Understanding the Role of DACs in a Simple Audio Setup

[Dragonmilenario]

Hi everyone,

I’ve been diving into the world of DACs and wanted to share some insights and get your thoughts. Starting with a basic setup of PC -> DAC -> AMP -> Headphones, it’s clear that we need a DAC that offers:

1. Uncolored Sound: The DAC should not add any coloration to the sound.
2. No Audible Distortion: It should have minimal to no audible distortion.
3. Linear Output Across All Frequencies: The output should be linear across the entire frequency range.
4. Balanced Output with 4Vrms: This ensures a strong and clean signal.

These specifications are measurable in lab conditions and can be clearly defined. For instance, the Topping D10 Balanced, which costs just over $100, meets these criteria excellently.

Despite this, I still see threads and messages where people claim that a DAC sounds “good” or “bad,” and discussions about DACs costing thousands of dollars that, in some cases, measure worse electronically than the Topping D10 Balanced.

So, my question is: Do audiophiles prefer a DAC that adds color to the sound, or do they seek a perfectly linear and accurate output? From a common-sense perspective, I believe a DAC should be as transparent as possible, leaving any desired coloration to the AMP.

Am I wrong in thinking this way?

I don’t consider myself an expert and could be completely mistaken. If so, I’d love to hear your opinions!

 

[Daiyama]

Your four bullet points are the basics for a DAC, but you need to listen more closely!
Try to hear the tiny, tiny, tiniest micro details, the black background or the space between the notes!

That’s where audiophile DACs excel! This is what makes them worth to spend kilo bucks.

 

[Dragonmilenario]

Let’s approach this topic with respect and information. Let’s avoid resorting to easy mockery.

Thank you.

 

[SpaceOperaVillain]

Quite literally, I think the only statement we can make with 100% certainty is that perception is everything. Confirmation bias and an appetite for colored sound are both rampant in this hobby. But I think this take, while not wrong, is a bit cynical.

The main problem, as I view it, is trying to get everybody to ignore the hump of quasi-science that attempts to explain away sound differences we each perceive differently with regards to timbre. Because harmonic overtones in the mids and treble are reliant on clean playback in the 10 kHz+ range in the spectrum, preference, tolerance, and age-related hearing loss all factor heavily in how timbre is perceived.

For example, clearly there are differences between planar, balanced-armature, and dynamic drivers in the onslaught of IEMs now flooding the market and the hobby itself. Each of these have a distinctive characteristic which shouldn't be any more controversial than declaring that middle C sounds different on a baritone versus a clarinet. When it comes to DACs, the processing has a lot of latitude when converting the source into a signal, which we perceive as attack, sustain, and delay on individual notes - even if that signal is distortion free. The analogue here is the difference between playing a scale on a piano with the sustain pedal depressed, and then playing it without. And how hard are the key strikes as the scale is played? And what wood species is the sound board? The nuance is similar between DACs, but the changes are so absurdly small that detecting it with our ears is all but impossible for the majority of music lovers, without even starting to question the abilities of self-proclaimed audiophiles.

Which begs the question when moving on from those obvious examples, at exactly what point our ability to detect changes in timbre, speed, and polytonal differences approaches zero is necessarily unique to our age, our anatomy, and how developed our sense of hearing may or may not be. Listening, as a neurologically supported skill, can spot tonal and timbre differences that an untrained listener cannot. Audiophiles, even the ones that believe in voodoo, still have a highly developed sense of hearing, and it is wading through these murky waters that we find differences between amplifiers, pre-amplifiers, tubes vs. discrete / solid-state, DACs, and then of course, cables.

It could be worse: you could be a guitarist chasing tone! If you're both a guitarist and an audiophile they your life is a purgatory and you're cry-laughing at this already.

So rather than attempting to rush to jump-to-conclusion inferences with basic physics, or reach for some kind of sci-fi nonsense invoking quantum voodoo to impress readers, I'm of the opinion that first and foremost we have to agree that there is a matrix for each listener that can account for:

* physical abilities, mainly limitations wrt to hearing loss
* neural abilities (eg the trained or golden ear, sense of pitch etc),
* the perception of one's abilities (consider tinnitus itself, which is perceptible noise that does not exist. also, brain burn-in is real.)

