Reducing DAC volume doesn't necessarily decrease bit-depth of audio data.

[Zapper]

I'm apparently on a mission to stop people from asking for 32-bit converters—if enough people ask, marketing (not engineering) will surely fulfill your requirement.

We need 32-bit converters so marketing can sell us 190dB SNR headphone amplifiers. If you've paid for 32 bits you want to be able to hear them all. ;-)

 

[Deleted member 58865]

You mean you need the output from the DAC to be attenuated by more than 48 dB before it enters the amplifier? If yes, then it still doesn't matter, because the resolution of the output of any DAC is limited by physics to 22 or 23 bits, no matter how many bits it is using internally. You will cut into the 16-bit of the source signal whether the DAC is using 24 or 32 bits.

At least, I'm lucky enough to not need more than the total of -48dB of digital attenuation.

I guess a passive stepped attenuator is beneficial for preserving at least 16 bits of resolution at the DAC output.

 

[earlevel]

At least, I'm lucky enough to not need more than the total of -48dB of digital attenuation.

I guess a passive stepped attenuator is beneficial for preserving at least 16 bits of resolution at the DAC output.

If I had time I'd draw a diagram...the reason this isn't a problem is...

First, I'm going to assume that we're not talking about a 16-bit CD player on a 16-bit CD player, with digital volume control ahead of a 16-bit converter. I don't know for sure if players like that were common, at one time, but I'm assuming you're talking about 16-bit audio, sent to a 24-bit converter with digital volume control in the 24-bit hardware (or, equivalently, playing back 16-bit audio on a computer connected to a 24-bit converter).

Second, I'll assume that the loudest you want to listen to is with no digital attenuation. Here, I'm just assuming rational gain staging. And note this caveat is not restricted to digital volume control—if your source must be turned down into the mud and compensated for with large amp gain, you will be amplifying mud, it doesn't matter whether the source is digital or analog.

Given that, say you're listening real loud (no digital attenuation) at time, but at this time you want a more relaxed volume to listen to, and you want to back off your digital volume control by about 12 dB. I'll keep this simple, and say this is two bits of reduction, the same shifting the 16-bit word two bits right (~-12.04 dB). Are we now listening to 14 bit audio? No, because we have a 24-bit converter. We're still listening to 16-bit, and the attenuation is basically the same as if we had stepped and analog volume know -12 dB, or a fourth of the original voltage amplitude.

I realize that you understand this already, which is why you called for 24-bit DACs. But I think this is pretty much universally the case at this point in time. And really, even if we were restricted to 16-bit DACs—which we are not—even effectively operating at 14-bit is not a problem is this scenario, because the 16-bit floor covers normal listening pretty effectively. What does change, though, is that optimal gain staging become more important. We don't want to waste bits by having loud amplification and having to turn the sound down digitally more than necessary. But if we're gain staged correctly, the digital noise floor is down more than 90 dB, and stays at that level as we turn down the volume digitally.

Of course, we don't want to just support shifting of bits, we want a real multiplication of gain, and we want a 24-bit result. We dont' want to dither to 16-bit, we want to dither to 24-bit (though as I've said many time, if you don't dither at the 24-bit level, it doesn't matter because you can't hear it anyway. The is 256 times as true if we start with 16-bit audio —which of course should be dithered—because the noise floor, the error, is already 48 dB above the the 24-bit error floor).

Hope that makes sense. The short version is that 24-bit DACs and math are a perfect fit for digital attenuation. True whether the source is 24-bit or 16-bit, but with 16-bit, the argument is trivial, you basically lose nothing, because there is digital room for the finest details to move down, well into the bits you can't hear. (With 24-bit, you do lose something, but it's something you couldn't hear anyway. Still, it will bother some that something is being lost, and they might disagree with digital attenuation being fine. It's OK, folks, we can disagree.)

I'll add that if you can think of why this isn't true, it's because you're thinking of a listening situation that is horribly gain staged. If a system is 100% analog and equally horribly gain staged, you're in the same situation, audibly degraded sound.

 

[earlevel]

Topping D10s comes with ES9038Q2M DAC which seems to process 32-bit audio samples internally.

With 32-bits, I would have 16-bits at -96dB of DAC volume.

-48dB may not be enough for a power amplifier or for an integrated amplifier at its max volume. Paul McGowan said at least some DACs use 50-bit words internally.

I would want at least 32 bits for a power amplifier.

How many bits are used in Topping D10s?

These and some other comments make me think you don't have proper perspective on resolution.

