Why you can't trust audio measurements

[andyr]

1 hour ago, MLXXX said:

As you would know, speaker load can affect audio power amplifier output as a function of frequency

 

Yes, strangely enough I do know this.  

 

1 hour ago, MLXXX said:

more so if the damping factor is low.

 

I have no idea what figure you regard as "low" - but both these amps are ss ... not tube amps having OPTs.

 

1 hour ago, MLXXX said:

Yes I think you mentioned that earlier in this thread, and I asked whether the graphs were of measurements of the two amplifiers driving the speakers, rather than a laboratory resistive load. You did not comment.

 

Actually, I only gave the HD graph of my Class A amp (not both amps).

 

But yes, the graph was with that amp driving a resistor - not my spkrs.

 

But both amps have been shown to be stable into 2 ohms - as they have been used with my Maggie 2 ohm true-ribbons.  Currently they have both been used to drive my 4 ohm, active 2-way spkrs - so 1x stereo amp for each spkr.

 

[andyr]

2 minutes ago, Satanica said:

My question, was each DAC levelled matched?

Louder might give the psychoacoustic impression of being bigger (wider).

 

Can you state the basis of your suggestion Paul?

 

Sure, component comparisons when one component is deliberately set to be louder than the other ... shows up differences between the 2 components - but a wider soundstage is not one (difference) that I've heard mentioned, before.

 

[Satanica]

7 minutes ago, andyr said:

Can you state the basis of your suggestion Paul?

 

A level playing field is logical which is the kind of thing debates should thrive on.

Edited October 3, 2022 by Satanica

[acg]

8 minutes ago, Satanica said:

 

What's the loaded question?

 

My question, was each DAC levelled matched?

Louder might give the psychoacoustic impression of being bigger (wider).

 

Loaded in that I've not provided enough info to answer a question I don't really expect to be able to be comprehensively answered. 

 

Level matching was limited to all dacs being really, really loud.   

[Satanica]

1 minute ago, acg said:

Loaded in that I've not provided enough info to answer a question I don't really expect to be able to be comprehensively answered. 

 

Level matching was limited to all dacs being really, really loud.   

 

Your question can't be compreshensively answered if you don't provide comprehensive information and your experiment is performed in a comprehensive manner (level matching goes a long way towards this).

[andyr]

1 minute ago, Satanica said:

It wasn't a suggestion, it was a question. I end questions with a question mark.

 

Umm - actually, your post had 2 lines, Paul:

• Your first line was your question: "My question, was each DAC level-matched?"
• Your second line was a statement: "Louder might give the psychoacoustic impression of being bigger (wider)."

 

It was this statement that I asked you to state the basis for.

 

[Satanica]

4 minutes ago, andyr said:

 

Umm - actually, your post had 2 lines, Paul:

• Your first line was your question: "My question, was each DAC level-matched?"
• Your second line was a statement: "Louder might give the psychoacoustic impression of being bigger (wider)."

 

It was this statement that I asked you to state the basis for.

 

 

 

A level playing field is logical which is the kind of thing debates should thrive on

[andyr]

Just now, Satanica said:

A level playing field is logical which is the kind of thing debates should thrive on

 

Please answer my request - which was (in case you've forgotten):

------------------------------------------------------------------------------------

 

------------------------------------------------------------------------------------

 

[acg]

2 minutes ago, Satanica said:

 

Your question can't be compreshensively answered if you don't provide comprehensive information and your experiment is performed in a comprehensive manner (level matching goes a long way towards this).

 

I don't really know what info is required to be honest, mainly because I don't know why this happened.  Soundstage for this dac extended outside the speakers at moderate to high listening volumes... none of the other dacs did this. 

 

@Nadano dsp as far as I am aware, but I've not really looked up the dac topology to be sure how it functions. 

[Satanica]

3 minutes ago, andyr said:

 

Please answer my request - which was (in case you've forgotten):

------------------------------------------------------------------------------------

 

------------------------------------------------------------------------------------

 

 

I have Andy, please think on it. 

[andyr]

2 minutes ago, Satanica said:

I have Andy, please think on it. 

 

Ummm - AFAIAC Paul ... you haven't!

 

So why don't you simply spell it out for a dumb f*ck like me.

 

Please state the basis for your - what I would consider condescending - statement that "Louder might give the psychoacoustic impression of being bigger (wider)".

