• Welcome to ASR. 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!

DAC audible differences are finally proven with ABX?

Sampling theorem, sampled signal must be limited to F <= Fs/2 for proper reconstruction. Sudden change between the samples means F > Fs/2 and as such it cannot be reconstructed properly. Anti-aliasing input filter. Or oversampling etc., There is enough literature.
 
First, we don't want to keep checking there. You are a member here with industry tag no less. Sending us there is violating our rules.

Going to push back on the 'rules violation'. I don't know if this is a veiled threat of banning. He's not spamming, advertising, or maliciously linking.

I think it would better serve to accept his reasons for limiting interactions at their face value. It's not a good look Amir.
 
Members could always pool money to purchase said comparator and group share along with the DAC's under test by Josh.
 
Gotcha, makes sense, but I read "between samples" as "only happening for limited times, at the sampling rate" which to me implies a minimum frequency of the sampling rate for any harmonics. To be honest I am not sure if "between samples" truly makes sense in the first place?
It doesn't. There is no "between samples" in a properly bandlimited sampled system. It's kind of hard to explain without a lot of theory, but basically, we first filter the analog signal to make sure it has a fixed and finite bandwidth (Let's assume a perfect filter for now). As long as you sample this filtered signal at a rate of at least twice this bandwidth the original filtered analog signal can be perfectly reconstructed from the samples. There is no additional information "between samples".
 
Going to push back on the 'rules violation'. I don't know if this is a veiled threat of banning. He's not spamming, advertising, or maliciously linking.

I think it would better serve to accept his reasons for limiting interactions at their face value. It's not a good look Amir.
above all it was not Josh who opened this thread. He chimed in on page 3, post #41
 
It doesn't. There is no "between samples" in a properly bandlimited sampled system. It's kind of hard to explain without a lot of theory, but basically, we first filter the analog signal to make sure it has a fixed and finite bandwidth (Let's assume a perfect filter for now). As long as you sample this filtered signal at a rate of at least twice this bandwidth the original filtered analog signal can be perfectly reconstructed from the samples. There is no additional information "between samples".
I get that, but I am wondering like... the voltage going into the filter "between samples" isn't going to be going out of control for half the time, (is it?) and even if it was, that would get filtered out. I think I'm overthinking something that doesn't make sense as stated.

If it's a nonlinearity, it's signal-correlated, so if it's not in the THD or IMD, where is it?

If it's not signal-correlated, it should show up as noise, but it's not there either... so...?
 
Going to push back on the 'rules violation'. I don't know if this is a veiled threat of banning. He's not spamming, advertising, or maliciously linking.
Malicious? That is not what the term is for. It is to avoid gaining commercial interest at our expense. His profile says this:
1785530652483.png


If he is part of AS team then sending people from here to go there violates our rules. We don't ask much but we do ask that.
I think it would better serve to accept his reasons for limiting interactions at their face value. It's not a good look Amir.
There was not a single reason in the world for him to have given the second installer to the admin there, wait for him to then post it. He should have post it here if all he wants it so inform of us of his work. Anything else is a bad look for him.

And keep in mind that I am banned from posting on AS:
1785530941523.png


So can't respond to him or others there.

Really, we are uniquely situated to review his work, not a forum that bans objectivists and doesn't allow much, if any discussion of audio science/engineering.

Very disappointed in your comment. :(
 
I get that, but I am wondering like... the voltage going into the filter "between samples" isn't going to be going out of control for half the time, (is it?) and even if it was, that would get filtered out. I think I'm overthinking something that doesn't make sense as stated.

If it's a nonlinearity, it's signal-correlated, so if it's not in the THD or IMD, where is it?

If it's not signal-correlated, it should show up as noise, but it's not there either... so...?
It is filtered by the input anti-aliasing filter, so it is not there. Of course real filters have some leak, but that is why we are to test it.
 
Last edited:
above all it was not Josh who opened this thread. He chimed in on page 3, post #41
And he post just today telling us to go there and read his posts and responses ("So please check the thread there."). Hence my admonishment to him.
 
That would sadly not answer the main question if the specific comparator unit Josh owns has a defect / tell :confused:

Well you can't strong arm him into loaning it out... So your next best option is to get one and replicate the testing and let that outcome speak for itself.

