Digital vs Analog Volume Control, does it matter? Options

[Stephan37]

Thanks to this forum I am an recovering audiophile. It was a very humbling experience to do some double blind testing of 128kbit mps but a necessary one. The cool thing about this process is that I've come to enjoy the music again, not the sound. Thus I listen more with more enjoyment. Thanks to everybody here who have made this learning possible. I have also learned that most audio equipment is not that much different. But since my amp doesn't have a remote control I am in the market for a receiver or new amp. And I like the new features like Airplay and the like.

Now I have been reading a lot about different models etc.

What sometimes comes up is the way the volume control is made - digital vs analog.

Here are my questions:
Should I be bothered? Is one really better? Will I hear a difference?

Thanks in advance for your time!

Stephen

[knutinh]

If it is sanely implemented and you avoid silly things like attenuating by a large amount, then amplifying by a large amount afterwards, it should not be something to worry about.

-k

[DVDdoug]

What sometimes comes up is the way the volume control is made - digital vs analog.

There's not much difference...

With an analog control turned-down, the quietest parts/details get lost in the analog noise. (And, since you are turning-it down, you can't hear those detals anyway.)

With a digital control turned-down, the quietest parts get truncated to silence and you loose resolution. (Again, since you are turning-it down, you can't hear those detals anyway.)

People who worry about "throwing away bits" seem to forget that essentially the same thing happens in analog.

As knutinh noted, in either case if you re-amplify the signal you may be able to hear the quality loss (loss of dynamic range, S/N, or resolution).

[.halverhahn]

Some readings:

http://www.esstech.com/PDF/digital-vs-anal...ume-control.pdf


This post has been edited by .halverhahn: Feb 19 2013, 23:09

[Arnold B. Kruege...]

Some readings:

http://www.esstech.com/PDF/digital-vs-anal...ume-control.pdf

Reference the slide titled "Analog Volume - variable noise"

The article is seriously flawed by an analysis of an incomplete system, especially on the analog side. In all cases it appears to presume that the analog stages following the volume control have infinite or at least vastly greater than 16 bit system dynamic range. While isolated analog stages can have dynamic range that equals or at least approaches 24 bit resolution, real world audio systems including consumer preamps, power amps, speakers and rooms don't even come close.

No surprise given that ESS has essentially bet their future on DAC chips with extremely high resolution. While advancing technology is laudable, their basic methodology which involves paralleling upwards of 8 lower resolution DACs is basically "Brute Force".

[bennetng]

Digital audio need not to convert to integer formats before it reaches DAC or SPDIF.

Floating point audio can achieve "variable noise" as well, we can even change volume by a large amount and many times without audible loss.
Does it mean digital wins?

[phofman]

Digital audio need not to convert to integer formats before it reaches DAC or SPDIF.

I am not sure I understand this correctly, but how do you transfer floating point over spdif or which DAC (chip) takes floating point as input?

[skamp]

Relevant: C5 Headphone Amplifier

THE CHANNEL BALANCE PROBLEM: Devices with conventional volume controls may have audible channel imbalance at very low volumes [i.e., one side is much louder than the other --JDS]. It’s extremely difficult to manufacture volume control potentiometers that maintain tight channel balance below about -40 dB (referenced to full volume). - NwAvGuy

It’s 2013, and it’s finally time to say goodbye to the analog potentiometer. C5 features 64 steps of audibly perfect digital attenuation […] C5 presents only +/-0.1dB of deviation all the way down to -50dB, and only +/-0.55dB at -60dB! [Yes, you can only see 28 steps here, as I'm manually racing the dScope test duration by making larger volume transitions.]

In other words, C5′s digital attenuation achieves perfect audible balance at volumes -20dB lower than the analog Alps RK097.

[pdq]

We seem to be comparing apples and oranges.

We all agree on how old-fachioned analog potentiometers work, and their limitations. When we discuss digital volume control, however, we are talking about two different things: there is the all-digital multiplication of the values by a scale factor before sending them to a DAC; then there is the device that digitally selects one of several analog dividers after the DAC.

So, what exactly are we discussing here?

[Arnold B. Kruege...]

We seem to be comparing apples and oranges.

We all agree on how old-fachioned analog potentiometers work, and their limitations. When we discuss digital volume control, however, we are talking about two different things: there is the all-digital multiplication of the values by a scale factor before sending them to a DAC; then there is the device that digitally selects one of several analog dividers after the DAC.

So, what exactly are we discussing here?

I'm responding on the topic of all-digital multiplication of the values by a scale factor before sending them to a DAC;

The selection of various analog dividers is digitatlly-controlled analog, and from a signal standpoint it is 100% analog.

This is BTW the means by which contemporary AVRs implement their volume controls, even when there is a DSP in the signal flow.

[bennetng]

Digital audio need not to convert to integer formats before it reaches DAC or SPDIF.

I am not sure I understand this correctly, but how do you transfer floating point over spdif or which DAC (chip) takes floating point as input?

What I mean is we can change volume in floating point before it reaches DAC and SPDIF. For example, typical DAWs and even freeware like Audacity can change volume and save in floating point. Only in the final stage we need to convert to integer.

This post has been edited by bennetng: Feb 21 2013, 01:57

[knutinh]

What I mean is we can change volume in floating point before it reaches DAC and SPDIF. For example, typical DAWs and even freeware like Audacity can change volume and save in floating point. Only in the final stage we need to convert to integer.

