Ok, I think I'mnot going to dive on this because I would have to ask too many questions and most likelly the thread would eventually die without me being able to finish whatever I woul start!
And at this moment I don't have the time to dedicate learning all the bits to be able to do it on my own!
But even so, I am a bit excited with this and I just tried to open a file my colleague was able to recover in his Windows machine, which was an MP4 file and mpv though out this error:So, I thought this could be a good candiate.
I open the file with hexdump and checked the first bytes (file signature) and compare it with the expecte file signature (which is 66 74 79 70 6D 70 34 32 for M4V MPEG-4 video, ftypmp42 with 4 bytes of offset) and I got:
So, in this case, can I assume that this file is probably corrupte somewhere further in the lenght of the file and cannot be recovered?
Edited;
Apparently, at least for this file, there is a section missing, called "moov". After digging a bit more I find that a playable MP4 file shows something like this, running the following command:one of the problematic files shows the following:So, now, I'm going to try to find out how to fix this, if possible, and give it a try!
And at this moment I don't have the time to dedicate learning all the bits to be able to do it on my own!
But even so, I am a bit excited with this and I just tried to open a file my colleague was able to recover in his Windows machine, which was an MP4 file and mpv though out this error:
Code:
$ mpv ./Unknown\ folder/\[000175\].mp4 [ffmpeg/demuxer] mov,mp4,m4a,3gp,3g2,mj2: moov atom not found[lavf] avformat_open_input() failed[ffmpeg/demuxer] mov,mp4,m4a,3gp,3g2,mj2: moov atom not found[lavf] avformat_open_input() failedFailed to recognize file format.I open the file with hexdump and checked the first bytes (file signature) and compare it with the expecte file signature (which is 66 74 79 70 6D 70 34 32 for M4V MPEG-4 video, ftypmp42 with 4 bytes of offset) and I got:
00000000 | 00 00 00 18 66 74 79 70 6d 70 34 32 00 00 00 00 |....ftypmp42....|Looking to my file's output and in my ignorance, I immediately thought I had a 1 byte offset there, because I only see 3 bytes of zeroes. So, I asked chatGPT if I had a misalignment there and he told me that I didn't have any misalignment and that that was actually correct for the given file type and also the byte 18 was actually the length of the set of bytes that defines the file type / format and compliance or whatever the correct terminology is.
So, in this case, can I assume that this file is probably corrupte somewhere further in the lenght of the file and cannot be recovered?
Edited;
Apparently, at least for this file, there is a section missing, called "moov". After digging a bit more I find that a playable MP4 file shows something like this, running the following command:
Code:
$ ffmpeg -v trace -i ./Unknown\ folder/\[000169\].mp4 2>&1 | grep -e 'mdat' -e 'moov'[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'mdat' parent:'root' sz: 22719721 32 22726446[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'moov' parent:'root' sz: 6701 22719753 22726446 <---------- moov section[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'mvhd' parent:'moov' sz: 108 8 6693[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'udta' parent:'moov' sz: 66 116 6693[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'meta' parent:'moov' sz: 121 182 6693[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'trak' parent:'moov' sz: 3919 303 6693[mov,mp4,m4a,3gp,3g2,mj2 @ 0x56246d0584c0] type:'trak' parent:'moov' sz: 2479 4222 6693Code:
$ ffmpeg -v trace -i ./Unknown\ folder/\[000174\].mp4 2>&1 | grep -e 'mdat' -e 'moov'[mov,mp4,m4a,3gp,3g2,mj2 @ 0x5603dc6b04c0] type:'mdat' parent:'root' sz: 818292713 32 39123456[mov,mp4,m4a,3gp,3g2,mj2 @ 0x5603dc6b04c0] type:'mdat' parent:'root' sz: 818292713 32 39123456[mov,mp4,m4a,3gp,3g2,mj2 @ 0x5603dc6b04c0] moov atom not foundStatistics: Posted by PsySc0rpi0n — 2024-02-01 12:14