 

[Satanica]

1 minute ago, acg said:

 

I don't really know what info is required to be honest, mainly because I don't know why this happened.  Soundstage for this dac extended outside the speakers at moderate to high listening volumes... none of the other dacs did this. 

 

@Nada

 

If somone was to truly attempt to answer your question, then they would need every single measureable metric. It would require a scientific study. My comment about the potential for sources to output at different levels is a hypothesis. In your words you asked a loaded question and I suggested a possibility. 

[MLXXX]

1 hour ago, stereo coffee said:

Better implementation of the products power supplies, and board layout, than the other two.  Measurements might be able to discover it in terms of crosstalk   https://en.wikipedia.org/wiki/Crosstalk

With vinyl disc tonearm cartridges perhaps, but modern stereo DACs have such excellent cross-talk figures that the sound stage wouldn't be noticeably affected.

 

You'd need to look at other factors to explain a major difference in the subjective sound stage width or depth.

 

2 hours ago, acg said:

Recently, I've been auditioning three very high end dacs which all sound very nice but different (of course) and one of them throws up a soundstage about 30 degrees wider than the others

 

If you were able to record the output of that DAC and of one of your other DACs and upload to say Google Drive, you could share those differences in apparent sound stage with forum members.  (Would work best as a demonstration if forum members could also access an original version of the digital music file for comparison. Presumably the original file would not throw up an extra-wide sound stage, only the special DAC would do that.)

Edited October 3, 2022 by MLXXX

[stereo coffee]

59 minutes ago, Nada said:

 

The most obvious way to get wider imaging is using DSPs. They can work really well. Tricks include phase shifting fold back.

 

Does that extra wide sound stage DAC have internal DSP? 

 

Measurements to take to explore this if someone was very curious would be to extract  the digital stream post-DSP and pre-DA conversion.

and move away from accurate audio reproduction  ?

[acg]

18 minutes ago, MLXXX said:

If you were able to record the output of that DAC and of one of your other DACs and upload to say Google Drive, you could share those differences in apparent sound stage with forum members.  (Would work best as a demonstration if forum members could also access an original version of the digital music file for comparison. Presumably the original file would not throw up an extra-wide sound stage, only the special DAC would do that.)

 

Unfortunately, that is not possible.  All dacs are back where they came from and I am mulling over buying one of them, just not the one with the larger soundstage, so I doubt it will ever return, although I did quite enjoy it.

[Satanica]

2 hours ago, andyr said:

 

Ummm - AFAIAC Paul ... you haven't!

 

So why don't you simply spell it out for a dumb f*ck like me.

 

Please state the basis for your - what I would consider condescending - statement that "Louder might give the psychoacoustic impression of being bigger (wider)".

 

 

The microphone may say the frequency response is the same at different output levels. But our brain interprets it as different SINAD and more importantly as different frequency responses, think the Fletcher-Munson curves. And different frequency responses will affect the imaging perception. 

 

Regardless of whether you agree with the above or not, how can level matching be anything other than an advantage rather than a hindrance for comparison? It's a logical advantage. 

Edited October 3, 2022 by Satanica

[rand129678]

3 hours ago, acg said:

 

I'd love to actually know how to measure what we experience.  Recently, I've been auditioning three very high end dacs which all sound very nice but different (of course) and one of them throws up a soundstage about 30 degrees wider than the others (and also wider than the dacs here that I actually own).  I would love to know which measurement boxes need to be ticked to cause this because it is a lovely effect... 

 

EDIT:  and no, it is not the most expensive of the three, it is actually the least expensive.

 

Which 3 DACs if you don't mind sharing the models?

 

I'm curious to search the web to see what measurements are out there for all 3.

 

[acg]

7 hours ago, rand129678 said:

 

Which 3 DACs if you don't mind sharing the models?

 

I'm curious to search the web to see what measurements are out there for all 3.

 

 

There will not be measurements for all three online,  just for one of them.   Of the three other dacs that I own,  there are measurements for two of them.  So,  six dacs all up, five with a similar soundstage width (including Holo May KTE, Phasure Nos1a G3, diy ValveDac, Aries Cerat Helene, Aries Cerat Kassandra II Ref),  one that is wider (Mola Mola Tambaqui).

