Hi, I noticed that the last frame of my mp3s, when decoded using ffmpeg's mp3 decoder, always contains a bunch (several hundreds) of zero-bytes. The mp3 decoder (which I'm using as a testcase here) indeed doesn't appear to handle "partial frames" (assuming any such a thing exists in the mp3 specs in the first place (?)), since mp_decode_layer3() just returns nb_granules*18 as the mount of data decoded, with nb being either 1 or 2, and then mp_decode_frame() returns that number *32, which is 576 or 1152 (i.e. the max. size of a mp3 frame). I don't mind diving into this, but since I'm not very good at mp3 (or any audio format for that matter), is anyone good at this and willing to do it instead? I noticed the /* XXX: should compute exact size */ in mp_decode_layer3(), I'm assuming it refers to this, but the relation between scale factors, exponents (calculated from scale factors??) and the final audio samples still isn't 100% clear to me, so I wouldn't know what needs to be done to know how many audio samples should remain "empty". Any pointers appreciated. :-). Thanks, Ronald
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4