I think as rossy that it's better to converting the shn>wave>flac so you got more downloaders on the wonderful music you have think to upload, but if you decode the shn to wave and then encode it to flac, then we that took it down decode it back to wave again, that mean we got the same wave as you encode the flac to, so we got the same sector boundary errors as it was in the first time when it was shn, because the wave is the same so on that it doesn't matter if it is an shn or flac.

But one thing that matter is as rossy say that more players can play flac than shn, the old winamp can play shn very great but the newer versions got problem with it, to decode flac or shn to wave to burn out an album is really the same, yeah flac is little faster to decode but who cares about a couple of minutes.

So it's like that flac is acepted by everyone and works with every player and the people have easy to decode it, when shn is an old way of compress the files that are really dying, and one thing more is that flac makes the files smaller too, yeah the lastest flac frontend make them as small as ape now.

So you can convert shn over wave to flac without loosing anything, and we got the same wave as you got when you decoded the shn.

If I may weigh in. The best advice in all five pages of this thread is:

Always encode to FLAC level 8. Yes, it may take a couple extra minutes, and the savings are not huge, but even 1 MB saved up front on a torrent downloaded 1,000 times saves a GB of internet traffic.

Next, MD5's do not lie. If you go from SHN > WAV(1) and then WAV(1) > FLAC > WAV(2) and the resulting MD5's match on WAV(1) and WAV(2), then you have a 100% digital replica.

I prefer flac as the compression of choice and I fully support converting SHN to FLAC (or even APE which typically has better compression ratios). There are so many tools out there that can decode or convert these lossless formats that there really is no reason to stick with the oldest, most bloated format.

Now... Let's talk of digital degradation...

There is a chance of error when converting file formats, but these errors would typically be caused by failing hardware or system instability -- and are pretty rare. For example, let us suppose that during the decode phase the bits being written to the HD are corrupted and the drive is unable to determine the error. Or at a later time, before being converted back to FLAC, the drive suffers from cross linked files. In cases such as these, there will be errors introduced to the file. Now, more than likely these errors would be caught by the encoder and the creation of the FLAC file would fail... But I suppose there is a very small chance the error would pass undetected and a damaged file would be spread.

The risk is much lower than going SHN > WAV > CDDA > WAV > FLAC. The reason for this is that CDDA has no checksum data and it is up to the drive/software reading the disc to determine a "good" read. Have you ever noticed that an 80 minute CD-R can hold 800 MBs of WAVs when written as an audio CD, but only 700 MBs of files when written as data? This is because each sector that is written on a data CD has a checksum that is used to ensure the data being retrieved is 100% correct. Over the length of the CD, this checksum data adds up and takes up nearly 12% of the disc.

CDDA doesn't have this checksum system, and often cheap(er) CD-ROM drives/software solutions are unable to determine if the read of the sector was 100% accurate. When bad reads are not detected, the resulting file typically ends up with clicks, digi-farts, and other strange noises.

Bottom line... I am all for the conversion of SHN to newer and better formats. However, it is often good form to denote this in the lineage of the upload so that purists may attempt to find the original SHN files.