Help GoldenSound / GoldenOne Get An Audio Precision Analyzer

[sphinxsix]

4 hours ago, firedog said:

Or just something he can't explain. Not the same thing.

Or something 'audio science' can't explain or can't explain yet.

I think that the willingness to accept the fact that the model of the world that we know at the moment is just approximation of truth among contemporary physicists is bigger that willingness to accept the fact that some things human ears can hear can not be explained by measurements, at least not yet, among the 'hard-core' objectivists. 

[fas42]

47 minutes ago, sphinxsix said:

Or something 'audio science' can't explain or can't explain yet.

I think that the willingness to accept the fact that the model of the world that we know at the moment is just approximation of truth among contemporary physicists is bigger that willingness to accept the fact that some things human ears can hear can not be explained by measurements, at least not yet, among the 'hard-core' objectivists. 

 

Measurements can determine what's going on - a trivial example I looked at years ago was a YouTube video which had a live performance being compared to a recording of those same live performers, played back over a "high end system" ... it was extremely easy to see, by examining the waveform, that transients were very poorly rendered by the reproduction rig - no wonder they sounded different!

 

A closer to home example was the microphone recordings by @manisandher, posted on a thread. They sounded different, and using the DeltaWave program to extract the difference file, it was easy to see that the treble energy in one version differed from the other - there was a clear pattern in the variations, throughout the piece. But objectivists have a deep reticence to more vigorously pursuing these avenues - it may get too close for comfort, threatening deeply held beliefs, 😉.

[Rexp]

1 hour ago, fas42 said:

 

Measurements can determine what's going on - a trivial example I looked at years ago was a YouTube video which had a live performance being compared to a recording of those same live performers, played back over a "high end system" ... it was extremely easy to see, by examining the waveform, that transients were very poorly rendered by the reproduction rig - no wonder they sounded different!

 

A closer to home example was the microphone recordings by @manisandher, posted on a thread. They sounded different, and using the DeltaWave program to extract the difference file, it was easy to see that the treble energy in one version differed from the other - there was a clear pattern in the variations, throughout the piece. But objectivists have a deep reticence to more vigorously pursuing these avenues - it may get too close for comfort, threatening deeply held beliefs, 😉.

If that gives you comfort that you're not missing out, fine. 

[Rexp]

On 6/14/2021 at 9:51 PM, GoldenOne said:

So I'm going to have an APx555 here for a couple weeks (not mine, not keeping it, just here temporarily). This likely won't be long enough to do some of the more in-depth stuff but I'm going to measure anything I can get my hands on just cause why not, but also will probably do a brief video explaining a little about why the APx555 is different to just using a good ADC etc.

If anyone has anything specific that they think I should do whilst it's here lmk!

Im more interested in measured difference between recordings than hardware, for instance can you do a 16/44.1 v 24/88.2 of Daft Punks Random Access Memories? I'll send files if necessary. 

[Confused]

13 hours ago, Superdad said:

 

Well 2,500 EtherREGEN owners--plus thousands more using and discussing their audible experiences with other switches, clocks, filters, SFP transceivers, and cables--will disagree with you there... 

Mindful that this thread is discussing things that could potentially be measured with an Audio Precision APx555, this perhaps begs a question.

 

@GoldenOne has already mentioned that he would be happy to measure an audiophile switch or router if someone were to loan one to him.  So lets say someone provided him with an EtherREGEN, would it be possible to use a APx555 to demonstrate measurable benefits from an EtherREGEN?  If so, what measurements or measurement methodology would you propose? 

 

This is a genuine question by the way, after all, I am one of those happy 2,500 EtherREGEN users.

 

[Confused]

7 hours ago, Rexp said:

Im more interested in measured difference between recordings than hardware, for instance can you do a 16/44.1 v 24/88.2 of Daft Punks Random Access Memories? I'll send files if necessary. 

Actually, this interests me also.  As an example, I can listen to something and think that my system sounds great, clean clear, no issues.

 

Then to the nest track, and maybe I hear some very slight distortions somewhere.  Of course, this can lead to paranoid thoughts that there is something wrong with my system somewhere, something I need to fix or tweak.

 

What would interest me would be if there was someway to analyze the actual digital file to identify and quantify such distortion.  I am thinking that in practical terms this is pretty much impossible to do.  For example, how would you discriminate between distortion that is supposed to be there, say from a guitar amp of effects peddle, and something that should not be there, say caused by a microphone or electronics used during the recording.