How many scalars exist in these three categories, and what kind of combinatorial complexity arises from it? Even a basic grasp on linear algebra should have you reeling. So these arguments about DACs and cables that play out, ad nauseum, have little to do with the hardware and everything to do with perception. In my view, the kind of simplistic measurements made by crude electronic equipment on an improvised lab bench we often see on forums is close to meaningless.

At best, we could approach some kind of mean calculation for what is deemed an average pair of ears for an average audiophile's background and sensibilities. We'd still be left with a high number of wild variables. This would be an interesting experiment for machine learning, sans voodoo and absent of commercial pressure and sales promises, but I digress.

so just ... yes - the cables do matter, and those DACs do sound different. Up to a point. We just can't be sure of where that point actually is when talking among ourselves.

edit, after getting carried away with defining the Neutral Zone:

I believe a DAC should be as transparent as possible, leaving any desired coloration to the AMP.

Am I wrong in thinking this way?

I agree with that assessment, but only you can answer that for yourself. If you want neutral, transparent sound, then sure. But look at this new wave of R2R DACs coming onto the market. Any device that warms up the mids and attenuates high treble is going to be a crowd-pleaser, regardless of the method it uses to achieve that, or what place in the audio chain it actually sits. There's no wrong and no right as the base case. The fun part of this hobby is exploring the combinations.

 

[Daiyama]

Let’s approach this topic with respect and information. Let’s avoid resorting to easy mockery.

Thank you.

Ok, direct answer.
In my opinion you are wrong, why leave colouring of the sound only to Amps and other stuff and leave DACs out? There is no reason for that. If you like colouring your sound use all options and maybe coloured DAC and Amp show great synergy (btw. best thing for colouring sound is an EQ).

On the other hand if you want to achieve transparency throughout your whole chain then certainly it is important to have a transparent DAC (and non colouring Amp).

 

[2leftears]

Hi everyone,

I’ve been diving into the world of DACs and wanted to share some insights and get your thoughts. Starting with a basic setup of PC -> DAC -> AMP -> Headphones, it’s clear that we need a DAC that offers:

1. Uncolored Sound: The DAC should not add any coloration to the sound.
2. No Audible Distortion: It should have minimal to no audible distortion.
3. Linear Output Across All Frequencies: The output should be linear across the entire frequency range.
4. Balanced Output with 4Vrms: This ensures a strong and clean signal.

These specifications are measurable in lab conditions and can be clearly defined. For instance, the Topping D10 Balanced, which costs just over $100, meets these criteria excellently.

Despite this, I still see threads and messages where people claim that a DAC sounds “good” or “bad,” and discussions about DACs costing thousands of dollars that, in some cases, measure worse electronically than the Topping D10 Balanced.

So, my question is: Do audiophiles prefer a DAC that adds color to the sound, or do they seek a perfectly linear and accurate output? From a common-sense perspective, I believe a DAC should be as transparent as possible, leaving any desired coloration to the AMP.

Am I wrong in thinking this way?

I don’t consider myself an expert and could be completely mistaken. If so, I’d love to hear your opinions!

Depending on your objectives, there is no reason why a DAC should be transparent and any colouring should be the domain of the amp.

On grounds of practicality in an ideal world I would suggest any colouration (if desired!) be done in a component that offers easy control to manipulate it depending on material/mood etc, rather than hard-baking it into any audio components. (The world is not ideal and re. e.g. speakers a certain degree of baked-in colouration has to be accepted when purchasing a preferred pair on ground of limits of physics and engineering. Room acoustics are also not fully controllable in practical terms).

Your question re. audiophiles cannot be answered really, other than to say that there are perhaps as many opinions & objectives as there are audiophiles. Suffice to say that some audiophiles are looking for transparency in as much of the audio chain as possible, whereas others prefer some colouration to their personal taste and may be happy to either select a colouring amp or a colouring DAC.

