R2R relay volume vs. digital volume

[olds1959special]

I know this question has been brought up before. Is there any way to quantify what kind of sound differences there are between these two types of volume control? I know that at very low volumes, digital volume can sound worse, due to bit loss, but that loss is not that significant and audible. I am now using an R2R volume control with my source at max volume (Apogee Duet 3 going to Topping A70 Pro) instead of an SMSL DO100 Pro with digital volume control. I am noticing improvements in the bass quality. Or are there other reasons I'm hearing differences, since I'm using a different source component that's being converted from balanced to unbalanced?

 

[voodooless]

sound differences

Volume controls have no sound… as long as the DAC SNR is high enough, a digital control will be just fine. And not all digital volume controls are made equal either…

… and it’s not like this hasn’t been asked before:

Search results for query: Analog digital volume control

www.audiosciencereview.com

 

[olds1959special]

Volume controls have no sound… as long as the DAC SNR is high enough, a digital control will be just fine. And not all digital volume controls are made equal either…

… and it’s not like this hasn’t been asked before:

Search results for query: Analog digital volume control

www.audiosciencereview.com

I have had good results with my SMSL DL200 and D0100 Pro using the digital volume control, so I am in agreement with you. I'm wondering if using an R2R volume control is worse?

 

[solderdude]

I'm wondering if using an R2R volume control is worse?

sound-wise it is not poorer nor better.
You can get weird 'spikes' and a bunch of rattling relays when adjusting the volume.

 

[ssashton]

Most DACs have lower distortion when not operating at full level - which unfortunately is just what many recordings do, trying to hit full scale output as much as possible to sound loud.

Running your DAC at -6dB digital attenuation will give better performance, even if you then use a physical volume control. (But I wouldn't bother, digital volume control in 24-bit or higher is basically perfect, unless you hesr background hiss and need to drop the noise floor)

 

[Purité Audio]

The opposite I believe less distortion ( higher SINAD) at max output.
Keith

 

[ssashton]

The opposite I believe less distortion ( higher SINAD) at max output.
Keith

Yeah, higher SINAD at full level usually, due to signal to noise (noise floor being random hiss). However, full level tends to give higher distortion in terms of harmonics and intermodulation. This is especially true of over-sampling DACs. A few dB headroom in the digital domain does wonders for conversion linearity.

 

[Mnyb]

What you realy need is good gain structure or what to call it . You should use most of your volume control to get the desired loudness .
Then a 24bit digital volume is really all that you need.

If your system are such that you only uses the initial % of whatever volume control you use , and get deafening playback levels.
You have to much gain somewhere.

In perfect world you only get a much higher absolute noise floor.

In extreme cases you can get linearity problems beween channels . More noise and cliks actual quantisation errors if its a not so well implemented digital volume.

With a good gain structure you simply remove all issues to the inaudible domain.
Even id they are not perfect.

Example if you play your typical 16 bit CD content, but your digital volume is a 24bit implementation you only risk loosing bits down at -48dB in volume.

So even a bad volume control that doesn’t convert anything but just truncated the bits would still be fine ( a good digital volume control is better than that ).

A reality check no recording are 16 bit or 24 bit in reality or use the full dynamic range of the CD system the actual noisefloor is usually higher. Let’s say that your typical good recording has 12-14 bits of practical resolution then .

So you don’t loose ”resolution” by using a digital volume control. You get stuff like perfect linearity and perfect channel balance.

 

[ssashton]

As a reality check / practical example - my main source is a Wiim mini streamer going digital to my DAC-preamp. I set the Wiim volume at 80 out of 100, so that my DAC never clips. After the DAC my system has a digitally controlled stepped attenuator going to active speakers. When I listen at normal background level it's around -30dB. When I listen loud, or a recording is quiet, it sits around -15dB.

To some people this may seem like my system is "not very powerful" because the volume control can sit near max level - but that is actually good gain structure making use of the source dynamic range and minimising noise floor. It's good, provided I'm not left wanting extra spl.

 

[melomane13]

another practical exemple:
I use the camilladsp volume control, 32 bit with dither( no loss of quality here ), amp is set to max, so it amplifies all the noise from my sound card (16 bit snr -95db). I only hear a faint noise 15 cm from my speakers (92 db/w), nothing at the listening point. I'm sure everything is fine. With very high-efficiency speakers, I would need a quieter sound card...

