[FFmpeg-user] Packet corrupt on cut, notice on concat

[Paul B Mahol]

On Mon, Feb 19, 2024 at 6:47 PM Mark Filipak <markfilipak.imdb at gmail.com>
wrote:

> On 19/02/2024 01.16, Greg Oliver wrote:
> > I usually do not chime in on your emails where people flame you for
> writing
> > a book or being a troll because I usually learn something from the
> > discussions.
>
> Hi Greg. I'm not writing a book. That's something someone made up.
>
> > Lately you have not posted one example clip or logs from anything you
> have
> > asked questions about.
>
> That's because there's no errors in my logs. What I'm doing is correct.
> What I have been posting is
> framecrc outputs that prove what I say.
>
> SUMMARY:
> I'm cutting solely on packets that contain I-frames. I'm joining packets
> that contain I-frames on
> both sides of the join. There's no decoding, no encoding. There's only
> remuxing. I'm well aware of
> DTS-order versus PTS-order.
>
> I began with
> ffmpeg -ss <starttime> -to <endtime> -i this.m2ts
> of course. It cut at the right place but the output included 52 audio
> frames and a B-frame that
> should have been cut. The join was not correct of course and the MPV
> player behaved as one would
> expect from a bad join.
>
> I tried
> ffmpeg -i this.m2ts -ss <starttime> -to <endtime>
> but that cut at the wrong PTSs -- hundreds of packets early. There were
> other issues, too. No one
> could explain that behavior.
>
> I asked if there is a way to discard specific packets based on PTSs. Gyan
> suggested I try
> ffmpeg -i this.m2ts -bsf noise
> I used '-bsf noise' to do _all_ the cutting and that appeared to work
> beautifully. The outputs
> appeared to be perfect in every way.
>
> It was when I joined (concatenated) those segments what I discovered
> another problem: FFmpeg had
> corrupted the final B-frame just before the ending I-frame. When I 'moved'
> the cuts and join around
> in time, the 'corruption' followed the moves, always the final B-frame
> just before the ending I-frame.
>
>

Perhaps this is bug how flush in code is handled after each reached EOF?
Might be bug somewhere. Best to create minimal reproducible usecase with
all required files to reproduce it, upload it somewhere and link it on
ffmpeg trac site.


> The video is 24/1.001fps, so ticks-per-frame is 3753.75. Because PTSs are
> integers, 3753.75 tick
> intervals must be turned into integer PTSs. That was when I discovered
> that FFmpeg is not truncating
> correctly, so PTSs are not correct.
>
> > You're being vague and Gyan just gave you something
> > to try and you refused to try with just more questions.
>
> Gyan suggested I add '-mpegts_flags +initial_discontinuity'. I tried it.
> ffmpeg -copyts -i "concat:this1.m2ts|this2.m2ts" -map 0 -mpegts_flags
> +initial_discontinuity -c copy
> -dn -muxdelay 0 "that.m2ts"
> The result:
> "Packet corrupt (stream = 0, dts = 504201210)" and the same glitch in MPV.
> Do I know whether the
> corruption is from the cutting or the joining? Of course not.
>
> > Sorry, but I find this list valuable and you are starting to put a bad
> > taste in my mouth.
>
> I'm so sorry about that. I don't know how to solve problems any other way
> than to try, try, try.
>
> > And Reindl and Paul - I don't need any smart ass
> > comments.  Reindal, I think that we are almost on all the same lists and
> > everyone hates you, and Paul, you're ego would be bad in certain circles
> of
> > technology people.
>
> I think both Paul and Harold are trying to be helpful in their own ways.
> I'm a behaviorist, so I
> don't 'talk' about motives at all.
>
> I think there are many bugs -- that's expected in an application that
> evolved as a collection of
> use-cases. I think I'm hitting them. I need help from a patient and
> persistent person who can read
> the source code.
>
> I have a few best friends on ffmpeg-user but they don't participate much.
> I have a sneaking
> suspicion that Paul just might become my best-best friend on ffmpeg-user,
> but don't tell him.
>
> I first got involved in video at Atari Advanced Engineering in the early
> 1980s where I redesigned
> its 2600 game system to produce NTSC/PAL/SECAM signals & timing that was
> 100%
> NTSC-/PAL-/SECAM-complaint. I have designed hundreds of hardware state
> machines for circuits, chips,
> and systems that I designed (and often invented) in Silicon Valley in
> California during the next 25
> years. They are/were used in games, test equipment, networking, NATO
> fighters and tanks, and various
> other design projects. Maybe half of what I designed was programmable. I
> program mostly in assembly
> language though I have written device drivers and test programs in about
> 10 other languages. In
> Silicon Valley, almost all my friends were codesmiths. :-)
>
> --Mark.
>
>
>
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
> To unsubscribe, visit link above, or email
> ffmpeg-user-request at ffmpeg.org with subject "unsubscribe".
>