Review and Measurements of Topping DX3Pro DAC and Headphone Amp

[derp1n]

DT880 600ohm and my trusty Radio Shack sound level meter level with the plane of the left earpad, playing Track 58 (stereo white noise) from the 192k version of http://www.chesky.com/album/ultimate-headphone-demonstration-disc-jd361, I am getting around 75-77 dbs with the DX3 at max volume, high gain mode. With most of my music, I am turning it down to around -8db for loud yet comfortable listening. Plenty of power imo, unless you are trying to *really* bump your most quietest tracks.

That's a flawed test and severely understates the achievable volume.

Firstly, you can't take accurate measurements of a headphone in a free field, you need to seal the cup to simulate coupling with the head.

Secondly, that track is mastered at -30 LUFS, so it's very quiet compared to real music.

 

[pauze]

That's a flawed test and severely understates the achievable volume.

Firstly, you can't take accurate measurements of a headphone in a free field, you need to seal the cup to simulate coupling with the head.

Secondly, that track is mastered at -30 LUFS, so it's very quiet compared to real music.

I concede it wasn't a rigorous test, I just wanted to give at least a (crappy?) data point for those with 600ohm cans and wondering if the DX3 should even be considered. I actually had dreamt up making a fixture using cardboard and bubble wrap but i got lazier than that contraption sounded on paper. Can edit my post to caveat that it's very rough numbers, unless the more professional folks here think it's better off gone.

 

[jsmiller58]

But how can you make researches, when u only can read about planctons, and glare and dark sound and many other impossible to grasp biased subjective audiophile terminology? Thats why ASR is better source, because these measurements are objective things, no nonsense, easy to see through the things. With other subjective reviews we can go back and let the companies steal our money for trash products. U know Yggdrasil is the best dac for 2k bucks, based on 90% of the reviews...

Hmmm... Not entirely sure of what you are trying to say... My advice remains unchanged. Do your own research. Make sure that anyone's advice you are following is using the product the same way you would use it. If you need more than that, here is another piece of advice - do not be an early adopter. Wait for others to try the product - that way there will be many more use cases explored.

I have high regard and respect for Amir, but he is a human being only. He has better toys with which to test gear than the vast majority of us. But he certainly is not testing everything in every possible way, and is not claiming to do so. His recommendations are just that, recommendations. Not commandments written on stone tablets.

 

[derp1n]

I concede it wasn't a rigorous test, I just wanted to give at least a (crappy?) data point for those with 600ohm cans and wondering if the DX3 should even be considered. I actually had dreamt up making a fixture using cardboard and bubble wrap but i got lazier than that contraption sounded on paper. Can edit my post to caveat that it's very rough numbers, unless the more professional folks here think it's better off gone.

I think it's fine to leave it up (although adding a caveat would be good). I just think it's important for everybody understand the limitations of certain measurement techniques and take care drawing conclusions from them. In your case, you came to the right conclusion anyway (the amp is powerful enough for these headphones), and that matches an earlier post I made using the amp's output and headphone's sensitivity to calculate the expected max SPL.

 

[bellyofthebizd]

I don't understand the argument here. All I have said is under full DAC volume the noise is beyond hearing range but when DAC is playing low volume signal, (signal - noise) is no longer > 16bit (- 96db), and you started to hear the noise. It is possible you are playing your music in 96db and the noise is 20db. It's entirely unrelated to the amp volume indicator.

Ok, so can anybody explain me what measure linearity test in simple words?

 

[yue]

Uhhh... Only voice? Where'd you come up with that? Of all the bizarre comments I have read on this thread, that one stands out as truly strange. And that is saying something.

There are many audio reviewers, each with their own methodologies, skills, and yes, even biases. As a consumer it is YOUR responsibility to ensure that you research what you buy. When you rely upon the opinions and writings of others you need to make sure that their assessments are based on conditions similar to yours. And, if the purchase is of sufficient value or importance to you that you get multiple inputs from different parties.

Sorry, I just had to comment. I will go back to quietly watching the spectacle that this thread has become.

But as far as I know this forum is the only one that gives objective tests to a wide range of audio equipments. There are a lot of subjective reviews elsewhere, and people have no way to research, analyze, and compare the results at all. If you want objective data, here is your only source.

