That's fair as well. Then again, I'm not sure if the volume controls on the respective software doesn't attempt some form of dithering anyhow. Plus, the native dither due to noise dominating the LSB.
I don't really sweat it either, but I use Sox, and that does dither if you tell it to. Given I'm about to do digital crustiness (crossover! Haha, silly phone) and my internal bit resolution will be higher (int32), I'll need to dither after everything anyways.
I don't really sweat it either, but I use Sox, and that does dither if you tell it to. Given I'm about to do digital crustiness (crossover! Haha, silly phone) and my internal bit resolution will be higher (int32), I'll need to dither after everything anyways.
Last edited:
I also try to make a sensible use of dither with brutefir. But have to admit that it all pretty sounds the same to my ears, in spite of many years of trying to train them into sophisicated golden devices...
But the half deaf will be half deaf, and that's fair as well...😎
Dithering is to be applied at the final stage (downsampling) when the lower X bits are truncated to fit the soundcard format - typically 24 bits these days. Actually, many people say dithering makes no sense for downsampling to 24bits (e.g. sox dithers only to 16 bits automatically). I agree with them as the 24th bit (the only one affected by dithering) is way below any DAC noise floor (I assume people do not have their audio DACs cooled with liquid nitrogen).
Volume control itself does not need dithering - it is just an arithmetic operation of division by a fixed number. IF done in larger bit depth, the subsequent final downsampling can be accompanied with dithering.
Volume control itself does not need dithering - it is just an arithmetic operation of division by a fixed number. IF done in larger bit depth, the subsequent final downsampling can be accompanied with dithering.
- Status
- Not open for further replies.