Is the volume control on DACs usable?

[martin900]

Hi all,
What is the current situation on volume control on DAC's like the Xsabre Pro, is it any good?
I've been told many times in the past not to use any kind of digital volume control due to loss of resolution on lower levels, is it still true?

 

[RayDunzl]

Given today's hardware implementations of digital volume control, it is the least of my worries.

 

[restorer-john]

Technically, modern digital volume controls are essentially flawless.

In the real world, they can have issues. They can 'forget' their attenuation settings, they can lock-up, or can have audible glitches, are slow or too fast and really aren't that intuitive.

Give me a pot anytime. And if that pot controls a digital attenuator with all the characteristics of a phyisical pot without the downsides, even better.

Benchmark use such a system I believe.

 

[Audiophonics]

Hi all,
What is the current situation on volume control on DAC's like the Xsabre Pro, is it any good?
I've been told many times in the past not to use any kind of digital volume control due to loss of resolution on lower levels, is it still true?

Hi,
You can take a look here : http://www.esstech.com/files/3014/4095/4308/digital-vs-analog-volume-control.pdf

 

[Purité Audio]

A combination of variable analogue output and digital attenuation works pretty well, the difference between the two analogue/digital is that with analogue you attenuate everything , with digital only the signal, then again with properly dithered attenuation I don’t think it matters.
Please do correct if wrong.
Keith

 

[Vincent Kars]

If you play 16 bits program material on a 16 bit DAC and reduce with 48 dB you end up with 8 bits

MSB LSB
1111111111111111
0000000011111111

Indeed a substantial loss of resolution
If you do the same on a 24 bit DAC

111111111111111100000000
000000001111111111111111

You can reduce with 48 dB without any loss.
Today there are DACs having a 32 bit internal data path….

All of this is a bit theoretical as in practice the noise floor will be higher than -144 dBFS but it gives you an idea about the impact of digital volume control.

 

[RayDunzl]

Visual Example.

The waveform represents voltage. The attenuation is 80dB using 32bit data. The attenuation is 1/10.000 of the original signal.

80dB attenuation represents from "quite loud" to "you'll be lucky to hear anything at all" in your room on musical content.

Full scale digital signal:

And at -80dB (visually zoomed in as far as possible with the software)

 

[martin900]

Technically, modern digital volume controls are essentially flawless.

In the real world, they can have issues. They can 'forget' their attenuation settings, they can lock-up, or can have audible glitches, are slow or too fast and really aren't that intuitive.

Give me a pot anytime. And if that pot controls a digital attenuator with all the characteristics of a phyisical pot without the downsides, even better.

Benchmark use such a system I believe.

This is also one of the main concerns. I've heard a DacMagic Plus going on a white noise bonanza through a Krell power amp. Almost played the brown note.
Not really something you want with a pair of small bookshelfs like the LS3/5A that on that day are connected to a 250W/8r amp.

 

[BDWoody]

This is also one of the main concerns. I've heard a DacMagic Plus going on a white noise bonanza through a Krell power amp. Almost played the brown note.
Not really something you want with a pair of small bookshelfs like the LS3/5A that on that day are connected to a 250W/8r amp.

I've had something similar happen with a careless 'what's this button for,' leading to the reflex of quickly covering both ears with my hands, and then realizing one of those ears is going to have to take one for the team when I reach to turn it off...

It is startling...

 

[RayDunzl]

Uh oh.

Just learned something.

"The dynamic range that can be represented by a 32-bit (floating point) file is 1528 dB."

https://www.sounddevices.com/32-bit-float-files-explained/

 

[Fluffy]

This is what I had to say about the subject last time it came up:

https://www.audiosciencereview.com/forum/index.php?threads/are-pre-amps-necessary.9327/#post-243960

 

[JohnYang1997]

Why the misconception is still dominating here
Digital volume doesn't attenuate the noise. So no matter what the SNR is going be reduced. It's not a matter of digital resolution but the residue noise of the analogue stage. If you don't hear noise at max volume then good go ahead. If it's a yes or you want a good SNR in your system, especially when paring with power amp who amplifies the noise further, you need a passive volume control or something in combination of the two to attenuate the noise with the signal to achieve high SNR.

 

[bennetng]

Hi,
You can take a look here : http://www.esstech.com/files/3014/4095/4308/digital-vs-analog-volume-control.pdf

My response to this pdf file:
https://www.audiosciencereview.com/...hould-i-set-my-dac-to.8956/page-4#post-255104

 

[JohnYang1997]

Hi,
You can take a look here : http://www.esstech.com/files/3014/4095/4308/digital-vs-analog-volume-control.pdf

This is quite nice.

 

[martin900]

So in the end it is still *not* the preferred way of controlling the volume it seems and so far only Sabre claims that it is because their DAC is 32Bit internally and thus has enough resolution for accurate volume control?

 

[JohnYang1997]

So in the end it is still *not* the preferred way of controlling the volume it seems and so far only Sabre claims that it is because their DAC is 32Bit internally and thus has enough resolution for accurate volume control?

The i/v and output buffer are also a big source of noise. Sabre's way is better than just attenuate the incoming datastream but still not optimal way.
Only way to judge is to see whether you can hear the noise or not. If the answer is no then for the most part why bother.

 

[bennetng]

So in the end it is still *not* the preferred way of controlling the volume it seems and so far only Sabre claims that it is because their DAC is 32Bit internally and thus has enough resolution for accurate volume control?

What I meant was the quality of the explanation is poor, not necessarily the quality of the ESS volume control itself. I highly suspect the actual math in a ESS DAC (and their competitors) is more advanced than those mentioned in the pdf file.

Anyway, here are some test files for you to try:
https://www.audiosciencereview.com/...alog-volume-controlled-dacs.11395/post-325447
https://www.audiosciencereview.com/...ume-control-guidance.10198/page-2#post-281405