You are right, It is user's responsibility to ensure that they research before they buy. But there is a precondition to this ---- @amirm has to make sure he performed very scientific test during the measurement. No user conclusion can hold if the data from @amirm is flawed. There're various ways to validate a scientific experiment, but none of them are viable to other people

- experiement should be reproducible. Others can buy an audio analyzer and repeat the same process, and post reviews here. But an audio analyzer is quite expensive, and performing such test is time consuming. Post bad review here may make manufactures angry and you may receive threats from them. So no one is willing to do that.

- manufacturers have audio analyzer and they can do the same thing as @amirm do. But unfortunately they are manufacturers so the review may be unconvincing.

- I do have access to audio analyzer from my employer but unfortunately I can't use company resources to pursue personal hobby.

- when there's no budget to buy another analyzer (such as the case we only have one LHC doing particle collision experiements), we can have 2-3 groups of scientists doing the experiment and analyze the data independently. This is how scientists around the world found Higgs boson. However in our case @amirm does not lend his equipment to another forum member. So again this is not possible.

IMHO @amirm made a few quite serious mistakes in some of his reviews. I can't tell if he made such mistakes in all his reviews. So in current situation, readers should be aware of these potential problems, and @amirm should try his best to perform scientific experiments.

 

[mi-fu]

But as far as I know this forum is the only one that gives objective tests to a wide range of audio equipments. There are a lot of subjective reviews elsewhere, and people have no way to research, analyze, and compare the results at all. If you want objective data, here is your only source.

You are right, It is user's responsibility to ensure that they research before they buy. But there is a precondition to this ---- @amirm has to make sure he performed very scientific test during the measurement. No user conclusion can hold if the data from @amirm is flawed. There're various ways to validate a scientific experiment, but none of them are viable to other people

- experiement should be reproducible. Others can buy an audio analyzer and repeat the same process, and post reviews here. But an audio analyzer is quite expensive, and performing such test is time consuming. Post bad review here may make manufactures angry and you may receive threats from them. So no one is willing to do that.

- manufacturers have audio analyzer and they can do the same thing as @amirm do. But unfortunately they are manufacturers so the review may be unconvincing.

- I do have access to audio analyzer from my employer but unfortunately I can't use company resources to pursue personal hobby.

- when there's no budget to buy another analyzer (such as the case we only have one LHC doing particle collision experiements), we can have 2-3 groups of scientist doing the experiment and analyze the data independently. This is how scientists around the world found Higgs boson. However in our case @amirm does not lend his equipment to another forum member. So again this is not possible.

IMHO @amirm made a few quite serious mistakes in some of his reviews. I can't tell if he made such mistakes in all his reviews. So in current situation, readers should be aware of these potential problems, and @amirm should try his best to perform scientific experiments.

Maybe it would be great for the whole community if you could specifically pinpoint what the mistakes are?

I'm sure @amirm would be happy to engage in any conversation that helps to improve his methodology. Only by calling "mistakes" would not be very helpful in improving the quality of scientific investigations.

I think we should take these reviews more as a "collective effort to know better" than just one person's work.

 

[jsmiller58]

But as far as I know this forum is the only one that gives objective tests to a wide range of audio equipments. There are a lot of subjective reviews elsewhere, and people have no way to research, analyze, and compare the results at all. If you want objective data, here is your only source.

You are right, It is user's responsibility to ensure that they research before they buy. But there is a precondition to this ---- @amirm has to make sure he performed very scientific test during the measurement. No user conclusion can hold if the data from @amirm is flawed. There're various ways to validate a scientific experiment, but none of them are viable to other people

- experiement should be reproducible. Others can buy an audio analyzer and repeat the same process, and post reviews here. But an audio analyzer is quite expensive, and performing such test is time consuming. Post bad review here may make manufactures angry and you may receive threats from them. So no one is willing to do that.

- manufacturers have audio analyzer and they can do the same thing as @amirm do. But unfortunately they are manufacturers so the review may be unconvincing.

- I do have access to audio analyzer from my employer but unfortunately I can't use company resources to pursue personal hobby.

- when there's no budget to buy another analyzer (such as the case we only have one LHC doing particle collision experiements), we can have 2-3 groups of scientist doing the experiment and analyze the data independently. This is how scientists around the world found Higgs boson. However in our case @amirm does not lend his equipment to another forum member. So again this is not possible.

IMHO @amirm made a few quite serious mistakes in some of his reviews. I can't tell if he made such mistakes in all his reviews. So in current situation, readers should be aware of these potential problems, and @amirm should try his best to perform scientific experiments.

