[heymian]

Before I start, keep in mind I'm strictly speaking about bit-depth only. Sampling rates, upsampling, resampling, etc. are a whole different discussion.

I've read you should preferably match your bit-depth (in Windows) to the source you're listening to. For example, if you're listening to a CD, you would want to set your bit-depth to 16-bit.

But I've also seen people recommend setting Windows to 24-bit bit-depth. Reasons being, supposedly an increase in sound quality, especially when using DSP's. And also because most DAC's these days are 24-bit. So again, regardless of source, you should set Windows to 24-bit.

What are your guys opinions or findings on this. Personally, I've been using 24-bit for a few days and all I've noticed is my sound is quieter.

 

[MindsMirror]

There is no significant processing or reduction in sound quality that happens when converting 16 bit audio to 24 bit. To go from 16 to 24, you just add 8 zeros to the LSB end of each sample. If you do not lower the volume or change the audio with DSPs, there is no difference either way. If you lower the volume of 16 bit audio and play it back in 16 bit mode, you will lose some precision and reduce the dynamic range. Because the 24 bit noise floor is lower, playing back in 24 bit mode theoretically allows you to lower the volume of 16 bit audio by up to 48dB without reducing the dynamic range. Your devices will not actually have 24 bit performance so in practice it's not that much better. 16 bits will be more than enough for practically every playback situation, but there are no downsides to choosing 24 bit playback mode so you might as well choose it.

 

[Yuri Korzunov]

Before I start, keep in mind I'm strictly speaking about bit-depth only. Sampling rates, upsampling, resampling, etc. are a whole different discussion.

I've read you should preferably match your bit-depth (in Windows) to the source you're listening to. For example, if you're listening to a CD, you would want to set your bit-depth to 16-bit.

But I've also seen people recommend setting Windows to 24-bit bit-depth. Reasons being, supposedly an increase in sound quality, especially when using DSP's. And also because most DAC's these days are 24-bit. So again, regardless of source, you should set Windows to 24-bit.

What are your guys opinions or findings on this. Personally, I've been using 24-bit for a few days and all I've noticed is my sound is quieter.

Recommended use ASIO or WASAPI driver for avoiding any processing in Windows.
 
For playback on DAC better use its native maximal resolution.
 
If your source have other resolution there are two ways:
 
1. Sent to DAC source as is;
 
2. Pre-convert to native DAC resolution offline.
 
As rule with bit-depth conversion must not be problems even for Windows drivers (nor asio, nor wasapi). Especially 16 to 24 bit.
 
24 to 16 bit demands qualitative dithering.
 
If bit-depth scaled correctly must not be difference in loudness. It can be trouble of your DAC too.

 

[castleofargh]

 you don't lose anything if you set foobar(or whatever) to output in 24bit bit. as mentioned, if it's 24bit music, then it works just perfectly fine. and if your song is in 16bit then it will had a few zeros at the end of the sample value(no real CPU usage as there is no complex job to do, and no bit lost in the process as it's just padding after the actual data. in both situations you really have nothing to lose.
 
also it can help for volume control, so it's really a non question, output as high as the DAC can effectively take and never think about it again.