Closed psstpsstpsst closed 3 years ago
If there is a way to predict start and length of those 'free' sequences, it is fixable. The better the prediction the better the resulting file.
If you want me to take a look, send me your files (healthy + broken).
Perfect, Anthon, I just did. Thanks a lot.
The main problems were:
The first one was easy to fix, the second was a little harder. There were more problems, see commit history if interested..
You are a master :-)
Thank you!
On Wed, Sep 30, 2020 at 9:20 PM Anthon Lockwood notifications@github.com wrote:
The main problems were:
- unknown atoms, scattered in the mdat atom 015d0f1 https://github.com/anthwlock/untrunc/commit/015d0f112f945d5fa2e716b3616f1c7bdf9c81f3
- false detection between avc1 and sowt. a17b173 https://github.com/anthwlock/untrunc/commit/a17b173e9fa3ea91b79c8e5142e33f478bb4d7ef
The first one was easy to fix, the second was a little harder. There were more problems, see commit history if interested..
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/anthwlock/untrunc/issues/66#issuecomment-701592584, or unsubscribe https://github.com/notifications/unsubscribe-auth/ARCKAQLWINO77SLBS3HYZALSIOAG3ANCNFSM4RTY3IHQ .
I'm having an issue with this, I'm getting:
Info: version '644978f' using ffmpeg '3.3.9' Info: reading 20151126_070833.MOV Info: parsing healthy moov atom ... Composition time offset atom found. Out of order samples possible.
Info: unknown track 'sowt' found -> fallback to dynamic stats Error: bad track: 'free'
Is it something that can be fixed or is it impossible?
Thank you very much for you work, really appreciate it.