I’m sorry, what are you claiming here and what is the evidence? You seem to be implying there are some out there so we could get a better understanding of the magnitude of differences between DACs?
1 Like
No. I watched it just now. While I agree with some of your points about value and marginal differences between many products, this is a HORRIBLE and USELESS comparison. Extremely rapid switching in a crude, compressed YouTube stream won’t let anyone hear much other than the general overall profile. That’s arguably the worst example of AB testing I’ve ever seen.
To have a chance to hear differences: One must start with a very high quality sample, avoid compression, and test and calibrate the playback device to ensure that the amp and drivers/speakers actually reproduce the full sample. The content of the music sample(s) must also stretch to cover the performance range of the DAC. This includes bass, complexity (e.g., soloist vs. orchestral), and especially treble purity. His basic mid-range pop music will never tell you much about a DAC’s quirks or potential.
Many people design p*ss poor online AB and ABX audio tests. To do testing right one must control and calibrate everything: source, DAC, amp, speakers/headphones, and background (room) noise. One must employ a reasonable range of music samples that span the performance envelope. One must allow sufficient time for habituation (familiarization) between changes. All of this has been done for many years in university and commercial research (e.g., internal Bose, Sony, Apple audio product development).
I do agree that there’s little difference between many DACs (and amps for that matter). I do agree that there’s little to no correlation between price and performance too.
4 Likes
Yes, I think you can tell alot about the performance of a piece of audio gear from its measurements.
What I think about is not the live performance. I think about the intended listening space for the recording, and how I might try to approximate that. This is what I mean by fidelity.
From my perspective, noise, distortion, and coloration would only degrade the integrity or fidelity of a recording… unless that is the listening space intended for the recording by the producer.
You’re referring to the number of bands or filters in the Mimir’s EQ.
3 isn’t alot in my book. But I suppose it would depend on the headphones or speakers you want to correct. And how much flexiblity there is in the filters. And whether or not you could adjust their frequency and Q values, in addition to their amplitude.
Some compact mixers I’ve used only had three filters: a high and low shelf, and midrange adjustment. So I have some experience with 3-band setups. I’ve used 31-band graphic EQs as well. My digital EQs are still mostly GEQs with a fair number of points or bands (so still in the stone age). Some of that is for smoothing though.
I wouldn’t be averse to trying a DAC with a built-in EQ though.
Some good point made here. Appreciate the feedback from everyone.
What got me to post this was auditioning the DX9 as a DAC was a bit of a shock. It sounded as good to me as expensive DACs I had on hand or have owned. I have noticed more differences with headphone amps than I have with DACs.
There is a lot written about the differences between R2R DAC’s and Delta/Sigma DACs. Still am curious if users here on the forum hear a meaningful difference with R2R and Delta/Sigma. Watching reviews on DACs from Golden Sound, he gives the impression that whilst many prefer the R2R sound, for Classical/Orchestra music, the Delta/Sigma may be a better option.
3 Likes
It’s about the least important thing in my book, there are good sounding DAC’s with both technologies, with the caveat that I can very clearly hear the difference in NOS vs OS DAC’s.
My 2c on this subject as someone who owns 1/2 a dozen DAC’s some very expensive, and some cheap, they all sound different to me. Among the cheaper DS designs with op amp output stages, I’d struggle to fast A/B them, bit I could probably get there with enough time.
Fast A/B switching DAC’s is hard to differentiate, but when I’ve spent time listening to them there are clear differences to me. This has everything to do with perception, we don’t hear just instantaneous tones, our mind builds up an impression over time. When I’ve had multiple expensive DAC’s here and spent time listening to them I’ve always had a preference for one over the other.
Neither of my expensive DAC’s measure spectacularly, they are both in what I’d term the not broken category of measurements, which is all I really look for in measurements.
Ones an NOS R2R DAC, and the other is using a DS chip with a tube output stage.
Having said that, the cheaper ones sound good, but I’d rather listen to the more expensive units in my system.
I don’t understand how anyone can’t hear the differences between amps once you get outside the high feedback low SINAD designs which do all sound very similar.
And at the end of the day the output section of a DAC is just an amp, it just has to drive a much higher impedance load.