First, "at least some DACs use 50-bit words internally": For DSP processes, we often need to retain precision, temporarily, that is far more that the actual digital to analog conversion. This is especially true of filters. I'll make up a dumb process here as an example, but it's not far off from what goes on in practical filters. n / 10... x 10. We should expect to get n back. But if n is a value 0-99, feeding a process that accepts 0-99, if our divide and multiple doesn't retain fractional values, then an input of 24 would yield a value at the output of 20, a huge error. If that math accommodates a range of 0.000-999.999, for instance, math like this comes out fine.

But for a gain change, we don't need this level of precision, because there is no penalty for losing unused bits. If we have 24-bit audio (fixed point, roughly representing -1 to 1). Now, the gain value could be 24-bit, because we're already working with that resolution, or it could be 16-bit or even 8-bit. Whatever it is, the results of the multiple yields (basically) the resolution of the digits added together (9 x 9 = 81, for instance, it take 2 digits to multiple two 1-digit values). If you don't, you lose the least significant digits. But, we're putting the output into a 24-bit DAC anyway, so we're going to toss those bits. We could retain one for dither, for folks who must dither 24-bit, but the point is we doesn't need Pauls 50-bit words just for gain. For filters, especially recursive, sure, but not just the gain.

Second, "With 32-bits, I would have 16-bits at -96dB of DAC volume": Again, recognize that 24-bit DACs are the limit of practicality, limited by physics (and fortunately, our ears anyway). And do you understand the implications of having 96 dB of volume control range? Basically, it means a range of being loud enough to cause hearing loss over time to being quieter than your own breathing at rest. I'm not saying it's bad to have that range, just giving a rough reference to what it means. More importantly, you seem to want this control down to the quietest sounds in the quietest rooms, while retaining the details of the recording that are another 90+ below that barely hearable level.

Two problems: The voltage levels represented by the bottom ~60 dB of that are below the floor voltage noise level of electronics. (No, you can't make it quieter.) And your ears are already less sensitive than that level anyway. For sound levels, if you're listening at a very loud 96 dB SPL, so that 0 dB SPL is the floor, and you turn down 96 dB so that that floor is now at -96 dB SPL...well, in a typically "quiet" room, you won't hear 0 dB SPL, so you can imagine what a stretch it is to think you might be able to hear another 96 dB down from that.

My point again, is that 24-bit is not only better than the best we can do (the error in the bottom bits is relatively huge, constrained by physics), it also outpaces our ears by a wide margin. And the only way you can make the latter not true is with horrible gain staging—an enemy of both digital and analog audio path, equally.

As an aside, I never see audiophile talk about this, but you don't want the loudest amp and speaker combination that you can buy. Say, for instance, you buy amp and speakers that can attain 150 dB SPL at you listening position. Well, since you don't want to be in pain and then hear very little the rest of your life, you might dial your preamp back, say 65 dB. But, the preamp still puts out its minimum noise floor, turning it to zero will not stop that. (The amp too has a noise floor ahead of its massive gain, so it's not a fault of the preamp, it's a fundamental limit of electronics, and it doesn't matter which stage you blame it on.) Ideally, you want the amplification to be louder than you will listen to at the loudest (you want some headroom) when your source is turned all the way up. Then use digital (or analog) attenuation for non-max levels. That will give you your full dynamic range and lowest noise floor.

Lastly, if you're worried about gear that doesn't use enough bits for the gain multiplication, you're fundamentally worrying that someone out there might make crappy gear. Yes, people make crappy gear, analog and digital. For the most part, the worst offenders go out of business, and that encourage the rest to try to not make crappy gear. And that includes cheap analog level controls that get noisy and scratchy, something that won't happen with digital volume control. Except for in crappy gear.

 

[antcollinet]

These and some other comments make me think you don't have proper perspective on resolution.

First, "at least some DACs use 50-bit words internally": For DSP processes, we often need to retain precision, temporarily, that is far more that the actual digital to analog conversion. This is especially true of filters. I'll make up a dumb process here as an example, but it's not far off from what goes on in practical filters. n / 10... x 10. We should expect to get n back. But if n is a value 0-99, feeding a process that accepts 0-99, if our divide and multiple doesn't retain fractional values, then an input of 24 would yield a value at the output of 20, a huge error. If that math accommodates a range of 0.000-999.999, for instance, math like this comes out fine.

