Understanding the Role of DACs in a Simple Audio Setup

[Lohil]

And lastly, I have not asserted that DACs are perfect, only that they are perfect beyond the level of audibility (unless they’re broken/faulty).

Yes good and fine, this is a very reasonable statement

Please give me an example of a real signal that is not band limited. In addition to instruments and mics being band-limited, as mentioned by VNandor, air itself band-limits any sound wave propagating through it. In addition to the “damping” effect of air that reduces sound level by roughly 6dB per doubling of distance, air also rolls-off high frequencies. For example, a 100kHz sound is rolled-off by about 100dB per 100ft (~3.3dB per meter). In order not to be band-limited, a real sound would have to break the laws of physics!

So, the only question is at what frequency the signal is band limited, not whether it’s band limited and as we’re dealing with human consumers, the obvious frequency would be the threshold of human hearing. Before the advent of digital audio, it was considered that the freq should be around 16kHz, as that is the average threshold for a human adult. Analogue recording media became more unstable/noisy/distorted above 16kHz and indeed, many studio mics were only specified to 14-16kHz, mics which are still in very common use in recording studios around the world today.

Well this is ironically a perfect example of what I'm finding on this forum. Yes you're right all signals are band limited in some way by their medium or carrier.

As for the comment on philosophy this is in general response to

Then you’re in the wrong place! I totally do not understand believing papers because belief has nothing to do with the facts/science. The introduction, methodology, results and conclusions of a scientific paper are either valid or invalid and what you choose to believe has no effect on that. And, one does not have to “confirm something for themselves”, I do not have to confirm that I will die if I jump off a skyscraper or measure my terminal velocity. That’s why we have science in the first place, once something is scientifically proven we don’t have to keep confirming it

Essentially I'm arguing one does have to confirm for themselves what a human body's terminal velocity is, if philosophically you want to claim that as knowledge and not a "belief" but this is mostly tangential.

As it pertains to the question at hand, do DACs color sound, then if one believes that human hearing is fully understood then yes all the arguments are likely a 100% valid and the issue of producing that aforementioned goal of a sound indistinguishable from hearing it first hand, lies in other parts of the audio chain. But as I find there are continually new discoveries on how we hear, I say "believes" as the knowledge it was based on turned out to be incomplete.

But this has all become a rather eristic back on forth on what was only meant to be some simple foundational statements on DACs in general.

 

[bigshot]

You can believe stuff, but it doesn’t make it a fact. Most DACs are audibly transparent. That’s a fact, verified by controlled listening tests, measurements, and established thresholds of human hearing.

 

[gregorio]

Well this is ironically a perfect example of what I'm finding on this forum. Yes you're right all signals are band limited in some way by their medium or carrier.

You mean this subforum is a perfect example of logic and the facts/science over marketing and audiophile myths? And, it’s not just the “medium or carrier” that band limits signals, it’s the transducers and what is producing the sound in the first place. EG. Only around 4% of the sound produced by a violin is higher than 20kHz, so a violin is effectively band-limiting itself, before we even consider the band-limiting of air, the transducers or the filters in ADCs or DACs. Your premise of non-band limited sounds cannot exist.

Essentially I'm arguing one does have to confirm for themselves what a human body's terminal velocity is, if philosophically you want to claim that as knowledge and not a "belief" but this is mostly tangential.

As science can be defined as the hierarchical organisation of knowledge and science (due to Newton, et al) can calculate “what a human body’s terminal velocity is”, then No, one does not “have to confirm for themselves” by jumping off a skyscraper in order to claim that as knowledge! You indeed do not seem to know what science is!

As it pertains to the question at hand, do DACs color sound, then if one believes that human hearing is fully understood then yes all the arguments are likely a 100% valid …

In the first instance, the “question at hand” has nothing whatsoever to do with hearing, let alone how much is understood about it! DACs do not produce any sound, they produce an analogue signal, so the question in the first instance is: Do DACs alter the analogue signal enough to cause a colouration of the output of the transducers (which is where the sound IS produced)? Sound itself has limitations, let alone the limitations of the amp and transducers downstream of a DAC. And, given a reasonable (safe) listening level, then many of the DAC artefacts which appear to concern audiophiles (and reportedly cause sound colouration) cannot in fact exist as sound. Therefore, human hearing (or our knowledge of it) cannot have anything whatsoever to do with any of this, because obviously we cannot hear a sound (or colouration of the sound) that does not exist.

Secondly, your assertion is a fallacy anyway, because we do not need to know everything in order to know anything! We do not and possibly will never “fully understand” everything about mathematics for example, does that mean the claim that 1+1=2 cannot be “100% valid” and is nothing more than some people’s belief? … Likewise, we do not need to fully understand human hearing, we only need to know the human hearing thresholds, which we’ve been studying for over 150 years and established nearly a century ago!

G

 

[CopperFox]

I stopped reading at "uncolored sound". There is no such thing as neutral or uncolored sound. It is 100% myth and something no one has ever heard.

A dac will have hundreds of components in its circuit, and every single one of them will tilt signal/power vs frequency in some way and/or have signal/power loss of some kind. This includes all of their attemps at correction systems.

 

[bigshot]

It’s easy to determine that uncolored sound exists. All you need is a null test.

 

[CopperFox]

bull.

Edit because of auto-censorship:

Taurine excrement.

 

[castleofargh]

bull.

Edit because of auto-censorship:

Taurine excrement.

Thank you for that elite example on how to argue and substantiate a point of view.

 

[gregorio]

There is no such thing as neutral or uncolored sound. It is 100% myth and something no one has ever heard.