AGAIN I'll save my money by using any number of quality, vetted, ~$500 DAC's and spending the $$ on speakers instead.

To say there is, on average, an overall difference in approach with ASR and other, more subjectivist leaning sites, is an understatement. Now if ANYONE at A.S. wants to race for pinks $20,000K setups where they have the typical 50-60% spent on not-speakers where I'll spend 90% on them then all the differences in streamers, DAC's, esoteric cabling, audiophile Ethernet, simply goes out the window. This is the standard I'm willing to prove on and even pay on.
 
And he post just today telling us to go there and read his posts and responses ("So please check the thread there."). Hence my admonishment to him.

He gave his reasons. He didn't link which is typical intent of such a rule. I think he did a solid for joining/posting #41, participating (now we are at post 630).

He's been asked some good questions, reasonable questions (ADA loop testing is huge), and we all have his response. I agree with Paul a perfect AB/X result stretches incredulity.

He's, IMO, at the limit of what he's going to participate here. Yours and others points stand.

I could care less if you can with incredible amount of training and an expensive AB/X box finally find that scintilla of a difference between two DAC's. I can think of better ways to spend money.
 
And he post just today telling us to go there and read his posts and responses ("So please check the thread there."). Hence my admonishment to him.
he didn't ask for his article being discussed here in the first place. You're the one asking something of him, not vice versa.
That he doesn't want to replay to everthing on 2 forums is understandable, isn't it?
 
I could care less if you can with incredible amount of training and an expensive AB/X box finally find that scintilla of a difference between two DAC's. I can think of better ways to spend money.
Which is NOT what we are dealing with here.
 
First, we don't want to keep checking there. You are a member here with industry tag no less. Sending us there is violating our rules.

Second, we have made simple requests of you which you keep denying. Instead, you wrote another missive which we did not ask for which took a lot more work.

We have all tried very hard to not question your personal motives. The fact that you don't want to cooperate with us is an objective fact and barrier to verifying what you have done is correct. For heaven's sake, you can't even be bothered to save a couple of files to be shared here.

As to not getting paid, what do you need money wise to do what we are asking you? For example, provide captured outputs of HPA4 for each input? If money is what motivates you to remove doubt regarding your work, we may be able to do that, as distasteful as that would be.

1) Insofar as it was a violation of ASR’s rules to point to my article, then the first post of this thread and everything that follows — including copying and pasting graphics from my article — is a violation, too.

2) You have not “tried very hard not to question my personal motives.” You have personally said “I find no logic in any of this other than fear of us finding something you didn't consider. Either you are in search of truth here or you are not.” You’ve also let ad hominem attacks such as “Would you trust a social scientist (an oxymoron right there, for me at least) to use instrument-based methods to check the volume matching or indeed do anything with kit convincingly?” stand from the first page of this thread. Moreover, you insinuated without evidence that my box was an older unit, despite my explicitly having said it wasn’t.

3) I actually reached out to you, as well as at least one other person in this thread for advice before I began my testing. In fairness, the advice was about level-matching, but at the time you raised no concerns about the ABX box’s transparency.

IMG_0266.jpeg

IMG_0267.jpeg


I also posted in the “Are Measurements Nothing or Everything” master thread about purchasing the ABX box and planning to conduct blind tests. I did so because I’m honest and wanted accountability regardless of the outcome.

IMG_5699.jpeg


Again, no one raised any concerns about the box’s transparency. Had you or anyone else said anything then, perhaps I’d have decided the box wasn’t good enough and abandoned the project. Instead, I bought a better DMM, went to great lengths to level-match precisely, and conducted ABX tests, since I wanted to resolve my own doubts about DAC audibility.

4) I don’t know why you took my plain statement of fact — that responding to requests on ASR is not my job and that I have an actual job, along with a personal life — to be a request for compensation for my services. It wasn’t. I was simply saying that I don’t have the time or desire to continue responding here. Indeed, had I known that posting in this thread would create the expectation that I continue responding indefinitely, I would’ve never started. I value my time far more than I value money. To that end, if you or anyone wants to make me an offer I can’t refuse for the box itself, you can investigate it to your heart’s content. I’m unlikely to accept any offer, but perhaps I should, since the emerging consensus here is that the box itself is flawed.

