Reynaud Venter wrote:Dither should always be applied whenever you are reducing bit depth, regardless of going from fixed or floating point. TPDF is more than enough.
Thanks but my question is: should I dither my 24 bit audio file when exporting from Ableton? I assume that Resolve is *not* dithering my audio when I export a video and so I should dither when exporting from Ableton (which then gets imported into Resolve). Thanks.
Honestly, if that critical just use video from Resolve and merge in audio using ffmpeg - c:a copy, shutterencoder or mkvtoolnix. No need to funnel audio through Resolve at all.
I don't think video encoders ever dithers audio. They just re-encode and re-compress. Dithering should always be applied just one time ever in the audio path. Going from 32bit internal in a DAW to 24bit export is when you apply it or you send a 32bit to mastering and they do it.
I think in Live the power 3 algorithm is most preferred.
Thanks for all the replies! What I'm getting is that since I work in 24 bit audio, I can just export it from Ableton Live with no dither. Resolve can take it in as is and export it as 24 bit, and YouTube prefers 24 bit, so there's never any need to reduce it. Thanks everyone!