But for a gain change, we don't need this level of precision, because there is no penalty for losing unused bits. If we have 24-bit audio (fixed point, roughly representing -1 to 1). Now, the gain value could be 24-bit, because we're already working with that resolution, or it could be 16-bit or even 8-bit. Whatever it is, the results of the multiple yields (basically) the resolution of the digits added together (9 x 9 = 81, for instance, it take 2 digits to multiple two 1-digit values). If you don't, you lose the least significant digits. But, we're putting the output into a 24-bit DAC anyway, so we're going to toss those bits. We could retain one for dither, for folks who must dither 24-bit, but the point is we doesn't need Pauls 50-bit words just for gain. For filters, especially recursive, sure, but not just the gain.

Second, "With 32-bits, I would have 16-bits at -96dB of DAC volume": Again, recognize that 24-bit DACs are the limit of practicality, limited by physics (and fortunately, our ears anyway). And do you understand the implications of having 96 dB of volume control range? Basically, it means a range of being loud enough to cause hearing loss over time to being quieter than your own breathing at rest. I'm not saying it's bad to have that range, just giving a rough reference to what it means. More importantly, you seem to want this control down to the quietest sounds in the quietest rooms, while retaining the details of the recording that are another 90+ below that barely hearable level.

Two problems: The voltage levels represented by the bottom ~60 dB of that are below the floor voltage noise level of electronics. (No, you can't make it quieter.) And your ears are already less sensitive than that level anyway. For sound levels, if you're listening at a very loud 96 dB SPL, so that 0 dB SPL is the floor, and you turn down 96 dB so that that floor is now at -96 dB SPL...well, in a typically "quiet" room, you won't hear 0 dB SPL, so you can imagine what a stretch it is to think you might be able to hear another 96 dB down from that.

My point again, is that 24-bit is not only better than the best we can do (the error in the bottom bits is relatively huge, constrained by physics), it also outpaces our ears by a wide margin. And the only way you can make the latter not true is with horrible gain staging—an enemy of both digital and analog audio path, equally.

As an aside, I never see audiophile talk about this, but you don't want the loudest amp and speaker combination that you can buy. Say, for instance, you buy amp and speakers that can attain 150 dB SPL at you listening position. Well, since you don't want to be in pain and then hear very little the rest of your life, you might dial your preamp back, say 65 dB. But, the preamp still puts out its minimum noise floor, turning it to zero will not stop that. (The amp too has a noise floor ahead of its massive gain, so it's not a fault of the preamp, it's a fundamental limit of electronics, and it doesn't matter which stage you blame it on.) Ideally, you want the amplification to be louder than you will listen to at the loudest (you want some headroom) when your source is turned all the way up. Then use digital (or analog) attenuation for non-max levels. That will give you your full dynamic range and lowest noise floor.

Lastly, if you're worried about gear that doesn't use enough bits for the gain multiplication, you're fundamentally worrying that someone out there might make crappy gear. Yes, people make crappy gear, analog and digital. For the most part, the worst offenders go out of business, and that encourage the rest to try to not make crappy gear. And that includes cheap analog level controls that get noisy and scratchy, something that won't happen with digital volume control. Except for in crappy gear.

Well done for taking on the baton - I've lost the will to live by this point.

 

[BeQuietZen]

Second, I'll assume that the loudest you want to listen to is with no digital attenuation. Here, I'm just assuming rational gain staging. And note this caveat is not restricted to digital volume control—if your source must be turned down into the mud and compensated for with large amp gain, you will be amplifying mud, it doesn't matter whether the source is digital or analog.

I'll add that if you can think of why this isn't true, it's because you're thinking of a listening situation that is horribly gain staged. If a system is 100% analog and equally horribly gain staged, you're in the same situation, audibly degraded sound.

Ideally, you want the amplification to be louder than you will listen to at the loudest (you want some headroom) when your source is turned all the way up. Then use digital (or analog) attenuation for non-max levels. That will give you your full dynamic range and lowest noise floor.

Great posts, right to the heart of the problem and beautifully summarized.

Unfortunately even the newest models of amps from famous brands like Denon, Marantz, NAD, Rotel, etc, adhere to the past standards of 200 mV for full power. My main DAC SMSL DO100 have 2V rms SE out and 4V XLR and I'm in search for a decent power amp.
I think even the Hypex NC-252MP have too much gain for balanced setup (without lowering the gain up to -50 dB on my DAC) for normal listening levels with classic dynamic speakers of 86-89 dB sensitivity.

I guess amps with 20 dB gain (or so) will bi more compatible with todays DACs with 2V output (so common these days).
I'd like to hear your (and others) opinions on this particular "problem".

 

[antcollinet]

Great posts, right to the heart of the problem and beautifully summarized.