A dac will have hundreds of components in its circuit, and every single one of them will tilt signal/power vs frequency in some way and/or have signal/power loss of some kind.

You seem to be confusing two different things! A DAC does have many components and some of them will have some effect on the output. However, the output of a DAC is NOT sound, it’s an electrical (analogue) signal, so the question is then: Will the components in the DACs’ circuits “tilt signal power vs frequency” or “signal/power loss of some kind” sufficiently to actually affect the sound output by the speakers or HPs and if so, is it by enough to actually be audible? The answer with pretty much any DAC is “no”, provided is it isn’t broken and that it’s output is volume matched!

bull.

Please explain how a null test does not determine if the output is coloured, it’s simple math!

G

 

[CopperFox]

Thank you for that elite example on how to argue and substantiate a point of view.

I don't have any need to raise this discussion's level from what it is. Why should I?

 

[CopperFox]

You seem to be confusing two different things! A DAC does have many components and some of them will have some effect on the output. However, the output of a DAC is NOT sound, it’s an electrical (analogue) signal, so the question is then: Will the components in the DACs’ circuits “tilt signal power vs frequency” or “signal/power loss of some kind” sufficiently to actually affect the sound output by the speakers or HPs and if so, is it by enough to actually be audible? The answer with pretty much any DAC is “no”, provided is it isn’t broken and that it’s output is volume matched!

This "no" simply is incorrect.

Please explain how a null test does not determine if the output is coloured, it’s simple math!

G

Why? There was no explanation presented why it would do that.

 

[gregorio]

I don't have any need to raise this discussion's level from what it is. Why should I?

Correct you don’t but then you’re not, you’re in fact doing the exact opposite. “This discussion’s level” is based on science/facts which can all be reliably substantiated, making false assertions and stating “bull” is obviously not substantiated at all and therefore your posts are way below “this discussion’s level” and way below what is acceptable in a science discussion forum, that’s why!

This "no" simply is incorrect.

Wrong, if colourations are at say -95dB and peak playback level is say 85dBSPL, how is that colouration audible?

Why? There was no explanation presented why it would do that.

There’s no explanation of why a null test “would do” what a null test does? Don’t you even know what a null test is?

G

 

[CopperFox]

“This discussion’s level” is based on science/facts which can all be reliably substantiated

The above statement does not correspond to reality.

There’s no explanation of why a null test “would do” what a null test does? Don’t you even know what a null test is?

G

A null test is something that can be used to determine if one or more pieces of variously flawed audio-related equipment are vaguely similar in some aspects of their operation.

What it has zero logical connection to is a massively complex and abstract concept of some sound being "neutral" or "uncolored" or such things existing at all.

 

[gregorio]

The above statement does not correspond to reality.

It corresponds to the facts/science. If that doesn’t correspond to your reality, then that’s your problem and you’re obviously posting in the wrong place given the name of this forum!

A null test is something that can be used to determine if one or more pieces of variously flawed audio-related equipment are vaguely similar in some aspects of their operation.

So that’s a “no” then, you do not know what a null test is. Calling “bull” on a test that you don’t even know what it is or does, is nowhere near the level of this discussion or pretty much any discussion anywhere, let alone in a science discussion forum, in fact it could hardly be a lower level!

What it has zero logical connection to is a massively complex and abstract concept of some sound being "neutral" or "uncolored" or such things existing at all.

Uncoloured is neither massively complex nor abstract, just making up BS in a science discussion forum could not be less acceptable or of a lower level! If the output signal is the same as the input signal then the device is “uncoloured” and if there is some difference but it’s below audibility then it’s audibly “uncoloured”. It doesn’t get much simpler than that and if something so simple is “massively complex” for you, then obviously you have problem but you shouldn’t be demonstrating that publicly on a science discussion forum.

Additionally, as a null test specifically detects if an output signal is the same as the input signal (or the amount of difference if not) then it has complete “logical connection” to the issue of coloured/uncoloured. Don’t you know what “logical” means either? Jeez!

G

 

[CopperFox]

It corresponds to the facts/science. If that doesn’t correspond to your reality, then that’s your problem and you’re obviously posting in the wrong place given the name of this forum!

Wonderful argumentation.

So that’s a “no” then, you do not know what a null test is. Calling “bull” on a test that you don’t even know what it is or does, is nowhere near the level of this discussion or pretty much any discussion anywhere, let alone in a science discussion forum, in fact it could hardly be a lower level!

Uncoloured is neither massively complex nor abstract, just making up BS in a science discussion forum could not be less acceptable or of a lower level! If the output signal is the same as the input signal then the device is “uncoloured” and if there is some difference but it’s below audibility then it’s audibly “uncoloured”. It doesn’t get much simpler than that and if something so simple is “massively complex” for you, then obviously you have problem but you shouldn’t be demonstrating that publicly on a science discussion forum.

Additionally, as a null test specifically detects if an output signal is the same as the input signal (or the amount of difference if not) then it has complete “logical connection” to the issue of coloured/uncoloured. Don’t you know what “logical” means either? Jeez!

G

Wonderful projection. Well I guess the ignore function exists for a reason.

 

[gregorio]

Wonderful argumentation.

As “wonderful” as calling “bull” on a test when you don’t even know what the test does?! The only argumentation you’ve presented is ad hominems, the exact opposite of “wonderful argumentation”.

Wonderful projection. Well I guess the ignore function exists for a reason.

You’re the one who falsely projected a very simple concept as “massively complex”. And if that’s not enough, you even guess wrong, the ignore function obviously doesn’t work unless it is engaged first. It is unacceptable for you to post BS to a science discussion forum in the first place, regardless of whether it can be ignored! Maybe this is another example of something too “massively complex” for you?

G