 

Comparing alternative versions of the same recording as you mention, is much more likely to reveal something meaningful I would think.

[sphinxsix]

13 hours ago, fas42 said:

Measurements can determine what's going on

I've seen so many examples of measurements not being able to explain 'what's going on' (not only in audio but e.g. in contemporary atomic and astro-physics  which both are, I hope you will agree, much more advanced in every way than 'audio science') that from the objective, scientific point of view  the word may would be more justified.

 

[JohnSwenson]

10 hours ago, Confused said:

Mindful that this thread is discussing things that could potentially be measured with an Audio Precision APx555, this perhaps begs a question.

 

@GoldenOne has already mentioned that he would be happy to measure an audiophile switch or router if someone were to loan one to him.  So lets say someone provided him with an EtherREGEN, would it be possible to use a APx555 to demonstrate measurable benefits from an EtherREGEN?  If so, what measurements or measurement methodology would you propose? 

 

This is a genuine question by the way, after all, I am one of those happy 2,500 EtherREGEN users.

 

This is a tough one. The AP measures two things that may be useful for this: spectrum of audio signal coming out of DAC and jitter of S/PDIF type signals (if the analyzer has the module for that). As the designer of the ER I know what it is trying to do and this results in decreasing jitter signals going into a DAC to levels much less than 1ns. So in order to measure the results of this takes measuring equipment that has clock jitter LESS than what we are trying to achieve. I don't think the AP has clocks that good.

 

I have recently acquired a new piece of test equipment which can make jitter histograms down to 1ps resolution and can handle pretty much anything, USB, I2S, S/PDIF, Ethernet etc, with this I DO see significant differences in USB feeding DACs. But I have only been working with this for a short period of time and need a lot more time to get things figured out. Unfortunately as has been discussed already the environment can have a big influence on measurements at this level.  How you lay the cables out, how close YOU are to the bench, is sunshine shining on the bench, things like this all effect the measurement. 

 

So far I have only been looking at USB. The big issue right now is getting signal INTO the test equipment, I have to make special adapter boards to properly get whatever in to the equipment without messing up the signal. Especially for things like USB and Ethernet which requires two boxes talking to each other in order to get the signal in the first place, the test equipment has to "spy" on the signal without changing it. This is not easy.

 

John S.

[GoldenOne]

I've been offered a loan which will cover much of the cost of the analyzer. So close now!
Selling anything I can that isn't tied down to try to cover the remaining amount.

But if anyone here is able to do so and has enjoyed the content I've made so far please do consider supporting even just for this month. Hoping to place the order in the next couple weeks!
 

[GoldenOne]

 

(Seriously, thank you so much to everyone who contributed to this! I'm really looking forward to making some good stuff with it.
Still need to work on getting a dummy load though)

[The Computer Audiophile]

1 minute ago, GoldenOne said:

 

I love it!!!

[firedog]

Congrats!

[GoldenOne]

Audio precision is allowing me to keep this unit until mine is ready. I'll actually be going to visit Audio Precision in person to drop this unit off and pick mine up which should be fun.

Dummy Load arrived today, this is controllable from the APx software so I can incorporate it into fully automated test sequences, and not only does it do pure resistive loads, but also does proper reactive load testing too so this should allow for some very comprehensive testing.

There's been an interesting thread about resistive/test condition measurements vs reactive/real world stuff lately so hopefully this can help with some similar stuff: https://www.superbestaudiofriends.org/index.php?threads/oh-shit-sinad-can-suck-my-s.11175

Also I've put my ifi iGalvanic on the test bench, and an iUSB 3.0 (thank you ifi) to ensure that no DAC measurements are skewed by the source itself. It's actually been quite interesting just how some dacs perform differently with one USB source vs another

[GoldenOne]

 

[The Computer Audiophile]

1 hour ago, GoldenOne said:

 

Love it @GoldenOne!

 

Your approach is fantastic. 

[GregWormald]

17 hours ago, The Computer Audiophile said:

Love it @GoldenOne!

 

Your approach is fantastic. 

Agreed.

It's very important to realise that not everything has been measured, and while our senses can be fooled they aren't always.

