinterlace

pdr0 pdr0 at shaw.ca
Sun Apr 19 19:06:11 EEST 2020


Mark Filipak wrote
>> The result of telecine is progressive content (you started with
>> progressive
>> content) , but the output signal is interlaced. 
> 
> According to the Motion Pictures Experts Group, it's not interlaced
> because the odd/even lines are
> not separated by 1/fieldrate seconds; they are separated by 1/24 sec. 

1/field rate seconds in the case of hard telecine. That's the reason for
telecine. If you take broadcast on a 1080i station, everything is
transmitted like that. "23.976p" native is not supported. If you watch a
23.976p movie on a 1080i29.97 station, it's hard telecined. That's why it's
called 29.97i (not p) . It's the transmission format. It's "23.976p in
29.97i" 

In the case of soft, it's a unique exception - and only applies to DVD and
it depends on the hardware. The flags output a 29.97i signal. If you have a
flag reading player it's 29.97i. That's the signal sent to the TV. Your TV
IVTC's it. If you have a cadence reading player, it treats it as 23.976p
signal sent over HDMI



>> That refers to interlaced content. This does not refer to your specific
>> case. You don't have interlaced content. Do you see the distinction ?
> 
> Well, of course, but why are we (you) discussing interlaced content? 

You were talking about decomb vs. deinterlacing . That's what this thread
title says.  Logically it follows you should talk about interlace content.
In general, you only deinterlace interlaced content ... How can you NOT talk
about it ? Notice everything in my prior posts in this particular thread was
in the general sense , other examples like 16mm, camcorder... no reference
was even made to your specific case. Or if you assume your specific case,
why are you talking about deinterlacing? It needs to be referred to in order
to to contrast against IVTC and weave

Anyways.... carry on...





>> The point is, visually, when looking at individual
>> frames, combing looks the same, the mechanism is the same.
> 
> (see above).
> 
>> The underlying
>> video can be different in terms of content (it might be interlaced, it
>> might
>> be progressive) , but you can't determine that on a single frame
> 
> Sure I can. If I put 1/60th second of comb next to 1/24th second of comb,
> I think anyone would see
> that the 1/24th second frame looks worse. 

Yes, in terms of degree of combing

But in terms of what the underlying video content actually is, there is no
way you can tell from 1 frame (2 fields) if it was originally a 23.976p
video, or anything else. You have to examine the adjacent fields next to
that frame as well. If the prior and next frames had 2 different fields
each, for 4 different times represented, it's interlaced content. That's
also why it should be called "combing" , not "interlaced frame" or
"interlaced content frame" 





>> What kind of motion compensation did you have in mind?
> ... [show rest of quote]
> 
> Hahahaha... ANY motion compensation. ...Is there more than one?

I want to understand what you mean by "motion compensation." In general
terms. What do you expect to happen or want it to look like. If there is no
motion, then what; if there is big motion, then what ....
And there are other programs than ffmpeg that can do this; I'm trying to see
if it's possible in ffmpeg




>> "recurse" works if your timestamps are correct.
> 
> Carl Eugen said that, too. How could the timestamps be wrong? I was using
> the test video ("MOVE"
> "TEXT") that you gave me. I proved that, for that video (which I assumed
> was perfect), ffmpeg
> traversal of the filter complex does not occur. 

Did you look at D2 output in my earlier post in that thread?  "testD2.mkv" ? 
If you were correct, there should be no blending in D2. There is blending in
D2... 

The blend filter changed the timestamps. You blended it with n-1
(select='eq(mod(n+1\,5)\,2)')

Split each intermediate node and check the output and timestamps eg. C2, H2
as well . C2 and H2 are the "working" proper timestamp branches  (without
the n-1) . Their timestamps are complementary. D (or D2) is off and does not
fit , so interleave does not work as expected









--
Sent from: http://www.ffmpeg-archive.org/


More information about the ffmpeg-user mailing list