 

I listen in nearfield.  Normally the soundstage is drawn more or less between the speakers,  say across 120 degrees in front of me.   The Mola Mola draws it wider, say 150 degrees all up... yes,  I crows flapped it, a very precise angular measurement method reserved for special loonies... point at each edge of the soundstage and guess how far short of 180 degrees your arms swing. 

[stereo coffee]

13 hours ago, MLXXX said:

 

 

 

For example, all electronic circuitry at room temperature will be noisy.

 

 

Sorry to hear of that,   in stark contrast absolutely black silence here.

[MLXXX]

1 hour ago, stereo coffee said:

Sorry to hear of that,   in stark contrast absolutely black silence here.

 

I was not referring to audible noise but measurable electronic  noise. When testing for an electronic null the residual may be treated as practically zero if it is comparable with electronic noise. Typically at the listening chair ambient noise in the room exceeds noise coming from speakers.

 

See https://en.wikipedia.org/wiki/Johnson–Nyquist_noise

[Monkeyboi]

16 hours ago, MLXXX said:

For example, all electronic circuitry at room temperature will be noisy.

 

Absolutely true.   In audio we are generally not dealing with extremely low signal levels so thermal noise at ambient temperatures in informal audio discussions largely gets conveniently dismissed or simply omitted.  However in other fields like the design of ultra sensitive instrumentation amplifiers and in radio astronomy it's a big deal where you are actually dealing with sub picovolts and femtoamperes.  

 

Most students studying electronic engineering are aware of the common types of noise and when asked to list them all come up with the well known types but not all immediately list the omnipresent background thermal noise which IMHO is surprising and at the same time disappointing.  It's widely accepted in the scientific community that noise generated by the random movement of charge carriers generates electrical noise at temperatures above 0 deg. Kelvin (absolute zero) although this has been difficult to prove absolutely simply because we haven't yet been able to cool things here on Earth to 0 deg. K albeit we've come very close.  So the assumption of what happens at 0 deg. K can be reasonably theorised by extrapolation of existing data about what we currently know about this at temperatures above and just above absolute zero.  We've come close to cooling to absolute zero, but not quite there.

[El Tel]

11 minutes ago, Monkeyboi said:

We've come close to cooling to absolute zero, but not quite there.

 

These guys clearly haven't experienced how cold my missus gets when I come home from the pub 3 hours later than I promised.

[davewantsmoore]

17 hours ago, stereo coffee said:

No we don't have the bases covered. The article here examines our perception of frequencies beyond 20khz

http://www.townshendaudio.com/PDF/The-world-beyond-20kHz.pdf

Townshend say this:

 

"Many listeners hear a great difference when 20kHz band-limited audio signals are compared with wide band signals."

 

Time and time and time again, it is shown to be untrue.... and the tests which do show a difference, are shown to be poorly constructed  (eg. it was not the same content, with the only difference being the band limiting, being tested).

 

I honestly wish it were true, and I spent about 10 years trying to prove it was true..... because that would be a very good way to improv playback fidelity  (because rendering these high frequencies accurately is extremely difficult).

 

 

FWIW... nobody should take my word for it.   Everybody should test it if they are interested..... but be warned to test it properly.   For example putting on the CD and the the SACD, usually has other variables which mean you are not only listening to bandlimiting.

 

18 hours ago, stereo coffee said:

I ask, as it was a surprise to me to achieve good to very good reproduction of sub bass frequencies

Definitely.   

[davewantsmoore]

20 hours ago, rocky500 said:

But do we have the technology available and really know we have all the bases covered yet?

 

In some sense yes, and others no.

 

Sound waves can be measured with a zillion times more precision than can be perceived...... so if we have A and B, and we think they sound different we can say with extreme precision, what differences there are between them.

 

OTOH, if you have some (for example, tiny) difference between A and B ... is it audible.   Hard to say.   There is an extremely large body of knowledge of what to expect and what is important..... and people try all the time (as they do) to contradict this knowledge through experiment.

[davewantsmoore]

17 hours ago, acg said:

one of them throws up a soundstage about 30 degrees wider than the others

Something is different about the output signal of this DAC to the others.

 

Those output signals can be examined in great precision.... and the difference(s) can be quantified.

It's a misnomer that "we don't know how to measure things like soundstage in electronics".  There is only two signals for each DAC .... They are not the same between the DACs.... someone just has to look.