• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Serious Question: How can DAC's have a SOUND SIGNATURE if they measure as transparent? Are that many confused?

ahofer

Major Contributor
Forum Donor
Joined
Jun 3, 2019
Messages
4,952
Likes
8,698
Location
New York City
I understood that the Topping E30 ii is a bit flat in sound and bright. I would like something more musical.
Pleasant sound is one that pleases the listener.
Please don't ask me to prove it scientifically.

What is 'more musical'? Is it a different sound due to the DAC? If you think it is, I have a canned response:

Welcome to ASR! You've made an unsupported assertion or a scientifically implausible claim that will cause most people in this science-oriented forum to react with skepticism (or scepticism if they are in the U.K.). Please don't take the reactions as overtly hostile - most of us are just frustrated with the many newcomers who have clearly come here just to "troll". Please do engage with the membership to find an objective, controlled method to support or discard your hypothesis. Our membership includes recovering subjectivists, many engineers/scientists, and several famous figures in the world of audio engineering research. Generally, they can cite scientific, controlled research to support their views. Most believe in the fallibility of human sighted judgement, and think blind testing and measurements are critical ingredients for assessing equipment contributions to sound quality. We'd love to have you, but if all you want is a) to fight or b) to have others cheerlead for your subjective views or anecdotal evidence, I'd suggest you will be happier elsewhere.
 

arnea

Member
Joined
Jan 7, 2023
Messages
9
Likes
1
@Thorsten Loesch - interesting reading, thank you very much for sharing this.

I have a Zen Dac v2 that is driving directly the Edition XS headphones. It's my first HiFi setup - I didn't know anything when I bought this from local shop. Now I've learned that Zen Dac doesn't have enough power to drive the very low impedance phones as Edition XS. I was actually planning to get rid of Zen Dac and get either amp+dac stack or combined unit. And yes, I was shopping by numbers and charts.

If I understood you correctly you are saying that the Zen Dac v2 is actually a better DAC than can be said by the measurement results? So perhaps I should keep it and just add a suitable amp?
 

ahofer

Major Contributor
Forum Donor
Joined
Jun 3, 2019
Messages
4,952
Likes
8,698
Location
New York City
If I understood you correctly you are saying that the Zen Dac v2 is actually a better DAC than can be said by the measurement results?
Oy vey.
 

Thorsten Loesch

Senior Member
Joined
Dec 20, 2022
Messages
460
Likes
530
Location
Germany, now South East Asia (not China or SAR's)
I returned a Topping E50 that plays intermittently via toslink. I liked the sound. Now I have 3 DACs in mind. Ifi zen one signature/ Cambridge magic 200m/ ProJect pre box S2. I have Marantz pm 6006+Dali Oberon 3. I'm tempted by the price of Ifi but I don't know how it sounds in my system. I used to have Cambridge magic 100. If anyone can help me. Thanks.

If you really liked the sound of the Topping, I doubt you will like the iFi.

Thor
 

Thorsten Loesch

Senior Member
Joined
Dec 20, 2022
Messages
460
Likes
530
Location
Germany, now South East Asia (not China or SAR's)
I have a Zen Dac v2 that is driving directly the Edition XS headphones. It's my first HiFi setup - I didn't know anything when I bought this from local shop. Now I've learned that Zen Dac doesn't have enough power to drive the very low impedance phones as Edition XS.

I had a look. The single-ended output should provide around 1.4V into your headphone. It seems that the headphone should produce 109dB SPL with 1V signal and 112dB SPL with 1.4V. This is quite loud.

So I think the Zen DAC can "drive" your headphones to high levels.

I was actually planning to get rid of Zen Dac and get either amp+dac stack or combined unit. And yes, I was shopping by numbers and charts.

Before you accept someone's numbers as evidence for something that matters to you listen to music, you should ask them for solid scientific evidence that their numbers somehow reliably (as in better numbers = better sound) correlate with what you seek.

If I understood you correctly you are saying that the Zen Dac v2 is actually a better DAC than can be said by the measurement results? So perhaps I should keep it and just add a suitable amp?

I am not saying any such thing.

I am saying that I designed the Zen DAC (V1) upon which the Zen DAC V2 is largely based to provide "good sound".

And I designed it to measure adequately well, so that normally it's noise and distortion are not the limiting factor in an audio system.

I am implying in this that there is actually no reliable correlation between "good sound" and "good measurements", but I am not explicitly saying this.