In the end, it is a personal hobby and whilst some ways of enjoying the hobby seem to make more sense then others, there is no "right" or "wrong" way to enjoy the hobby. The problems arise when some people (sales people and marketeers included!) throw science into the wind and try to convince others that all sorts of perceived effects are real (and therefore merit investment of time and money) when science provides hard evidence to the contrary. That may be tolerable on other forums but on the Sound Science forum this will obviously get challenged.

Re. the claims you read on "good" vs. "bad" sounding DACs: those qualifications are almost always exaggerated. The (early) days of easily noticeable differences between (then) expensive vs cheap DACs are long gone. In a feat of both engineering and manufacturing abilities catching up with much older established theory, DACs have reached a point where even relatively modestly priced ones can be audibly transparent.

But he psychology of very expensive audio components "inevitably" sounding much better than cheap ones is hard to eradicate (marketing is at work here also), leading to expectation bias and easily biased perception in an uncontrolled test environment. This is not to say that bad sounding DACs don't exist, but the likelihood of easily finding a significant improvement in sound quality simply through buying a much more expensive DAC have diminished rapidly over the years.

 

[castleofargh]

Hi everyone,

I’ve been diving into the world of DACs and wanted to share some insights and get your thoughts. Starting with a basic setup of PC -> DAC -> AMP -> Headphones, it’s clear that we need a DAC that offers:

1. Uncolored Sound: The DAC should not add any coloration to the sound.
2. No Audible Distortion: It should have minimal to no audible distortion.
3. Linear Output Across All Frequencies: The output should be linear across the entire frequency range.
4. Balanced Output with 4Vrms: This ensures a strong and clean signal.

These specifications are measurable in lab conditions and can be clearly defined. For instance, the Topping D10 Balanced, which costs just over $100, meets these criteria excellently.

Despite this, I still see threads and messages where people claim that a DAC sounds “good” or “bad,” and discussions about DACs costing thousands of dollars that, in some cases, measure worse electronically than the Topping D10 Balanced.

So, my question is: Do audiophiles prefer a DAC that adds color to the sound, or do they seek a perfectly linear and accurate output? From a common-sense perspective, I believe a DAC should be as transparent as possible, leaving any desired coloration to the AMP.

Am I wrong in thinking this way?

I don’t consider myself an expert and could be completely mistaken. If so, I’d love to hear your opinions!

2 and 3 make 1
I'm not sure where you got 4? Is it 4V because you expect balanced output to double the old CD player standard of 2V(that engineers followed when they felt like it)? Or does it just happen that you have a DAC you like for whatever reason, and it outputs 4V FS? Even the balanced choice is arguable. I see no issue with that, but I also know that there is a vast offering of single ended outputs that are more than clean enough.

When the source is the computer, I agree that there is probably not much need for a DAC with EQ and extra DSP features. For those with more or other sources, having extra features within the DAC is great IMO. We can always have some extra box like miniDSP devices, but it adds an extra AD/DA to the chain.
As for DAC that would somehow manage to have very noticeably inferior fidelity on purpose(NOS DAC with massive treble roll off or absurd amount of aliasing, or some weirdo idea to introduce tubes in a DAC ), I have two thoughts:
1/ this is freaking moronic.
2/ people can and should do what they want. I'm nobody's mother.

As for people saying DAC X sounds good and DAC Y doesn't, my opinion is to ignore them. For starters, most don't qualify to passing a judgement on sound. They don't measure and almost never actually discuss fidelity(even when many pretend to). As for listening impressions, they're almost exclusively sighted(and if the 2 DACs don't output the same voltage, almost nobody will match the output within 0.1dB for a more relevant listening impression). @2leftears says the comments on good and bad are almost always exaggerated, I'd argue that exaggerated is overly optimistic and sort of the best scenario we can hope for in this case. Because to exaggerate the right differences, you need the right initial impression, and that's rare with DACs and play pretend testing.

