Hello. I did a quick search for FLAC and didn't see anything relevant, so here goes...
I am trying out the CD-ripping/format-converting software Ze Converter, which uses BASS libraries. I noticed that the FLACs it produces are corrupt; something is wrong with the final frame. I am going to report the problem to KastorSoft, but I noticed something really strange when trying to diagnose this.
They are using BASSFLAC 2.4.1. The documentation for BASSFLAC says that it uses libFLAC 1.2.1. However, foobar2000 says the corrupt FLACs were created with libFLAC 1.1.2. I don't know if this is the problem causing the corruption. Regardless, it's a mystery how they ended up with the wrong libFLAC. Any ideas?
The screenshot below is showing the number of samples in the file as a multiple of 4608, a standard FLAC block size, rather than as a multiple of 588, the audio CD sector size. For example the track in the screenshot is supposed to be 11946396 samples (that's how long it is on the CD), but the FLAC is 11948032 samples. I checked the FLAC and found that all the blocks are size 4608, even the last one, whereas in a properly encoded file, the last block will only be as big as it needs to be. So I suspect that there's a problem with how the end of the input stream is being handled; the last batch of samples won't fill up a full-size block, but they get dumped into one, anyway. Would this be something BASSFLAC is doing wrong, or are they invoking it wrong, or what?
(The screenshot also shows a spurious ID3v1 tag, but I confirmed that removing it doesn't affect anything).
Any feedback appreciated. Sorry if these are dumb questions.