Some other DAC's are also designed for "good sound" as priority (they often have unhappy panthers in Amirs tests) and tend to be very expensive.

Most cheaper DAC's from the far east tend now to be designed to get happy panthers. If that makes you happy when listening, so be it, but it cannot be reliably be expected.

If you ask me for buying advise?

First, understand your "taste". Understand what kind of "sound" makes you want to listen to more more music and gives you greater pleasure listening, improves your moode and get's you an emotional connection with the music.

This is like finding which wine you like. To me most white wines are the same, a touch sour, drinkable but expensive ones are still a little sour. Now there are white wines like Tokaji, sauterne, Spätlese, Icewine etc. that are different and I love these and yes, I taste differences between them and have preferences.

But Saugvignon Blanc? Chardonnay? Cheap? Expensive?

On the other hand among red wines, while I love cheap easy drinking sweetish reds for casual drinking I get different Red's just fione and some of the really good ones I had a chance to try (I said"good", not expensive) were spectacular experiences. If I could afford them I'd drink a glass or two everyevening.

Then find gear that suits your taste in sound. Maybe it is exactly the sound produced by the highest rated products here? Maybe (like to me) these are more to you like dry white wine to me. It's ok, but I would not drink it if I have a different choice.

Go to concerts of music you like. If possible, acoustic/unplugged is what I recommend and what informs my consideration of "good sound".

So go to shows, shops etc. Ignore anything sales people say. They are experts in misdirection, it's their job.

Ignore what reviewers say. Ignore what experts on the internet say. Ignore what I say about products. Just because I like what I created doesn't make it "good" to you.

Bring your own music, your own headphones. Match levels when listening (use a multimeter). Relatively small differences in loudness can give a strong preference for the louder item that disappears or even reverses with level matching.

Listen without pressure and try not get to be amazed and befuddled by sonic fireworks that impress on first listen but are tiring after one hour. Check how you react emotionally as well as what you perceive as "sound".

Because you need to try a lot of stuff, buy second hand and buy cheap. Haggle like you would with an armenian rug seller over price. If you sell, try to get a lot of your money back, the less you pay to start with, the less you loose.

This may not be what you want to hear. If so, just ignore what I say.

I'd like to leave you (and others reading) with the following:


"The high-fidelity initiate, bewitched, bothered, and thoroughly confused by the staggering selection of components he must choose from, often turns to a high-fidelity expert to assist him in assembling his dream system. The expert may be a local consultant, a dealer, or a magazine that the prospective buyer trusts as a source of accurate, down-to-ear information.

...

The best that any expert can do is to lead you to components that are intrinsically excellent. You will still have to make up your own mind about such matters as cost and appearance and flexibility, and you should try out a few different loudspeakers in your home to find out which ones suit your acoustical environment and your taste in reproduced sound. The expert cannot, and will not if he has any sense, choose the components for you, because your ear is the final judge in the last analysis. If no combination of really good components sounds good to you, then you probably don't really want high fidelity, and can forget all about the expert opinions. They don't agree anyway."

And my own sign off using a quote from Anthony De Mello "On waking up":

"Waking up is unpleasant, you know. You are nice and comfortable in bed. It is irritating to be woken up.

That’s the reason the wise guru will not attempt to wake people up.

I hope I’m going to be wise here and make no attempt whatsoever to wake you up if you are asleep.

It is really none of my business, even though I say to you at times, "Wake up!"

My business is to do my thing, to dance my dance. If you profit from it fine; if you don’t, too bad!

As the Arabs say, "The nature of rain is the same, but it makes thorns grow in the marshes and flowers in the gardens."

Thor
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,051
Likes
12,150
Location
London

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,051
Likes
12,150
Location
London
Does it have that quality that is yet undiscovered( and of course unmeasurable) but will be revealed one day long after we are all dead?
Keith
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Does it have that quality that is yet undiscovered( and of course unmeasurable) but will be revealed one day long after we are all dead?
Keith
No it isn't undiscovered and not unmeasurable, just not revealed. You are supposed to read between the lines sort of. Oh and listen with your emotions and your heart.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
I had a look. The single-ended output should provide around 1.4V into your headphone. It seems that the headphone should produce 109dB SPL with 1V signal and 112dB SPL with 1.4V. This is quite loud.