 

[Dragonmilenario]

Thank you all for your responses! If I understood correctly, buying the Topping D1 Balanced ($130) doesn’t make a difference compared to the E70 Velvet ($350) when connecting to the L70 amplifier? Both measure very similarly in terms of linearity and zero distortion. Is it just for aesthetic reasons (they have the same style) and connection possibilities, but the sound is expected to be the same?

 

[bigshot]

Sound differences are most pronounced with different models of headphones. To have your headphones presented at their best, they may need to be amped properly. DACs (aside from NOS DACs and ones with non standard filers) are calibrated to all perform to perfect digital specs. If they don’t sound the same, there is something wrong with one of them.

 

[gregorio]

So, my question is: Do audiophiles prefer a DAC that adds color to the sound, or do they seek a perfectly linear and accurate output?

Generally audiophiles prefer a DAC that costs more, has a better brand name and/or high praise from reviewers. Colouration doesn’t really come into it because extremely few DACs audibly colour the sound. The only exception being some NOS DACS and even then, only under some conditions.

Try to hear the tiny, tiny, tiniest micro details, the black background or the space between the notes!
That’s where audiophile DACs excel! This is what makes them worth to spend kilo bucks.

No, that’s not where audiophile DACs excel. Audiophile DACs perform “tiny, tiny, tiniest micro details” the same or sometimes worse than pretty much every other DAC, including many very cheap ones. What makes them worth (to many audiophiles) spending kilo bucks is their visual appearance, brand name and a false belief based on incentivised reviews and misleading marketing that they are better fidelity.

Quite literally, I think the only statement we can make with 100% certainty is that perception is everything.

That’s not really true, it depends on what we’re claiming. If for example, we’re making claims about what we personally like or prefer, then your statement is absolutely correct. But if we’re making claims about the actual performance of a DAC then perception is relatively little, the actual objective properties of the analogue output is everything and that can only accurately be ascertained by measuring them.

[1] Listening, as a neurologically supported skill, can spot tonal and timbre differences that an untrained listener cannot. [2) Audiophiles, even the ones that believe in voodoo, still have a highly developed sense of hearing, and it is wading through these murky waters that we find differences between amplifiers, pre-amplifiers, tubes vs. discrete / solid-state, DACs, and then of course, cables.

1. Yes, that’s been repeatedly objectively demonstrated for centuries.
2. Hang on, you seem to be confusing hearing with listening. Listening is a skill that can be trained as you stated, hearing can’t be trained and as audiophiles tend to be older/middle aged adults, their hearing abilities are actually lower than many others (children or significantly younger adults). However one big difference with audiophiles is that they’re exposed to far greater amounts of misleading marketing and almost exclusively that is where we find differences between amplifiers, DACs and cables!

So these arguments about DACs and cables that play out, ad nauseum, have little to do with the hardware and everything to do with perception.

Agreed, which is why they should stick to making claims only about their perception and not about the actual hardware performance. This is why we get the vast majority of arguments in the first place, they don’t seem to realise that hardware performance and their personal perception are two entirely different things that are measured entirely differently.

In my view, the kind of simplistic measurements made by crude electronic equipment on an improvised lab bench we often see on forums is close to meaningless.

Sometimes that can give meaningful information, it depends what we’re measuring and of course quite a lot of the measurements we see on forums is cited from measurements using highly sophisticated measuring equipment.

so just ... yes - the cables do matter, and those DACs do sound different. Up to a point. We just can't be sure of where that point actually is when talking among ourselves.

Firstly, do those cables and DACs actually sound different, are the differences of sufficient magnitude to even be resolved into sound? If the answer is “no”, they’re too small to be resolved into sound, then obviously they cannot possibly sound different. And, a surprising number of the differences we find with DACs and cables fall into this category! Even if they can be resolved into sound, we have around 130 years worth of objective data about the “point” at which some sound property/difference is audible, so typically we have at least a reasonable, if not a very good idea of “where that point actually is”. This is assuming of course we’re aware of and understand that objective data, which very few audiophiles are, hence why we only find these arguments/assertions in the audiophile corner of the audio world.

