I would like to publish music on Internet with the best quality possible. Originally, the stereo masters are in analogic (studer) or in 24 bits at 192 khz. Because it’s a downloading thru a server (and not thru a p2p), the size could be higher than usual.

I’m trying to find the best coding.

Does a losy coding be better than a loseless at the same size?
For example, a PCM 16/44.1 and a MPC 24/96 with bitrate @ 1.4 MB (I don’t know if it’s possible, but it’s for the example).

Thank you for your expertise.

Sorry, I'm in the wrong category (Genaral - (AAC)). I should be in Hydrogen Audio Discussion (General).

Currently, there are no lossy compression formats (that I'm aware of) that support such high sample rates. The format/setting you should choose depends a great deal on your target audience. For example, noone with a modem connection will download losslessly compressed files such as Monkey's Audio (.ape) @ 50 MB or more. And, they won't be happy about bitrates over 200 kbps either.

But if your idea is to present the highest possible quality for download with reasonable filesizes, I suggest you put the 24 bit / 44.1 kHz (downsampling might be necessary if you can't capture at this sample rate) thru mppenc.exe, in other words to use Musepack with the --xtreme setting. For a wider audience, go with Ogg or AAC, and if server space isn't an issue, why not provide several formats?

Cheers,
Uosdwis

Thanks Uosdwis,



It's more broadband oriented.



Not at 48 or 96 khz



Sure, I need also a smaller size and widely recognize (mp3) and 5.1 (dolby or DTS). But in this discussion I would like the reference (the best) in stereo.

As far as I know, AAC can handle up to 96khz, and dolby 5.1. It's not as well recognized as mp3, but you can play it in Winamp & Sonique, so it won't be hard for your downloaders to reap the benefits of the increase in quality :o)