A 112dB continuous tone is quite loud. 112dBA noise level is very loud.
Music signal consists of an addition of all amplitudes.
So when a bass 'impulse' + rest of the signal is to be reproduced, certainly with Harman EQ in the bass, 112dB total peak SPL is not quite loud at all as it consists of short peaks at various frequencies.
All very dependent on the DR of the recording of course. With highly compressed pop/rock music it is unlikely one wants to listen to 120dB peaks :)

It'll be fine for 'sensible' listening levels even up to 'comfortable loud' levels.
The problem is when people (shortly) want to play more impressively loud. Then a headphone should be able to reach 120dB SPL peaks (not average) cleanly.
 
Last edited:

Thorsten Loesch

Senior Member
Joined
Dec 20, 2022
Messages
460
Likes
530
Location
Germany, now South East Asia (not China or SAR's)
⬜No it has that quality where both DACs have inaudible distortion, but the iFi has a little more and therefore sounds more pleasing in terms of preference. I'm sure Thorsten Loesch will say I've misunderstood him and yet won't explain how this is so.

You did misunderstand and on purpose.

You believe in one thing that my personal experience over decades tells me is wrong. And I am sick an tired of the arguments going nowhere.

My point is simply that beyond a certain point improving traditional measurements, specifically THD&N (aka SINAD), SNR/DNR, Jitter and frequency response stop being reliable predictors of subjectively perceived fidelity.

What this means is that if we take a given high quality, high resolution recording of music file and a playback system substantially free from noise and distortion etc. and in isolation add differing amounts of HD, Noise etc.

Very large amounts (relative to zero) will materially impair fidelity.

However a point comes where the stimulus is no longer detectable as impairment of fidelity.

At this point further reductions reductions of the stimulus that leads to impairment of fidelity in larger quantities no longer improve fidelity.

This has not stopped marketers and salespeople in claiming audible improvements anyway or suggesting that even without such audible improvements making these numbers better than necessary is a "good thing" and one should buy the equipment with the best numbers, however the lack of logic coherence of such a position is patently obvious. I tend to ask who pays for such a position being broadcast whenever I encounter it.

The quantities have a long history of being reported and while passing decades have led to refinement of thresholds and introduced new fidelity impairments into the definitions, generally speaking for the core measurements listed above have been established, documented and validated many times since the 1950's (1990's for Jitter).

It should also be noted that so far no scientific evidence has been presented to the effect that these specific fidelity impairments are the only ones possible and that no other mechanism of fidelity impairments exist. In fact, over the decades since "fidelity limits" were first established in the 1950's we have seen the introduction of a number of previously unobserved and unknown fidelity impairments and their quantification, e.g. Group Delay, Phase Response, Slew-induced Distortion etc. et al, Jitter and it is likely that more such potential fidelity impairments remain to be discovered.

I would indeed suggest that this position in itself is hardly controversial and that no scientific evidence to contrary has ever been produced. If I am in fact mistaken about the scientific evidence, I'd like see it presented, reference will suffice, I have the required access.

Now while up to now I see no controversy, I think that even though I feel that there should not be controversy on my next point, it is controversial it seems.

My next point is that even through a pair of Devices under Test (DUT) may both show objective performance with the quantities of the specified fidelity impairing stimuli (shall we call then distortions in short from here on?) below audibility; there can be a reliable preference (that is of statistically significant reliability) for listeners for one DUT over the other under double-blind conditions (but not ABX for obvious reasons, for starters ABX does not test preference).

I further posit that these results are an indication of more or less subtle fidelity impairments not currently covered by standardised and common objective tests and that reducing such posited fidelity impairments while possibly creating a worse but still inaudible level of fidelity impairments covered by common tests is the underlying reason for the observed preference of Item ◯ over Item ⬜ in our blind testing.

Thor
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
You did misunderstand and on purpose.
No I didn't.
You believe in one thing that my personal experience over decades tells me is wrong. And I am sick an tired of the arguments going nowhere.
That makes two of us at least.
My point is simply that beyond a certain point improving traditional measurements, specifically THD&N (aka SINAD), SNR/DNR, Jitter and frequency response stop being reliable predictors of subjectively perceived fidelity.

What this means is that if we take a given high quality, high resolution recording of music file and a playback system substantially free from noise and distortion etc. and in isolation add differing amounts of HD, Noise etc.

Very large amounts (relative to zero) will materially impair fidelity.

However a point comes where the stimulus is no longer detectable as impairment of fidelity.