Unfortunately even the newest models of amps from famous brands like Denon, Marantz, NAD, Rotel, etc, adhere to the past standards of 200 mV for full power. My main DAC SMSL DO100 have 2V rms SE out and 4V XLR and I'm in search for a decent power amp.
I think even the Hypex NC-252MP have too much gain for balanced setup (without lowering the gain up to -50 dB on my DAC) for normal listening levels with classic dynamic speakers of 86-89 dB sensitivity.

I guess amps with 20 dB gain (or so) will bi more compatible with todays DACs with 2V output (so common these days).
I'd like to hear your (and others) opinions on this particular "problem".

Two words:

Attenuator pad.

 

[sonder]

Two words:

Attenuator pad.

New guy, learning a ridiculous amount here very quickly.

Are these fixed attenuators (resistor based?) with high transparency that you can stick between lines to lower the signal in without using variable attenuation / cheap pots, and not have the worries if imbalance or distortion being introduced at magnitude.

Effectively a firewall for ensuring the volts in cannot get too high?

 

[MaxwellsEq]

New guy, learning a ridiculous amount here very quickly.

Are these fixed attenuators (resistor based?) with high transparency that you can stick between lines to lower the signal in without using variable attenuation / cheap pots, and not have the worries if imbalance or distortion being introduced at magnitude.

Effectively a firewall for ensuring the volts in cannot get too high?

That's a nice way of putting it! You can work out the ratio between maximum DAC output and the loudest you will ever want to listen, then pick a resistor network to replicate that ratio whilst not messing up impedances too much. Then if the DAC goes to accidental full beans, it will just be very loud, but not dangerously so.

When CD came out, it's line-out was much higher level than tape, cassette and tuners which made it very loud but it could also cause some amplifiers discomfort. The trick with older amplifiers was to put a pad in to halve the maximum voltage. I had a preamp I loved, but it predated CD, so I used a pad for the 30 odd years I used it!

 

[antcollinet]

New guy, learning a ridiculous amount here very quickly.

Are these fixed attenuators (resistor based?) with high transparency that you can stick between lines to lower the signal in without using variable attenuation / cheap pots, and not have the worries if imbalance or distortion being introduced at magnitude.

Effectively a firewall for ensuring the volts in cannot get too high?

Pretty much, yes.

You have to select the attenuation to match what you need based on the amp gain and your source ouput level, but they won't introduce any distortion to the signal.

Ideally choose the level of attenutation such that the volume with the DAC on max is a little bit louder than you ever want to listen to (but not so loud it can cause damage) then turn the volume down from there using the DAC.

If your amp has a volume control you can do the same with that, rather than attenuators - if you want to use the DAC to control volume.


EDIT : Ninjad by @MaxwellsEq

 

[Ken Tajalli]

Theoritically it can have a stepped attenuator - are you aware of any DACS that do? (The RME has a crude one)

Yes, I sent one in! Leema DAC indeed implements one. A four channel version to control balanced outputs.
But it only was necessary, as it also has a dedicated analogue input/output and no ADC.
Otherwise, digital volume controls are just fine.

However, the main point is that even with that resolution reduction, it doesn't matter since the noise floor of any half way decent DAC will still be inaudible.

Right.

 

[Blumlein 88]

New guy, learning a ridiculous amount here very quickly.

Are these fixed attenuators (resistor based?) with high transparency that you can stick between lines to lower the signal in without using variable attenuation / cheap pots, and not have the worries if imbalance or distortion being introduced at magnitude.

Effectively a firewall for ensuring the volts in cannot get too high?

Yes available for XLR or RCA. Common values are 6db,12 db, 10 db and 20 db. Place them at the amp input and they'll be just fine.

Home

Harrison Labs 12 dB RCA Line Level Audio Attenuator PairAttenuators reduce the drive signal to an amplifier so that the input level or gain control can be adjusted in a more useful range. For example, if the line driver has an 8 volt output and the amplifier has a 100mV (0.1V) to 2 volt range of...

www.parts-express.com

Home

In-Line XLR Attenuator Pad 10dBSome audio devices will overdrive an audio input whether by design or by accident. Mixing consoles in theaters are often designed to output a 3 volt signal that can overdrive the 1 volt input of a studio mixing console. An attenuator or pad will reduce the overall...

www.parts-express.com

 

[JustJones]

Benchmark DAC 3 volume control is 32 bit DSP and it has choice of -10db and-20db pads on the XLR output. Does that cover all the requirements so far?

 

[mdsimon2]

Great posts, right to the heart of the problem and beautifully summarized.