I agree with that assessment, but only you can answer that for yourself.

Again, not really. We can ascertain transparency/fidelity far more accurately and precisely using objective measurements.

In my opinion you are wrong, why leave colouring of the sound only to Amps and other stuff and leave DACs out? There is no reason for that.

There is a reason for that, the vast majority of DACs do not audibly colour the sound and therefore obviously have to be left out of “colouring the sound”. Indeed, that’s true of the vast majority amps as well, assuming they’re the right amp for the job and don’t deliberately strive for poor fidelity.

If you like colouring your sound use all options and maybe coloured DAC and Amp show great synergy (btw. best thing for colouring sound is an EQ).

Surely you wouldn’t “use all options”, surely you’d only use those options that actually audibly colour the sound? In which case, EQ is pretty much the only option (with the tiny number of exceptions already mentioned).

G

 

[Sonic Defender]

Ok, direct answer.
In my opinion you are wrong, why leave colouring of the sound only to Amps and other stuff and leave DACs out? There is no reason for that. If you like colouring your sound use all options and maybe coloured DAC and Amp show great synergy (btw. best thing for colouring sound is an EQ).

On the other hand if you want to achieve transparency throughout your whole chain then certainly it is important to have a transparent DAC (and non colouring Amp).

Really? That doesn't seem at all like a good idea. First off, ideally no gear should colour the sound, period. And absolutely under no circumstances should a DAC do that.

If people want to colour the sound you do that with DSP.

 

[Daiyama]

My gear (any gear) can colour my sound, period.

 

[bigshot]

If a source colors the sound, then every source will sound different. If you have a preferred sound, you’ll have to EQ or apply a DSP individually for each source. That can lead to unnecessary complexity. It’s better to have all the sources calibrated to the same sound and apply coloration at the amplification stage. That way you just have one coloration setting instead a different one for each source.

 

[bagwell359]

The desire for .00000001 distortion and perfect linearity seems like an obvious goal.

But hearing isn't totally objective. Trained hearing is more objective, but still is subjective. Add in that recording techniques/technology isn't perfect, and hearing differes between people.

It's true that the audiophile side of things can and does contain all sorts of biases and nonsense - some of it very costly.

I couldn't stand digital for many years at its outset (from TELARC vinyl up). But I heard digital sound like music when I first heard the Krell KPS 20i, but it was very expensive, then some 10 years later when I heard the Schiit Gungnir A1 at a much lower cost. Ran 1000 cuts on my very good vinyl rig vs the "Gumby". It was essentially a draw, and I could and did make lots of money on selling my vinyl stuff, with zero regrets. I'm sure neither of those two DAC's equal the specs of your $100 DAC - but do you listen to stats or music? Building a system that is accurate and enjoyable isn't just a checklist.

To avoid the placebo/bias inherent in all of us: do double blind testing will rule out bias of cost, perceived brand quality, looks, and all the rest. And remember transducers is where the differences are crucial and easy to differentitate, and what should be chosen first, and tailor the rest to fit the budget and any special requirements of the transducers. Amps, pre-amps, cables, and all the rest tend to be the same/similar all things being equal.

 

[Daiyama]

If a source colors the sound, then every source will sound different. If you have a preferred sound, you’ll have to EQ or apply a DSP individually for each source. That can lead to unnecessary complexity. It’s better to have all the sources calibrated to the same sound and apply coloration at the amplification stage. That way you just have one coloration setting instead a different one for each source.

You are absolutely right, when your aim is a reference sound. Then you should keep it simple and stick to defined parameters.
That is no longer my aim. I tried this for 25 years and never achieved it, for a reference sound you need a reference grade room. I am not able to do that, so results were more and more not to my satisfaction and I got less and less interested in listening to music.
Now I do not care about reference sound anymore. I only listen to gear I enjoy wether it colours the sound or not.