Quote:
Originally Posted by MattGray  Hey Ed, can you explain why dither wouldn't be needed in this scenario? If the Lavry is capturing at 24bits & you're recording into wavelab at 16bit wouldn't that effectively be truncation?
Are you talking about the master fader in Wavelab? Again if you're raising the gain in Wavelab at 16bits, it would need re-dithering at this point (all internal DSP including a gain change would be calculated at 32bit float) unless I'm mistaken. It would make more sense to capture into Wavelab at 24bit 44.1kHz with your technique, so it could be dithered at the end. Also if you raise up the gain on the MF in Wavelab to digitally clip does that mean you're printing a file that peaks at 0.0dbfs? or are you lowering by 0.3 some place after the gain change?
Matt |
yes, i usually capture 24 bits & yes, use the master fader for clipping. the 24bit recorded file ends up hitting 0.0db. If I set the record dialog to 32bit this would not work as the full dynamic range would still be intact. I use the over shoot of the faders as a 'Limiter" per say, maybe a"clipper" is a better term. I do realize that I'm truncating from 32bit to 24 bit. this sounds more transparent to me than using dither to 24 bits. doing a null test between the two is telling...
I then apply a limiter(to control the output ceiling)& dither to 16 bits. That's the norm for me.
You are correct in saying that if capturing at 16 bits either set the ADC to 16 bits, or if using the above technique put a limiter at 16 bits & lower the output ceiling at the end of the capture chain.
Ed