Unfortunately even the newest models of amps from famous brands like Denon, Marantz, NAD, Rotel, etc, adhere to the past standards of 200 mV for full power. My main DAC SMSL DO100 have 2V rms SE out and 4V XLR and I'm in search for a decent power amp.
I think even the Hypex NC-252MP have too much gain for balanced setup (without lowering the gain up to -50 dB on my DAC) for normal listening levels with classic dynamic speakers of 86-89 dB sensitivity.

I guess amps with 20 dB gain (or so) will bi more compatible with todays DACs with 2V output (so common these days).
I'd like to hear your (and others) opinions on this particular "problem".

For amps like the NC252MP with reasonable gain (26 dB) / input sensitivity (1.7 V) I find that 4 V from a DAC is fine, assuming the DAC is somewhat low noise (>116 dB SNR at 4V). The extra input voltage gives you more flexibility in dealing with lower level recordings which is very helpful while still having low enough system noise to be silent with typical sensitivity speakers.

I agree that amps with really low input sensitivity (0.2 V) are not a good match for a 4 V output voltage DAC and a fixed attenuator can be helpful.

Michael

 

[Deleted member 58865]

Yes available for XLR or RCA. Common values are 6db,12 db, 10 db and 20 db. Place them at the amp input and they'll be just fine.

Home

Harrison Labs 12 dB RCA Line Level Audio Attenuator PairAttenuators reduce the drive signal to an amplifier so that the input level or gain control can be adjusted in a more useful range. For example, if the line driver has an 8 volt output and the amplifier has a 100mV (0.1V) to 2 volt range of...

www.parts-express.com

Home

In-Line XLR Attenuator Pad 10dBSome audio devices will overdrive an audio input whether by design or by accident. Mixing consoles in theaters are often designed to output a 3 volt signal that can overdrive the 1 volt input of a studio mixing console. An attenuator or pad will reduce the overall...

www.parts-express.com

• Some amazon reviews for Harrson Labs 12dB RCA Line-level attenuator pair say the in-line attenuators color the sound. Are they lying? Or, is there a truth to it?
• Do I want a stepped attenuator or a series of in-line attenuators? A stepped attenuator would be more flexible, but in-line attenuators will be more compact.

 

[sonder]

For amps like the NC252MP with reasonable gain (26 dB) / input sensitivity (1.7 V) I find that 4v

How are you controlling the volume though, I'm assuming your not just putting it out at 0db without lowering somewhere..

 

[mdsimon2]

How are you controlling the volume though, I'm assuming your not just putting it out at 0db without lowering somewhere..

I have two systems. For my main one I use the digital volume control of an Okto dac8 pro. This system uses Hypex NC252MP amplifiers (~26 dB gain).

For the one in my office I use a MOTU Ultralite Mk5 as a DAC with digital volume control via CamillaDSP. This system uses ICEpower 300A2 amplifiers (~24 dB gain). As the DAC output voltage is 8.6 V and this setup is nearfield I have 18 dB of permanent digital attenuation added via CamillaDSP.

Both systems are completely silent.

Michael

 

[Blumlein 88]

• Some amazon reviews for Harrson Labs 12dB RCA Line-level attenuator pair say the in-line attenuators color the sound. Are they lying? Or, is there a truth to it?
• Do I want a stepped attenuator or a series of in-line attenuators? A stepped attenuator would be more flexible, but in-line attenuators will be more compact.

I haven't used the Harrison Labs pads. I have used their Fmod inline crossovers. I doubt they color the sound. Your mainly talking about a pair of resistors in the signal path in these in-line pads or with stepped attenuators. The stepped units are just switching between different value resistors.

 

[antcollinet]

• Some amazon reviews for Harrson Labs 12dB RCA Line-level attenuator pair say the in-line attenuators color the sound. Are they lying? Or, is there a truth to it?
• Do I want a stepped attenuator or a series of in-line attenuators? A stepped attenuator would be more flexible, but in-line attenuators will be more compact.

Lying would be the wrong word. More likely placebo effect/cognitive bias at play.

Purely resistive pads can't colour the sound, unless there is significant capacitance between them and the amp - which there won't be with a properly designed input stage on the amp.

 

[Hayabusa]

Why all this debate when Amir presents it to us visually?

Here's an example,at -30db you get 94-95db SINAD.
More than enough:

Edit:I'm not sure about the -30db but that's his usual attenuation with this test.

What you see is a line that goes down exactly with the volume, this means the absolute level of the noise/distortion remains constant.