3 Likes
There is a difference between fidelity to the recording and fidelity to the performance/experience. Exclusively using measurements of a amp/DAC/headphone can only indicate the former. If something measures well, this is good enough for many people. And there’s nothing wrong with that, within reason. Others may prefer something they feel better replicates the experience of hearing the music live rather than through the recording, even if it means adding distortion. And there’s nothing wrong with that either, within reason.
So really there is no conflict between stating that modern inexpensive DACs can be transparent, and that DACs can sound different.
4 Likes
Here’s a blind A/B comparison of two DACs that indicates a difference: Schiit Bifrost 2 Vs. Bifrost 2/64 - A Schiit Shootout! - AudioHead
I am not suggesting this test is conclusive since participants were asked about which they preferred and not tested for whether they could reliably tell a difference. However according to the author the comments on the differences were consistent, so that gives some support that the participants were really hearing a difference.
5 Likes
Bingo! This was my second smoking gun - after hearing modern dongle differences for myself - that told me there’s something going on here: the consistency between reports of which DACs are warm/lush/musical and which are cold/sterile/clinical from those much-maligned subjectivist reviewers. Sure they don’t all agree all of the time, sure it could be some kind of “social placebo” where whatever the first reviewer said is then repeated by many of the others because they went in primed to hear it that way, but… they’re so consistent across different devices though, and many of them do take care not to go in primed by exposure to other reviews. It can’t be rational to outright dismiss the possibility they’re hearing something real. How many scentific discoveries have started out in just this way - with interestingly consistent reports from otherwise low-down dirty subjectivist/unprofessional evaluations? 
2 Likes
So, yes. I generally agree with all the above, AudioTool. But not sure how one idea necessarily follows the other, on your last point.
Some DACs are transparent, and some probably are not. And add some distortion or coloration. And some may sound the same, and some may sound different. I see no conflicts in any of that, regardless of your objectives.
I also understand the desire to try to better replicate live performances. Especially the ones that you may have been to personally. Axel Grell also had some interesting things to say about that in the recent livestream he did with Resolve, Blaine, and Listener. I’m sorry I don’t have a timestamp on this either, because it’s a long video. But I think his opinions on the definition of “fidelity”, along these lines, were part of what informed the design goals for the OAE-1.
The whole video is worth a listen though. Especially the first half where Grell discusses his goals and philosophy for audio.
If my memory is correct though, he also started out believing that fidelity to the performance, and to the sound of live instruments was the most important thing. But over time he realized this was a very difficult objective. And so he began to focus more on the artist’s intent, and how they might hear or experience a recording in a good studio.
I try to take this notion a step further, and think about how the music is intended to be consumed, which may or may not be how an artist might have heard it during some of the different stages of production in a studio.
1 Like
Art or
Artifice or
Artificial
Eternal question with all things creative
3 Likes
I’m not sure what you’re trying to suggest above, generic… Maybe that focusing on the point of consumption is somehow more artificial than the point of creation? I’m not one of the deepest thinkers. 
But I understand that many enthusiasts also want to be part of the creative process, and decide how music is to be consumed in their own home, headphones, or other listening spaces, rather than having this prescribed by some other “expert”.
As I think about all this some more though, I think my own goal really is transparency, rather than fidelity to either the performance, or recording, or the artist’s intent.
The target or intended listening space is still important to me (more important than trying to replicate a live or artist’s experience that I have no knowledge of). But only to a point. Because what I think I really want to hear in my gear is the recording stripped completely bare of audible noise and distortion, with most or all of its warts on full display! And with a completely neutral (or “flat”) coloration.
I may not always want to listen that way. But I want the ability to at least approximate that condition for some of my more critical listening applications.
1 Like
This forum I feel as if is combining two diverging conversations.
First and foremost, subjective perception of how you enjoy the audio chain and music: great! At the end of the day this is a luxury hobby, where you personally decide the validity of your purchases.