Uhhh. Sorry. Amir and ASR are not a government run consumer protection agency or goods testing service. He has no duty to you or to anyone. He runs the tests he runs, and he runs them the way he sees fit. I have seen him engage in constructive discussion if changes are called for. But I also see him taking the other party to task to be specific and constructive in their feedback.

He offers his time, his efforts, and his conclusions freely. And as this is free advice from him, that is precisely what he charges you, and how much absolute faith you should put into it.

You should consider a crowdfunding program to buy your own measurement gear and start your own measurement forum where you can hold yourself to these standards you want to hold Amir to.

Your attitude and expectations are somewhat foreign to me, I must admit. As such I am really trying very hard to not be condescending towards you. Please don’t keep making it so hard for me to avoid doing that.

 

[yue]

Maybe it would be great for the whole community if you could specifically pinpoint what the mistakes are?

I'm sure @amirm would be happy to engage in any conversation that helps to improve his methodology. Only by calling "mistakes" would not be very helpful in improving the quality of scientific investigations.

I think we should take these reviews more as a "collective effort to know better" than just one person's work.

In almost all experiments he did not ground the APX555 analyzer. Some audio devices are sensitive to grounding. As a result, this undermines the reliability of these measurements.

 

[Thomas savage]

There’s some tea stains on amirs lab jacket, bloody unprofessional. I mean how can you trust a guy that can’t keep tea and kebab ( yes I think there’s some kebab on it too) off his coat.

Let’s not even get started on his choices of footwear and lack of hairnet..

 

[amirm]

In almost all experiments he did not ground the APX555 analyzer. Some audio devices are sensitive to grounding. As a result, this undermines the reliability of these measurements.

The analyzer is always grounded. Would not be safe otherwise.

If you mean audio input, balanced gear doesn't need it care about grounding so your comment doesn't apply there.

For unbalance input, on purpose AP inputs are floating. I have the option to ground or not. I try both and post the results of best outcome. For grounding I actually try half a dozen ways to find best outcome.

I have commented on these issues a number of times. I have also said how broken unbalanced audio is to cause all this complications.

So your points don't apply at all. You want to tell us where you got these false ideas about my testing?

 

[yue]

Uhhh. Sorry. Amir and ASR are not a government run consumer protection agency or goods testing service. He has no duty to you or to anyone.

Your attitude and expectations are somewhat foreign to me, I must admit. As such I am really trying very hard to not be condescending towards you. Please don’t keep making it so hard for me to avoid doing that.

Well you are right at some point. @amirm is a free man and have freedom to express whatever he believes. But I am not talking about it in that sense.

What I point out is the fact, that at some point, because @amirm is the only such source, his measurements may mislead customers. By mislead I mean sometimes he didn't test those equipments in a correct way, or gives a wrong recommendation based on a correct measurement. Again @amirm may be unintentional if this happen, but such things could happen, and it did happen here.

We want constructive conversations in the forums, and I all that I did is point out such a possibility, a possibility that people are misguided because @amirm's review.

Again you are right he is not the government. But it doesn't mean we can not be critical, or he can not improve his measurement in the future. I believe everyone here involved (including @amirm) wants to have constructive communications and perform more reliable measurements here.

 

[Jmudrick]

Don't be fooled by software like Jplay, just read what Jriver developers have to say :

https://jriver.com/jplay.html

Sound quality wise, in windows Foobar2000 & wasapi are the best you can get, or if more functionality and better library management is needed, Jriver is a good commercial option, there's also Roon which is more expensive among others.

ASIO is an interface that provides bit perfect playback but is geared more towards professional use, so wasapi is usually more practical for music listening where things like latency are mostly irrelevant.

Been there done that, prefer Jplay thanks.

 

[yue]

So your points don't apply at all. You want to tell us where you got these false ideas about my testing?

I don't represent my employer but with access to the original measurements of v1 and v2, I don't think your pixel dongle experiments are properly performed. Again bear with me I can't discuss it here further because of company policy. I do not work on the product so if that team decides not to join the conversation here this is as far as the discussion can go.

 

[amirm]

I don't represent my employer but with access to the original measurements of v1 and v2, I don't think your pixel dongle experiments are properly performed.

Well then represent us to them. Tell them how terrible it is that they don't publish any measurements for those audio products. That they have them but don't share is even worse. If companies like yours served the customer well I wouldn't need to spend so much time, effort and money doing this work.

And oh they could contact me in private and with confidentiality with their data.