5) My article was straightforward and even-handed. I’m not claiming “My wife ran into the room agog at the difference between DACs” or “The heavens opened and I heard music like I’d never heard before.” I reported hearing small differences that may or may not be of value to others. I’m not trying to score points for either “side” of the silly “subjectivist” versus “objectivist” debate. I love music and audio, and I want to bring more people (particularly young people with limited budgets) into the hobby. That’s why most of my reviews are of affordable IEMs. I don’t want to fight with strangers online about something that should bring joy. Life’s too short.

6) To be clear, this will be my last post in the thread.
 
1) Insofar as it was a violation of ASR’s rules to point to my article, then the first post of this thread and everything that follows — including copying and pasting graphics from my article — is a violation, too.
Hotlinking to an image from another site is adjudicated in US to be legal as a form of fair use. See Perfect 10, Inc. v. Amazon.com, Inc. That is what is allowing google to show images in its searches: https://en.wikipedia.org/wiki/Perfect_10,_Inc._v._Amazon.com,_Inc.

"Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir., 2007) was a case in the United States Court of Appeals for the Ninth Circuit involving a copyright infringement claim against Amazon.com, Inc. and Google, Inc., by the magazine publisher Perfect 10, Inc. The court held that framing and hyperlinking of original images for use in an image search engine constituted a fair use of Perfect 10's images because the use was highly transformative, and thus not an infringement of the magazine's copyright ownership of the original images.[1]

As a professional courtesy however, if you want to formally request that any such images be deleted, provide a link here or in private message to me and I will delete them.
 
Moreover, you insinuated without evidence that my box was an older unit, despite my explicitly having said it wasn’t.
No. I mentioned what Erin told me:
I hope Erin doesn't mind me sharing this but he told me he passed on an AVA ABX comparator in Seattle due to it being the older model and hence, the one with the "tell." Josh says his is newer but what are the chances of two different units coming from same area of US?
I have acknowledged what you said as well.

Doubt remains as to whether there are audible tells in your comparator, whatever someone told you notwithstanding.
 
Malicious? That is not what the term is for. It is to avoid gaining commercial interest at our expense. His profile says this:
View attachment 548474

If he is part of AS team then sending people from here to go there violates our rules. We don't ask much but we do ask that.

There was not a single reason in the world for him to have given the second installer to the admin there, wait for him to then post it. He should have post it here if all he wants it so inform of us of his work. Anything else is a bad look for him.

And keep in mind that I am banned from posting on AS:
View attachment 548475

So can't respond to him or others there.

Really, we are uniquely situated to review his work, not a forum that bans objectivists and doesn't allow much, if any discussion of audio science/engineering.

Very disappointed in your comment. :(

Fair enough I stand corrected.
 
Again, no one raised any concerns about the box’s transparency.
We didn't know what we didn't know. If you had written to me saying what you found out in this testing, that would have been my immediate question and concern. Per your posting of our communications, you had full access to me. But you went from asking to simple questions to posting results that strain credibility.

When you find results like this you should question them more than publishing them. Your instinct was right to come to me initially. But wrong in not following up before publishing this.

Your instinct continues to be wrong in withholding options for us to either strengthen your case, or find critical information about how such a test could have gone wrong. Maybe that is countered by walking around saying your hearing is so good that blind tests have shown the same results so folks have to believe your review. I know which way I would have gone.
 
4) I don’t know why you took my plain statement of fact — that responding to requests on ASR is not my job and that I have an actual job, along with a personal life — to be a request for compensation for my services. It wasn’t. I was simply saying that I don’t have the time or desire to continue responding here. Indeed, had I known that posting in this thread would create the expectation that I continue responding indefinitely, I would’ve never started.
We would have had this discussion with or without you. Your participation level is not an issue. What is an issue is that we have made some simple requests and your reasons for refusal don't make sense to us, or at least to me. You worry about your comparator getting lost if you loan it to me for measurement? I will pay you $10,000 if that happens. Now you are going to do it or will there be another excuse?

You don't have time to run the same test and capture the output and post it here? That is the definition of posting here indefinitely?

We have spent so much time and energy here because you are a highly reluctant witness. You equate effort on your part to contributing to our questions. They are not the same thing. We, ourselves with our own fingers and ears, need to somehow analyze and replicate what you have done. Ultimately we will get it done if enough people find your work credible. That, you can bank on.
 
Back
Top Bottom