Speaker Level bit depth

Hello Mastering Fellows,

I’m now experimenting on how to implement Acourate DRC into my monitoring chain which consist of Wavelab 8, Mytek DSD DAC and K&H O870 and K&H O410. Pretty simple path !

I’m now using the speaker level to compensate for the necessary drop in level introduced by the DRC (which I put in the dither pane).

So, using the speaker level (or not for that matter, the same question applies): what is the bit depth going through the Asio stream ? Is it dithered to 24 bit ?

Bit depends on the ASIO driver, it is usually 24 bit and WaveLab does not add any dithering.

PG,

Using Speaker Level, one would think that it’s a calculation and it would occur at 32 bit float. If it’s the case, it would have to be dithered to 24 bit. Unless I don’t understand. Would be nice to have a complete schematic of the signal flow for Wavelab 8.

Calculation is done at 32 bit float, but when sent to the audio card driver, it is reduced to 24 bit if the audio card driver only accept 24 bit (some accept 32 bit float).

Reduced… Is it dithered or truncated ?

Truncated. WaveLab never does dithering automatically.

PG,

I understand that you don’t want Wavelab to dither “automatically” for rendering a file. BUT: what is sent to the DAC has to be dithered to 24 bit ! It wont affect any dither in the dither pane for any given file to be rendered.

I find the stand on doing “nothing” without it being specified is admirable. But I’ve always been uncertain about it with regard to dither. You could argue that dither is a basic part of retaining the audio correctly, and that not dithering is a choice to introduce distortion, which therefore should not be done automatically!

The problem arises, of course, when a reduction of bit depth is done at a different time from the final processing (e.g. when saving to a file after listening), and I guess there’s a problem if dithering was automatically done at that stage because of the possibility of creating a clipped peak - though frankly that’s only an issue if you’re playing loudness wars. There’s also the question of choosing a specific noise-shaped dither for final production (though I’ve never been convinced that it’s a significant concern - dither noise is well below the noise-floor of virtually all recordings, and listening venues for that matter).