Master Thread: Are measurements Everything or Nothing?

[voodooless]

If I had to guess, I'd say a combination of very slight treble droop, and/or borderline audible distortion levels.

The slightness of the treble droop might not be so slight. It’s high dependent on the volume setting:

Which likely means it modulates this dependency with the signal as well. But probably I’m just rehashing what is already mentioned in the topic…

… what I always find disappointing is that after these blind tests show some difference, very little effort is made to figure out what actually caused this. As with Mythbusters, after you busted the Myth, you try to reproduce the results. Was it the distortion? Was it the frequency response (modulation)? If we don’t know we’ve really learned very little from the experiment.

 

[Blumlein 88]

The slightness of the treble droop might not be so slight. It’s high dependent on the volume setting:
View attachment 325877
Which likely means it modulates this dependency with the signal as well. But probably I’m just rehashing what is already mentioned in the topic…

… what I always find disappointing is that after these blind tests show some difference, very little effort is made to figure out what actually caused this. As with Mythbusters, after you busted the Myth, you try to reproduce the results. Was it the distortion? Was it the frequency response (modulation)? If we don’t know we’ve really learned very little from the experiment.

Yes, now that we have Deltawave it should be implemented whenever there is a positive result.

 

[voodooless]

Yes, now that we have Deltawave it should be implemented whenever there is a positive result.

That's just half the story though, Deltawave gives you the differences, it does not tell you which ones are audible.

 

[Blumlein 88]

That's just half the story though, Deltawave gives you the differences, it does not tell you which ones are audible.

If used in all of its capabilities, it will usually point to something. You can hear what is left that is different. You can see FR differences with ease using just music files. You can compensate for FR so if that dramatically lowers the null value, then you know FR is the only big difference. A few other tricks which is helpful in ways difficult to do otherwise.

 

[bodhi]

Ofcourse I can have the live performance as a reference - knowing it never can make it onto the recording in a very flawed 2-channel playback system.

True high fidelity with only two channels are only a wet dream in an audiophiles mind . Its a very bad copy of the real event in the concert hall. Its like making a movie without colours with 480 interlaced resolution on a 100 inch screen.

Ok, so you still must understand that what people are trying to do with perfectly measuring gear is get the best possible reproduction of the source material, whatever it is? When this is the goal then there is something to discuss at least.

Pondering about what kind of system would reproduce certain persons specific memory of some specific performance in some venue with specific set of conditions on listeners mood, place related to performers, intoxication etc. is completely useless.

As for using some device that supposedly creates just the right kind of coloration to make sound reproduction closer to aforementioned memory, well, I guess it's best to leave that at "you do you".

 

[Galliardist]

I presume you acknowledge a Benchmark pre-amp is functionally transparent. If that's the case, all else held equal in the test, and I could detect a difference in the blind test between it and a tube preamp, it suggests the tube preamp was adding some coloration. Correct? I passed the blind test, preferring the tube preamp. If, from the supplied stereophile measurements of this preamp, you can't figure out how it could possibly color the sound you can say so.

1) There is a difference that you know to look for, and can then prefer as you did when sighted, If I read this correctly, you had tested sighted first.
2) The standard objection I bring to this kind of test, including the ones I've done myself, is that you knew what was being tested. It's not a fully blinded test for that reason.

I'm happy to admit that you can hear a difference in these circumstances, but the value of the test to me or another reader isn't much more than your sighted difference, because of those two things.

It's confirmation of your sighted preference, but I don't see what else it tells us. Even if you preferred the Benchmark both sighted and blind... which would no doubt please more people here.

 

[pma]

The slightness of the treble droop might not be so slight. It’s high dependent on the volume setting:
View attachment 325877
Which likely means it modulates this dependency with the signal as well. But probably I’m just rehashing what is already mentioned in the topic…

… what I always find disappointing is that after these blind tests show some difference, very little effort is made to figure out what actually caused this. As with Mythbusters, after you busted the Myth, you try to reproduce the results. Was it the distortion? Was it the frequency response (modulation)? If we don’t know we’ve really learned very little from the experiment.

Good point, enough to be audible.

 

[GXAlan]

That's just half the story though, Deltawave gives you the differences, it does not tell you which ones are audible.

PK metric which is generated by DeltaWave does.

 

[voodooless]

PK metric which is generated by DeltaWave does.

No, it just tells you maybe something about audibility as a whole, not specifically what caused the audibility. Is it the distortion, the modulation, or the frequency response? PK metric will not tell you.