[numlog]

On 7/15/2021 at 2:55 PM, GoldenOne said:

Also I've put my ifi iGalvanic on the test bench, and an iUSB 3.0 (thank you ifi) to ensure that no DAC measurements are skewed by the source itself. It's actually been quite interesting just how some dacs perform differently with one USB source vs another

Do you have any data you can share from this? there was a thread a while back that was unsuccessful in showing differences created by different sources or source conditions with a few dacs, it would be an good and interesting counter.

 

Good to have another alternative to ASR although I never had a real problem with their measurements, they are usually pretty thorough. It's definitely wise to just ignore all of the added subjective commentary and ''conclusions'' from them.

 

So your aim is to use measurements to explain why certain equipment sounds the way it does VS pretending/assuming it all sounds the same aka audio 'science' review method.

Personally I dont really see this explaining much, I've tried to find connections between different measurements and how they influence how something sounds by comparing various, mostly budget, equipment.

There was never a consistent link between sound quality and various measurements of low level distortions, just taking an in-depth look at the circuit design will give you more hints about the sound quality ime (but since for that to work you have to understand how certain design choices might impact sound it is not very useful).

 

New or alternative measurements techniques are needed I think... I really have no clue what they would be, though maybe you do or it is your goal to find them?  

 

 

 

[Confused]

I watched the unboxing video yesterday. For me it was less of an unboxing, more of a mission statement from @GoldenOne.

 

And an excellent mission statement it is too, I think I pretty much agreed with every word.

 

I think that everything that can be heard almost certainly can be measured, but I suspect that in some cases the wrong things are being measured, and the right things not at all. The key is correlating measurements with the subjective experience, something that is not happening at the moment, in some areas at least.

 

I wish @GoldenOne the very best of luck with this new venture, it will be interesting to see how this develops.  

[Archimago]

22 hours ago, numlog said:

Do you have any data you can share from this? there was a thread a while back that was unsuccessful in showing differences created by different sources or source conditions with a few dacs, it would be an good and interesting counter.

 

Galvanic isolation of USB does make a difference and can be easily shown with complex measurement setups where there are potential interference and ground loops. Can't speak to the ifi specifically but can be shown with other inexpensive devices. [Note I'm talking true galvanic isolation, not whatever the JitterBug does 😉.]

 

22 hours ago, numlog said:

Good to have another alternative to ASR although I never had a real problem with their measurements, they are usually pretty thorough. It's definitely wise to just ignore all of the added subjective commentary and ''conclusions'' from them.

 

Subjective opinions can by definition be highly divergent. So not sure if one could say their "conclusions" are necessarily wrong regardless of whether one agrees or disagrees. I think you might be disappointed if you're looking for an "alternative" to ASR if it's the objective testing you're looking for pointing to different answers or something.

 

22 hours ago, numlog said:

So your aim is to use measurements to explain why certain equipment sounds the way it does VS pretending/assuming it all sounds the same aka audio 'science' review method.

 

Just have to be careful about painting ASR too broadly. There are some things I would highly disagree with Amir on but not everyone on ASR are of the same opinion as him either, just like not everyone here agrees to the same ideas. I don't think most "objectivists" are "pretending/assuming it all sounds the same". Rather, the testing leads many of us to that conclusion and further controlled listening as far as I can tell usually can confirm the measurement "experiments".

 

22 hours ago, numlog said:

Personally I dont really see this explaining much, I've tried to find connections between different measurements and how they influence how something sounds by comparing various, mostly budget, equipment.

There was never a consistent link between sound quality and various measurements of low level distortions, just taking an in-depth look at the circuit design will give you more hints about the sound quality ime (but since for that to work you have to understand how certain design choices might impact sound it is not very useful).

 

New or alternative measurements techniques are needed I think... I really have no clue what they would be, though maybe you do or it is your goal to find them?  

 

Just because we might believe a certain thing, or think we hear a certain thing does not mean it's there, right? Many people believe the Earth is flat because that's what they see every day. Many truths may not be intuitive like recognizing that Nyquist-Shannon sampling theorem works "perfectly", or that 16/44.1kHz has enough temporal resolution beyond anything tests of human time-domain perception have suggested regardless of whether you see advertisers use "stair-stepped" waveforms.

 

I'm curious, maybe you can suggest what budget equipment sound differences you're perplexed about specifically... I suspect those are some of the things @GoldenOne might be able to answer (and likewise myself if it's something I have access to).

