drewfx1

Internally MP3's don't really have "bit depth" in the sense of uncompressed audio. The MP3 codec will work using floating point, so there's nothing really to dither to.

"Exactly what happens" is a rather complicated and very technical process.

Thanks.

What I forgot to mention is, I do export my projects at 44.1kHz/16bit (with dithering) if I need to burn to CD, which honestly, I do less and less of all the time.

What I normally do is export at whatever my project settings are, and make an MP3 directly from that .wav. It's usually a 96kHz/32bit .wav converted directly to a 320k MP3 using Sound Forge.

What I see a lot of people do is, they'll take a *insert khz of your choice here*/24bit project, export/convert it with dithering to 44.1kHz/16bit, then convert it to an MP3. My way of thinking on this is, the less alteration of the original project export you do, the better the end result will be.

This is some interesting stuff...thanks for sharing guys. I must confess, I had no clue about the "no need for dithering" thing in an mp3 situation. When I do export right out of Sonar to mp3, I always change my bit and sample rate in the options and set it to dither at pwr3. This is great that I don't have to do that.

Most of the time though, I export out of Sonar as a full blown wave in the bit and sample rate it was recorded in. Then I master the file, dither, CSR and then export to 16/44 wave as my master file. From there I bring that file into Wave Lab and export as an mp3.

As for hearing the differences in mp3's, anything over 192 kb and I have to really listen to hear differences. I can't hear ANY difference at 320 at all though no matter how hard I try.

-Danny