Now I believe there is still a need for an objective approach to make audio tangible. There really is an objective truth to dacamps, although not an end all solution, that performs as a standard that is contrasted for further research. I believe this is what resolve means when he refers to a ‘good dac’, where it does its intended job perfectly. You cannot expect the DAC to create an out of body experience for its purpose is to translate the digital file into an analog signal. If it does; objectively it’s a poor performing DAC, even though subjectively it may not be. Performance of the source gear is measurable, whereas the externalities which arises when testing the source gear, and the person’s experience is not.
Hyperbole example:
Let’s say you have a dacamp setup with perfect measurements that sounds horrible to 99% of the population. Does this make it a horrible setup? Subjectively yes, but objectively no. The data extracted here is still very valuable, because this brings cases for other avenues of hypothesis testing.
Data/statistics are there to analyze, interpret, and present explanations to the experiences which the world offers. The experience should not be used to nullify the data, but rather another catalyst to further it.
“When there is no objective truth, when everyone gets to make up their own reality, their own script, authoritarians thrive.” Peter Wehner(https://archive.is/lgMD1)
1 Like
I think many of us feel that (1) good measurements are easy to achieve and therefore not very useful, and (2) the value of a DAC as a hobby purchase comes from delivering a superior qualitative personal experience.
Many $10 or $20 DACs – such as the Apple dongle – measure well. Long, long ago DACs achieved “objective truth” as you characterize it. Reaching 99%+ accuracy in reproduction is now taken for granted.
My position is not that “subjective” factors dominate, rather, that arbitrary human perception and fatigue are what matters to humans. Electrical measurements (classic audio “objective” standards) typically correlate well with human perception – the interesting stuff happens on the margins.
But some of us don’t care or don’t hear the differences between devices and become absolutists or move on.
8 Likes
But we are not authoritarians. We are aquariums. Underwater from perchases. I’m adding an under gravel filter to my system.
4 Likes
Why would a DAC/AMP with perfect measurements sound horrible? What would be the reason for that? And what would be reason for not showing up in the measurements?
Torben
The problem I at least see with measurement first is the obsession with measurements above good enough, and the obsession with specific measurements.
Why do we care if something is 0.00001% vs 0.01% THD even 1% THD which was considered good for decades and is inaudible if your just listening to test tones.
The problem with this isn’t the measurement, it’s the way you have to design to reach it.
Grossly simplifying here, lets say you need an amplifier with 10x gain you can get there by having 10x gain in your gain stage and no feedback or by having 100x gain and 90% feedback.
The Second design will have MUCH lower distortion numbers and output impedance for that matter, but adding that much feedback impacts other parts of the signal like phase, bandwidth and to a lesser extent slew rate.
You can measure all of that but it’s much less clear what audible is on those scales.
By picking a single metric and optimizing for it you basically dictate the design, and your ignoring all the compromises that design choice has.
That’s not to say that measurements are bad, every amp manufacturer or DAC manufacturer is going to measure, the only difference is what they are optimizing for. Nelson pass used to publish a number of papers on novel amp designs (I don’t know if they are still on the First Watt website) it’s interesting to see his comments on and interpretation of measurements. He often for example comments on the phase of the first harmonic noise component because to him it’s an indicator of the sound the amp produces.
It’s just a complicated subject once you get beyond my distortion figure is lower than yours comparisons.
3 Likes
Floyd Toole also talks about “subjective measurements” though, which are highly repeatable. So I’m not sure I agree that personal experience isn’t measurable or predictable. The Harman research seems to refute this idea.
Your opinions and perceptions can be predicted with a pretty high degree of precision. At least under blind/controlled circumstances.
Probably needless to say, but this has not been my experience.
I haven’t found good measurements easy to achieve. And I’ve also found measurements, in general, to be quite useful!
If you’ve found them to be easy though, that’s probably a good thing.
I think these are all subjective though.
I can’t really refute this, because I don’t really understand what you mean by “interesting stuff” or “on the margins”.
In my experience, the closer I get to no noise, no distortion, and a “flat” response, the better and more consistent my results seem to be. And getting there certainly has not been an easy journey, especially with headphones.
For a long time, the objective standard for headphones was diffuse field, which doesn’t really correlate well with human perception (most find it unpleasantly bright). This was one of the findings of the Harman research. Perhaps you were referring only to source gear though, where things tend to be a little more cut and dried.
Absolutely love this haha