Anthon Lockwood

no problem :)

If you have `snap` (e.g. by using ubuntu), you can install my [fork](https://github.com/anthwlock/untrunc) by using `sudo snap install --edge untrunc-anthwlock`. Otherwise the build process (of that fork) is relatively easy...

Your .cpp files seem to be different from master. Because `AVFrame *frame = avcodec_alloc_frame` is commented out in master. If you want to add functionality consider starting from [this branch](https://github.com/anthwlock/untrunc).

Feel free to send me some samples. Maybe I will be able to refine the relevant matching patterns a bit, which could improve your results a bit. > The SD...

> https://rotoglup.blogspot.com/2009/06/few-performance-tests-on-sequential.html Am I missing something, or did this flag only have an impact on Windows XP? Adding this flag to the code would also add some complexity. First because...

> Codec::was_bad_ = 1 -> skipping I think this is the problematic line. It also appeared in #83. So please try again with the newest changes. It may be important...

Since untrunc is heuristic based, sometimes invalid samples are added to the resulting file. In these cases this often goes unnoticed, since video-players usually just ignore such invalid samples. However...

Currently not. However you could write a script which iterates through your files and executes untrunc for each of them. For example in python you could use `os.listdir` or `os.walk`,...

Interestingly, the corrupted file itself contains a valid moov atom. But it looks like some junk(?) data got injected to certain places, leading to an (from the original moov unaccounted)...

I guess the filesystem of the SD card (ntfs?) did not store your files (completely) sequentially. So those sequences of junk-data could be data from other files. In that case...