 

[Miska]

42 minutes ago, Archimago said:

whether you see advertisers use "stair-stepped" waveforms.

 

Unfortunately, many DACs produce such using for example RedBook source. IOW, have fairly high measurable images, due to incomplete reconstruction...

 

[Archimago]

1 hour ago, Miska said:

 

Unfortunately, many DACs produce such using for example RedBook source. IOW, have fairly high measurable images, due to incomplete reconstruction...

 

 

Agree. 

 

The irony is that some in the 'hi-end' seem to desire this... 

[numlog]

2 hours ago, Archimago said:

 

Subjective opinions can by definition be highly divergent. So not sure if one could say their "conclusions" are necessarily wrong regardless of whether one agrees or disagrees. I think you might be disappointed if you're looking for an "alternative" to ASR if it's the objective testing you're looking for pointing to different answers or something.

I did not say they are wrong or right, I dont care what subjective things they have to say, I go there for objective data, it is supposed to be a 'science' forum.

An alternative source of measurements using the same tools is helpful weed out measurement errors, I would hope they all match... tbh I don't trust amir as being fully competent and/or honest 100% of the time so this is nice to have.

Quote

Just have to be careful about painting ASR too broadly. There are some things I would highly disagree with Amir on but not everyone on ASR are of the same opinion as him either, just like not everyone here agrees to the same ideas. I don't think most "objectivists" are "pretending/assuming it all sounds the same". Rather, the testing leads many of us to that conclusion and further controlled listening as far as I can tell usually can confirm the measurement "experiments".

I basically mean Amir when I say ASR, he is a strong influence and many do follow and spread the stuff he says blindly.

 

Quote

I'm curious, maybe you can suggest what budget equipment sound differences you're perplexed about specifically... I suspect those are some of the things @GoldenOne might be able to answer (and likewise myself if it's something I have access to).

Most of my experience is based on sighted listening, many will discount that (GoldenOne doesn't though from what I have seen) understandably so. Where possible and appropiate there was some controlled blind testing to have undeniable proof, not that specific differences exist necessarily, but that the ear can be reliable in sighted listening(only for myself, as I still cant proof of any of it)

Blind tests included USB cables (cheap printer vs supra), sources (desktop vs Rpi), some DAC tests (topping D50s, D50,  E30)

Being able to tell them apart in test was not important, the nature of the differences is what mattered.

For example D50s and D50 sound so distinctly similar, just as you'd expect as the circuits are very similar, while E30 had it own's (much preferable) character yet it's objective performance more closely matches the D50s. 

Sighted experiences with more DACs showed more distinct characteristics that could be not connected to the measurements.

It never made much sense to me anyway that minor differences in THD, IMD, noise etc. at extremely low levels could explain these distinct differences in character and levels of detail/resolution.

Then you have sources and USB cables, these shouldn't effect the analog measurements at all.

 

Is there any software DSP that can add different kinds of distortion in a controlled way to make 2 different dac measure almost identically? I bet they would not lose any of their distinction.

 

 

 

 

 

 

 

 

[fas42]

The value of improved measurement approaches will be in investigating the various types of noise induced and related distortion, which occur at low levels, and in complete systems, working in normal domestic environments - the test bench approach, of a single component, of the standard metrics, is very close to being completely useless for detecting subjectively disturbing anomalies. If one is practiced in the Art, 😆, then it is trivially easy to register these distortions just by listening to the output of a speaker driver - however, this ain't considered scieenteefic 🙂; so, developing a new set of input waveforms, on the right System Under Test arrangement, will certainly start getting us somewhere ...

[Miska]

3 hours ago, numlog said:

Is there any software DSP that can add different kinds of distortion in a controlled way to make 2 different dac measure almost identically? I bet they would not lose any of their distinction.

 

I would say you can use software to reduce distortions to make two difference DAC measure more similar. However, there are so many factors in play that it is very hard to make two DACs sound the same. And there will still be measurable differences too.

 

Most of the argument is about what is audible. ASR people think that none of the measurable differences are audible. Yet they still prefer to keep ranking list based on one single measured number and somehow think that defines the performance (yet not being audible).

 

[KenMoreira]

The people that live n die by THD and sinad need to slowly understand that their are more important numbers to human hearing.  And even still measurements of distortion don't tell a whole story either. 

 

To me, that logic would equate to, two headphones of different makes should sound the same with the same frequency curve.