At this point further reductions reductions of the stimulus that leads to impairment of fidelity in larger quantities no longer improve fidelity.
Agreed.
This has not stopped marketers and salespeople in claiming audible improvements anyway or suggesting that even without such audible improvements making these numbers better than necessary is a "good thing" and one should buy the equipment with the best numbers, however the lack of logic coherence of such a position is patently obvious. I tend to ask who pays for such a position being broadcast whenever I encounter it.
Agreed more or less.
The quantities have a long history of being reported and while passing decades have led to refinement of thresholds and introduced new fidelity impairments into the definitions, generally speaking for the core measurements listed above have been established, documented and validated many times since the 1950's (1990's for Jitter).

It should also be noted that so far no scientific evidence has been presented to the effect that these specific fidelity impairments are the only ones possible and that no other mechanism of fidelity impairments exist. In fact, over the decades since "fidelity limits" were first established in the 1950's we have seen the introduction of a number of previously unobserved and unknown fidelity impairments and their quantification, e.g. Group Delay, Phase Response, Slew-induced Distortion etc. et al, Jitter and it is likely that more such potential fidelity impairments remain to be discovered.
Do not agree with some of these. Especially the jitter bugaboo.
I would indeed suggest that this position in itself is hardly controversial and that no scientific evidence to contrary has ever been produced. If I am in fact mistaken about the scientific evidence, I'd like see it presented, reference will suffice, I have the required access.

Now while up to now I see no controversy, I think that even though I feel that there should not be controversy on my next point, it is controversial it seems.

My next point is that even through a pair of Devices under Test (DUT) may both show objective performance with the quantities of the specified fidelity impairing stimuli (shall we call then distortions in short from here on?) below audibility; there can be a reliable preference (that is of statistically significant reliability) for listeners for one DUT over the other under double-blind conditions (but not ABX for obvious reasons, for starters ABX does not test preference).

I further posit that these results are an indication of more or less subtle fidelity impairments not currently covered by standardised and common objective tests and that reducing such posited fidelity impairments while possibly creating a worse but still inaudible level of fidelity impairments covered by common tests is the underlying reason for the observed preference of Item ◯ over Item ⬜ in our blind testing.

Thor
And where is your evidence? Where are your references? Give us examples of DUTs with objective measures below audibility that somehow still show a reliable preference. Certainly the last paragraph holds, but you first have to supply evidence that what you are saying is so in the previous paragraph.
 

Thorsten Loesch

Senior Member
Joined
Dec 20, 2022
Messages
460
Likes
530
Location
Germany, now South East Asia (not China or SAR's)
A 112dB continuous tone is quite loud. 112dBA noise level is very loud.
Music signal consists of an addition of all amplitudes.

The THX recommendation is 85dB SPL @ -20dBFS. This is equal to 85dB average SPL with a programme having 20dB dynamic range giving 85dB SPL average and 105dB peak.

85dB average SPL is quite loud actually. Sound at this level should be kept to below 8 Hours per day to avoid noise induced hearing loss.

For more common modern music (12dB Dynamic range) this equates to 93dB average and 105dB peak. Exposure to these levels should be limited to between 1 - 2 Hours daily maximum.

For 100dB average which with modern music is 112dB peak the exposure limit is 15 minutes per day.

For 108dB average and 120dB peak the exposure limit is 0 minutes per day. Just saying.

https://blogs.cdc.gov/niosh-science-blog/2016/02/08/noise/

Of course, I am guilty of designing products that can exceed these levels with pretty much any headphone and far be it for me to dictate how people should listen.

However, for safe and normal listening levels with any music I can think of, 112dB peak is a lot. It should not pose a limit unless just trying to demonstrate "look how loud this goes" for which I commonly take headphones off.

You are correct of course that EQ will eat into the reserve, however even that should be manageable.

The reason I wanted to inject some common sense and context is that the OP did not complain he felt the Headphone driven by Zen DAC did not play loud enough, but that he was told (by others) that the Zen DAC could not provide enough power for his headphone and this made him want to spend more money, which I have my doubt as being essential.

Thor
 

Thorsten Loesch

Senior Member
Joined
Dec 20, 2022
Messages
460
Likes
530
Location
Germany, now South East Asia (not China or SAR's)
Do not agree with some of these. Especially the jitter bugaboo.

You do not have to. However, empirically established audibility limits for jitter as "fidelity impairment" exist and have been documented. This is regardless of your agreement or not.

323888802_5584518801670199_3886520853393252434_n.jpg


Note that this strictly "phasesnoise" and NOT discrete components.