If you have a CD, a CD-player, some dac/pre-amp, an amplifier and a set of loudspeakers, doing floating-point digitalvolume is just another way to implement digital volume. You would still have the fundamental challenge of multiplication by a gain <1, (hopefully) dithering, requantizing to 16 or even 24 bits.

-k

This post has been edited by knutinh: Feb 21 2013, 14:46

[bennetng]

If you have a CD, a CD-player, some dac/pre-amp, an amplifier and a set of loudspeakers, doing floating-point digitalvolume is just another way to implement digital volume. You would still have the fundamental challenge of multiplication by a gain <1, (hopefully) dithering, requantizing to 16 or even 24 bits.

-k

Your example is pretty safe. I agree with you that we still need to face the fundamental challenges you mentioned but requantizing to 24-bit is very safe for 16-bit CD audio as we can reduce volume by more than 40dB without quality loss.

But it is not my main point. What I mean is we can take an digital audio stream, convert to floating point, then apply gain like +33dB then -66dB then +33dB then repeat the above steps 10 times without significant loss.

It sounds silly to do such things when we are only listening to music like the scenario you stated, but I do mixings and sequencings so it is very important to me.

[Stephan37]

Wow, that is some interesting stuff to read here.

This is BTW the means by which contemporary AVRs implement their volume controls, even when there is a DSP in the signal flow.

Do I understand that correctly that your typical AVR volume control is rather bad? Audibly bad?

Stephan

[dhromed]

Well, mine is, but not in terms of any kind of noise or distortion, so what is the definition of "bad" here?

[Stephan37]

Actually a rather good question....

Should I avoid AVRs and look for something more "sophisticated"?
Can one buy a stereo amp for the same money (about 1000 Euros max.) that has a better volume control and do you know any that do?
Or is this a trivial problem, i.e. lets forget about it and buy whatever has best feature/price relationship?

[probedb]

Well my AVR is perfectly fine for music. It's maybe 8 years old now but was a nearly top of the line model in it's time.

[MrEnergizer]

Check out the Onkyo TX-8050

[.halverhahn]

Should I avoid AVRs and look for something more "sophisticated"?
Can one buy a stereo amp for the same money (about 1000 Euros max.) that has a better volume control and do you know any that do?
Or is this a trivial problem, i.e. lets forget about it and buy whatever has best feature/price relationship?

In my opinion, just skip this problem and buy some nice digital gear you can afford and don't care about analog volume control.
Analog has other problems as already mentioned in #8 http://www.hydrogenaudio.org/forums/index....st&p=824917

For 1000€ you can get excelent AVRs:

Yamaha RX-A820
Denon AVR-3313
Onkyo TX-NR717
Pioneer SC-2022

or check the prizes for "last 2 year" products.


This post has been edited by .halverhahn: Feb 22 2013, 00:29

[greynol]

Analog has other problems as already mentioned in #8 http://www.hydrogenaudio.org/forums/index....st&p=824917

Reading further down suggests that the issues raised are no longer relevant.

There is no way I would ever spend a grand on an AVR.

[mzil]

For anyone who, like me, uses their AVR volume knob over a large range of settings, I encourage them to try it out in person, via remote control, before they buy. The Marantz I bought recently (over $1000 USD) is rather awkward in how the volume up and down ramp speed varies depending on your start location, the time you depress the up/down button, and also differs depending on whether you are raising or lowering the volume.[ My previous Yamaha unit didn't have this issue at all. ] I would assume it is the same on all current Denon units as well, in fact it may be a problem with almost everything on the market, for all I know.

This post has been edited by mzil: Feb 22 2013, 04:27

[Arnold B. Kruege...]

Wow, that is some interesting stuff to read here.

This is BTW the means by which contemporary AVRs implement their volume controls, even when there is a DSP in the signal flow.

Do I understand that correctly that your typical AVR volume control is rather bad? Audibly bad?

No. At this time the typical digitally-controlled analog volume control chip has very low noise and distortion as well as excellent channel tracking. For example the Cirrus CS 3318
http://www.cirrus.com/en/pubs/proDatasheet/CS3318_F1.pdf
has 127 dB dynamic range and -112 dB nonlinear distortion. However, its presence in a device with a DSP and DACs with dynamic range > 110 dB seems a little strange unless it is used to implement an analog pass-through feature which some consumers desire for use with digital players with their own high quality converters. In that case it still doesn't make any actual technical sense, but it makes some consumers feel better.

[Arnold B. Kruege...]

Analog has other problems as already mentioned in #8 http://www.hydrogenaudio.org/forums/index....st&p=824917

Reading further down suggests that the issues raised are no longer relevant.

There is no way I would ever spend a grand on an AVR.

I completely agree. The last AVR I purchased (Yamaha RX -V371) was B stock and cost me $118 with full warranty. Other than having fewer channels and no so-called room correction (but it does have independent eq on every channel) I doubt that it would fail an ABX with equipment costing > $1K.

[mzil]

Your B-stock RX-V371 does have a full warranty, a full B-stock warranty, which is half as long as that of a new RX-V371 warranty (two years, parts and labor). As is common with other brands, cosmetic defects on the faceplate, cabinet, and remote are not covered on Yamaha B-stock units, although their occurrence is probably rare and obviously they don't detract from the electrical performance.

[greynol]

Doubt it would fail ABX means what, exactly?

Sounds to me like you're saying you believe your AVR will be distinguishable from the more expensive ones.

This post has been edited by greynol: Feb 22 2013, 17:47