In absence of either one of these and your ad-hoc stone throwing, I am not feeling like there is any positive intent here.

 

[amirm]

What I point out is the fact, that at some point, because @amirm is the only such source, his measurements may mislead customers.

That may happen. The best antidote is to insist that manufacturers post full measurements of audio products. To the extent you work for a company who chooses not to do that, then I say you are part of the problem, not the solution. Go and speak to your teams to be more transparent with their customers. What are they worried about? That the truth cost them sales? If so the solution is to build better products. Not to come here with vague criticism that is not actionable.

 

[yue]

Well then represent us to them. Tell them how terrible it is that they don't publish any measurements for those audio products. That they have them but don't share is even worse. If companies like yours served the customer well I wouldn't need to spend so much time, effort and money doing this work.

And oh they could contact me in private and with confidentiality with their data.

In absence of either one of these and your ad-hoc stone throwing, I am not feeling like there is any positive intent here.

I did talk to one of their engineers and if they decide to make a reply, they will.

 

[yue]

The best antidote is to insist that manufacturers post full measurements of audio products. To the extent you work for a company who chooses not to do that, then I say you are part of the problem, not the solution.

What are they worried about? That the truth cost them sales? If so the solution is to build better products. Not to come here with vague criticism that is not actionable.

This is pretty ridiculous --- Similar to Apple's dongle, Google's dongle is not a hifi device and customers simply don't care about releasing the measurements at all. Google or Apple needs to make sure the dongle performed "reasonably well" for mass consumers and that's the reason they performed the audio test. For each of the products those big companies released, there were dozens of tests being performed --- things such as radiated power or total isotropic sensitivity. There will be tests that focus on durability, safety, reliability, connectivity, etc. But like radiation power, high fidelity is never the marketing points. It's meaningless to publish all the test results to the general public, as no one is interested in those data at all.

Simply put, big companies don't care about releasing the test you did at all. Neither do customers. This doesn't mean they didn't perform the tests. They did all kinds of tests (including ones we can't imagine) in a serious way, and they make sure the product performs reasonably well for their customers. You were VP for Microsoft so you know better than I do.

Also to clarify I'm not promoting my employer's products here. I don't work for the hardware team and I don't care about dongles ---I don't even think the company cares about how many dongles they sell as compare to other revenue sources (such as youtube ads), dongle sales number is infinitesimally small and means nothing.

For people like me who have access to internal tests reports, they will have reason to cast doubt on the measurements here. I do not blame you for anything. All I want to suggest is people should hold a more skeptical view when reading reviews because "Objective review" does not mean the whole story. Your report may serve as a good reference, but in the end people need to use their brain to make the choice.

 

[maxxevv]

This is pretty ridiculous --- Similar to Apple's dongle, Google's dongle is not a hifi device and customers simply don't care about the measurements at all.

A 100 years ago, food and health safety were not a "concern" with the general public either.

But that simply wasn't true. Not that they didn't care, they didn't know better.

Look at how things have changed since FDA became a requirement ?

OR if you want to delve into the less savory stuff, how religious and political abuse was "never a concern" , but in reality its was because people didn't know any better as the information was withheld from them.

Likewise people waste money and resources on less than well performing audio products in their pursuit of audio bliss quite simply because information was withheld from them.. Or that they were mislead into believing that some products were worth far more than the manufacturers purported engineering chops garnered.

Information is power. Withholding it is deceit at best, false representation of it should be a crime. ( I don't want to go into the political scene in the US, but many of the politicians there deserve to go to jail simply for that.).

Not that people don't care about such information, rather they never knew it existed. And they never knew how it could empower them.

 

[yue]

A 100 years ago, food and health safety were not a "concern" with the general public either.

did you read my reply in full? I expressed pretty clear that the company did serious measurement, but there is no value to release it to the general public.

To clarify, the company did care about all kinds of measurements (and you don't even know what most are for), but they don't care about releasing it to the general public.

To reiterate I do not represent my company here. All I posted are personal opinions.

Information is power. Withholding it is deceit at best, false representation of it should be a crime.

Oh CRIME! I think you are free to to report the crime and let the prosecutor sue Google/Apple for not releasing the reports.

Not that people don't care about such information, rather they never knew it existed. And they never knew how it could empower them.

Good point. You can do a poll and see how many people here have good music theory knowledge. A better performing DAC vs a better understanding of music, I would confidently say the former means nothing and the latter empowers much more.