 

[Mike123]

I use a balanced volume control on the 4-resistor stage. I can't hear the difference with my ears. But I can see the difference when measuring the amplifier. The amplifier is very low-noise, and the amplified DAC noise at its outputs exceeds its own noise. When adjusting the volume with resistors, with attenuation of 10 dB or more, the DAC noise at the amplifier output becomes lower than the amplifier's own noise and disappears. I repeat: there is no difference in the music. Pure engineering perfectionism.

 

[Mnyb]

I have in the past used inline attenuators to amp's that did not have gain settings , it's a great way to drop 3, 6 or 10db of gain .

I had a HT with passive speakers and an assortment of used power amps with very disparate sensitivity for all channels with attenuators i could bring them closer to each other .

Gain settings on power amps is a necessity nowadays or atleast that you can order them with different gain .

Another very simple example i always used rather cheap monitor speakers like fostex adam or similar as computer speakers , I simply use their volume controls to adapt gain to something usable with my computer or DAC connected to the computer .

 

[boxerfan88]

I know that at very low volumes, digital volume can sound worse, due to bit loss, but that loss is not that significant and audible.

My viewpoint- assuming our room noise floor is 35dBC, and we listen at average 85dBC, thats a 50dB range. Many DAC these days can do 110dB Sinad. At digital volume of -70dB probably nothing useful can be heard because of the room noise, yet there is a further 40dB of nice & clean headroom before hitting the DAC’s distortion/noise floor.

I would just stick with digital volume control.

And if we use digital software volume control like foobar, that operates in floating point instead of integer, even better.

 

[ssashton]

Do the volume controls built in to DAC ICs (AKM, ESS) use floating point digital attenuation? It's hard to find clear info on the internal workings. 32-bit is common, but floating point or dither? No much said.

Edit: I did find this which states the digital attenuation in the ESS DAC is done before over-sampling or ASRC which means a bit of attenuation will avoid over 0dB clipping in the reconstructed waveform.

The Sabre DAC Digital Volume Control: Better than Analog?

The ESS Sabre DAC built-in volume control The volume control is done just before the oversampling filter on the digital data. It can be set to 0.5dB increments and in order to avoid clicking upon v…

hifiduino.wordpress.com

 

[Jazigo]

I am by no means an expert on this, but from what I have gathered: SNR (signal-to-noise-ratio) is what you should try to maintain.

Say you have a volume control on your player, dac and your amp. Find out which one have the largest SNR. The one with the largest SNR is the device you should turn the volume down most on.

By example:

1. Windows:
   1. SNR: 144db
2. DAC:
   1. SNR: 126db
3. AMP:
   1. SNR: 127db

This means that the best you can get from this setup is 126 db. Turn down Windows volume to -18db and amp to -1db.

Still to loud? Turn down all 3 devices with -10db: Windows: -28db, DAC, -10db, AMP -11db.

I may be wrong about this, but I think this is the correct way to to get the most out of your setup.

 

[ssashton]

In practice, it’s not just the signal-to-noise ratio (SNR) that matters, but the actual signal and noise levels.
For example, imagine a DAC with a 120 dB SNR and a 10 V output, feeding a preamp that also has a 120 dB SNR but only accepts a maximum of 5 V input. In this setup, the DAC’s noise will be 6 dB louder than the preamp’s own noise, even though both devices have the same SNR. This indicates that the system’s gain structure isn’t optimized.

Digital volume controls won’t solve this issue. The only fixes are to lower the DAC’s output level—either through a modification, a built-in switch (if available), or by adding passive resistive attenuators. Attenuators work because they reduce both signal and noise together, preserving the SNR while bringing the level down to something the preamp can handle.

 

[Mike123]

For example, imagine a DAC with a 120 dB SNR and a 10 V output, feeding a preamp that also has a 120 dB SNR but only accepts a maximum of 5 V input.

If the owner is concerned about minimal noise, they should first think about level matching. I have 160-watt, 4-ohm speakers. The RMS voltage is 25 volts. The supply voltage of the DIY Class D amplifiers is 25 x 1.41+1 = 37 volts. The DAC output voltage is 4 volts. The amplifier gain is 25/4 = 6.2/16 dB. In this case, the noise level will be minimal. If the amplifier supply is 48 volts, the noise will be higher. If the amplifier gain is 20 dB or higher, the noise level will be higher. If anyone is trying to gain ultra low noise from a resistive volume control, remember that it only makes sense if all of the above are met.