I tend to agree that Jitter, like Harmonic Distortion, Noise etc. are overrated as fidelity impairment at very low levels, which is what we nowadays find in most equipment we test. .

Which is why always ask for context to numbers, that is what number of a given fidelity impairment is actually reliably audible and which level is reliably inaudible (p=0.05) to most people and how do these numbers compare to the number provided by the Device under Test.

And where is your evidence? Where are your references?

I referenced tests I performed in the process of product development. I commented elsewhere on details of methodology.

I made that clear. If you find this to be insufficient evidence to convince you, you are free to say so.

Give us examples of DUTs with objective measures below audibility that somehow still show a reliable preference.

I gave such an example of a device I designed and which was actually measured here at ASR.

Certainly the last paragraph holds, but you first have to supply evidence that what you are saying is so in the previous paragraph.

If my tests showed such a preference and they were done adequacy controlled, I consider this evidence sufficient for me personally. And they do.

I did not do these tests to make a scientific publication.

I made them to develop a product that I had solid reason to believe many listeners would have a reliable preference for, over other alternatives.

I mention my experiences here in specific conversations aimed at specific individuals and their part of the conversation.

I further find it funny that when a certain objectivist designer after adding extra NFB to an existing amplifier resulted in what I can only describe as "miraculous" sonic improvements, without providing evidence, methodology or anything else about the test, still all of ASR rally behind him against any criticism and claim it all as good and scientific and when I do something similar with a lot of detail on test and outcome I am told it is all terribly unscientific. Pot, kettle, black? But I cannot be arsed to do anything to do about it. I actually genuinely don't care, even though I might point it out and point a finger and say at times "Look, Cargo Cult Science in action!".

Thor
 
Last edited:

Mulder

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2020
Messages
624
Likes
871
Location
Gothenburg, Sweden
It should also be noted that so far no scientific evidence has been presented to the effect that these specific fidelity impairments are the only ones possible and that no other mechanism of fidelity impairments exist. In fact, over the decades since "fidelity limits" were first established in the 1950's we have seen the introduction of a number of previously unobserved and unknown fidelity impairments and their quantification, e.g. Group Delay, Phase Response, Slew-induced Distortion etc. et al, Jitter and it is likely that more such potential fidelity impairments remain to be discovered.
This is a variant of "Absence of evidence is not evidence of absence". https://quoteinvestigator.com/2019/09/17/absence/ I would say it is a very thin line between the "absence of evidence" type of argument and pure rhetoric. You can basically claim anything supported by this type of argument - e.g. there is no evidence god do not exist. As a minimum, if someone claims the existence of unmeasured impairment of fidelity, a valid hypothesis ought to be presented.

 
Last edited:

Gringoaudio1

Addicted to Fun and Learning
Forum Donor
Joined
Sep 11, 2019
Messages
585
Likes
743
Location
Calgary Alberta Canada
I have a friend who has gone through a lot of gear and probably ten DACs. Many very expensive. He likes the iFi Zen DAC II the best. I do not have such golden ears but I suspect that there is something to his preferences. I love the analysis here at ASR but the hardcore objectivist viewpoint based on what we can measure today will ultimately be debunked just as the ‘academy’ always is seen to be stuffy and conservative in science itself with resistance to new and ground breaking discoveries.
 

BDWoody

Chief Cat Herder
Moderator
Forum Donor
Joined
Jan 9, 2019
Messages
6,948
Likes
22,625
Location
Mid-Atlantic, USA. (Maryland)
I love the analysis here at ASR but the hardcore objectivist viewpoint based on what we can measure today will ultimately be debunked just as the ‘academy’ always is seen to be stuffy and conservative in science itself with resistance to new and ground breaking discoveries.

Only thing missing is actual replicable evidence...

They've only had a few decades to provide it, so I'm sure it's right around the corner.
 

Nationalizer

Member
Joined
Aug 27, 2019
Messages
25
Likes
33
Location
Roam-ania
I have a friend who has gone through a lot of gear and probably ten DACs. Many very expensive. He likes the iFi Zen DAC II the best. I do not have such golden ears but I suspect that there is something to his preferences. I love the analysis here at ASR but the hardcore objectivist viewpoint based on what we can measure today will ultimately be debunked just as the ‘academy’ always is seen to be stuffy and conservative in science itself with resistance to new and ground breaking discoveries.
Same, I prefer the Zen dac V2 versus the Gustard X16 and many others.
 
Top Bottom