 

[GXAlan]

No, it just tells you maybe something about audibility as a whole, not specifically what caused the audibility. Is it the distortion, the modulation, or the frequency response? PK metric will not tell you.

It does show you what range of frequencies are accounting for the difference, which, when combined with the when, would allow you to narrow down the source of the audibility

 

[fpitas]

If old-fashioned means resistant to the modern trend whereby people haughtily assert their right to deny solid science whenever it is too inconvenient for them, and go on pushy lecture-to-the-forum campaigns where they cherry-pick odd data points and erect arguments-from-ignorance about what that data point 'means', then yes, you may be old-fashioned.

I'm old enough to remember when tube gear bragged about its low levels of distortion. At some point since then tubes have become somewhat of a farce.

 

[ads_cft222]

Question. For instance: You have 100 SINAD dac and 85 SINAD amp. What is the sinad that goes out of the speaker outputs of the amplifier? How can you know if it transforms to an audible or inaudible sinad? What is the minimum inaudible sinad in this case?

 

[voodooless]

It does show you what range of frequencies are accounting for the difference, which, when combined with the when, would allow you to narrow down the source of the audibility

That only works if you accept the PK metric as trustworthy. So far, I'm inclined not to do that yet. It needs way more validation for that.

 

[voodooless]

Question. For instance: You have 100 SINAD dac and 85 SINAD amp. What is the sinad that goes out of the speaker outs? How can you know if it transforms to an audible or inaudible sinad? What is the minimum inaudible sinad in this case?

Your speakers have distortion orders of magnitude larger than these two amps. It will most probably mask the amps distortion completely.

 

[ads_cft222]

Your speakers have distortion orders of magnitude larger than these two amps. It will most probably mask the amps distortion completely.

Does the following relationship holds? Dac distortion = a% , amp distortion = b% and speakers distortion = c%. Final distortion = a% x b% x c%? If this product holds true and c% is the bottleneck , could we extrapolate that maybe the distortion in isolation of a and b can be inaudible but in the final product we need the lowest possible for a and b?

 

[voodooless]

Does the following relationship holds? Dac distortion = a% , amp distortion = b% and speakers distortion = c%. Final distortion = a% x b% x c%?

Not quite. The distortion profiles of these devices will not be the same. So the resulting addition is quite complex.

If this product holds true and c% is the bottleneck , could we extrapolate that maybe the distortion in isolation of a and b can be inaudible but in the final product we need the lowest possible for a and b?

In theory yes, in practice with well-designed electronics, that is unlikely. As I said, the differences are generally orders of magnitude apart.

 

[HarmonicTHD]

Does the following relationship holds? Dac distortion = a% , amp distortion = b% and speakers distortion = c%. Final distortion = a% x b% x c%? If this product holds true and c% is the bottleneck , could we extrapolate that maybe the distortion in isolation of a and b can be inaudible but in the final product we need the lowest possible for a and b?

No.

Somewhere in the forum you can find the mathematical correlation and its discussion. Sorry can’t find it for you on the run.

Edit. Oops overlap with @voodooless

 

[fpitas]

Does the following relationship holds? Dac distortion = a% , amp distortion = b% and speakers distortion = c%. Final distortion = a% x b% x c%? If this product holds true and c% is the bottleneck , could we extrapolate that maybe the distortion in isolation of a and b can be inaudible but in the final product we need the lowest possible for a and b?

What those guys said: it's complicated. I'll comment, if you're close to the hearing threshold of about 70dB SINAD in any component, you have serious equipment problems.

 

[Killingbeans]

Does the following relationship holds? Dac distortion = a% , amp distortion = b% and speakers distortion = c%. Final distortion = a% x b% x c%? If this product holds true and c% is the bottleneck , could we extrapolate that maybe the distortion in isolation of a and b can be inaudible but in the final product we need the lowest possible for a and b?

Root sum squared (and gain factors):

The Signal Chain: How do noise and distortion propagate through my system?

This article is to explore how noise and distortion change through our system, and how different components can influence the final output. A common question is “If I have DAC A, preamp B, and amp C, where should I set the volume and gain controls for the best performance?” Usually this...

www.audiosciencereview.com

 

[fpitas]

Root sum squared:

The Signal Chain: How do noise and distortion propagate through my system?

This article is to explore how noise and distortion change through our system, and how different components can influence the final output. A common question is “If I have DAC A, preamp B, and amp C, where should I set the volume and gain controls for the best performance?” Usually this...

www.audiosciencereview.com

Kind of. Phase matters, so you can actually get true addition and even cancellation.