🚀 go-pugleaf

      I can hear a difference in a slight white noise-esuqe type of quality
      but im not exactly sure what im listening to and at which point it takes
      affect on the mixdown.




Not a thing wrong with the foregoing at all, but let's do a Dither 101, in view of what you just said:

Just as a way of explaining what's happening in actual bit terms, dither is applied at the very bottom end of the dynamic range - the last very few level transitions (depending on how it's implemented), so we are talking an almost vanishing-point arefact here. If something like a reverb tail gets down to this level, then at the last level, the output drops randomly between this and zero, and if you listen to it doing this, it's not a very pleasant noise at all, and if this happens with a really low bit depth integer signal (like a 12-bit one, for instance), it's very audible, and unpleasant. With a 16-bit CD though, dither is applied therefore at approximately the -96dB level - hence Arny's comments.

It was discovered years ago that if you artificially manipulate these lowest levels in a fixed pattern, that you can make this annoying transition disappear, and that's effectively what all the different dither options do - they manipulate the fixed pattern, which is what Neil was talking about. What's been discovered is that some of these patterns appear to produce more aesthetically pleasing results than others. You can only apply real dither to an integer-based signal, but even with a floating point one, you can artificially generate your dither signal at the right levels, and then do your bit depth conversion without adding any more dither - and that's what mastering systems like Ozone recommend that you do with their dither algorithm.

Anyway, that's the background story to dither, and I hope puts a little context around the previous posts, in view of your original comments about this.