"Things that cannot be measured"

[Jimbob54]

In my very humble opinion, what cannot be measured is the individual preference.

I have a pair of RP-600M, and while I rationally know that they have bad specs, I can't help keep using them, at least for rock and pop, because I like that sound despite some of the problems that they have and that I can hear quite clearly.
On the other hand I had the opportunity to listen to various models of Neumann monitors in many occasions, and I find them flipping annoying! Completely lifeless! I understand a sound engineer may need that type of sound, but they are terrible if you listen for pleasure.

So measures will tell you how a product will sound, but you may just not like the sound.

I know Amir doesn't agree because there is research that shows how everybody tends to like the same sound in a blind test, so maybe I am 5 sigmas out.

Yes , but you can have a very good go at measuring what factors cause you to prefer the RP to the Neumann. And when you need to buy a new pair, if you knew how the RP measured , you could ease your shortlisting for the next purchase by looking for similar measurements.

Which I think is what @scott wurcer was saying.

 

[Pennyless Audiophile]

Yes , but you can have a very good go at measuring what factors cause you to prefer the RP to the Neumann. And when you need to buy a new pair, if you knew how the RP measured , you could ease your shortlisting for the next purchase by looking for similar measurements.

Which I think is what @scott wurcer was saying.

Which is a good point, I agree, but then it casts a shadow on the language we use.
Should we keep saying that Neumann is great and recommend it because of their linearity, or should we rather say that they are great IF you like absolute linearity, otherwise look elsewhere?

 

[Jimbob54]

Which is a good point, I agree, but then it casts a shadow on the language we use.
Should we keep saying that Neumann is great and recommend it because of their linearity, or should we rather say that they are great IF you like absolute linearity, otherwise look elsewhere?

As long as you know the parameters of the review/ recommendation , why does it matter? We know what Amir is looking for in electronics and transducers and its (usually) clear from the measurements what drives the recommendation. If you know your preference deviates from his/ any given target, you can take what you need from the charts.

This isnt music, where the review can only really be based on perception and prior experience and an attempt to convey in words why you like or not. Something some audio reviewers either fail to grasp or choose to ignore.

 

[mocenigo]

There are no things that cannot be measured. Saying otherwise is bollocks.

There are many things we imagine. And there are just a few things that we just happen not to measure, or do not want to measure. In fact the single-digit SINAD has the explicit purpose to distract from a lot of other things that would be easy to verify but would be "distracting" to the single-value-truth doctrine (as the multi-tone test was before Amir was persuaded to perform it as well).

Sometimes we bury our heads (and ears) in the sand, like some people here that remind me that 16-bit 44100-hertz recordings can capture and represent phase delays of a few picoseconds. True, of course please say at which frequencies, and tell me the resolution for very high frequencies (cough) – and stop reminding me of this each time I recall you cannot determine the *start* of a transient with a higher precision than 1/44100 of a second, a value that is much bigger than the resolution at which the brain can process time differences. Because this is a completely different matter (in fact, given a specific ADC topology, it is a simple mathematical exercise to devise arbitrarily many pairs of L/R signals that start at relative time intervals much smaller than 1/44100 sec and from a given point in time are otherwise identical, and would be sampled at 44.1Khz with exactly the same values. The brain would be able to perceive the time differences as different positions in space for the original signals, and since there would be just one pair of sampled signals, no reconstruction filter would be able to differentiate them).

But everything the ears can hear can be measured. We just need to want to do it.

 

[mansr]

Sometimes we bury our heads (and ears) in the sand, like some people here that remind me that 16-bit 44100-hertz recordings can capture and represent phase delays of a few picoseconds. True, of course please say at which frequencies, and tell me the resolution for very high frequencies (cough) – and stop reminding me of this each time I recall you cannot determine the *start* of a transient with a higher precision than 1/44100 of a second, a value that is much bigger than the resolution at which the brain can process time differences. Because this is a completely different matter (in fact, given a specific ADC topology, it is a simple mathematical exercise to devise arbitrarily many pairs of L/R signals that start at relative time intervals much smaller than 1/44100 sec and from a given point in time are otherwise identical, and would be sampled at 44.1Khz with exactly the same values. The brain would be able to perceive the time differences as different positions in space for the original signals, and since there would be just one pair of sampled signals, no reconstruction filter would be able to differentiate them).

I could agree with you, but then we'd both be wrong.

 

[scott wurcer]

Why would you think that "assuring that the sound field presented to each person is the same as possible" is something desirable baffles me. If you like your speakers and I don't, who's "right"?

Listening preference does not fit here, I don't believe in using measurements to tell people what they should like. If you compare the speakers with a SET amp on one and a Benchmark on the other some pretty easy to do measurements can show that in all likely hood you are not doing a fair comparison. of the speakers.

 

[Blumlein 88]

There are no things that cannot be measured. Saying otherwise is bollocks.

There are many things we imagine. And there are just a few things that we just happen not to measure, or do not want to measure. In fact the single-digit SINAD has the explicit purpose to distract from a lot of other things that would be easy to verify but would be "distracting" to the single-value-truth doctrine (as the multi-tone test was before Amir was persuaded to perform it as well).

Sometimes we bury our heads (and ears) in the sand, like some people here that remind me that 16-bit 44100-hertz recordings can capture and represent phase delays of a few picoseconds. True, of course please say at which frequencies, and tell me the resolution for very high frequencies (cough) – and stop reminding me of this each time I recall you cannot determine the *start* of a transient with a higher precision than 1/44100 of a second, a value that is much bigger than the resolution at which the brain can process time differences. Because this is a completely different matter (in fact, given a specific ADC topology, it is a simple mathematical exercise to devise arbitrarily many pairs of L/R signals that start at relative time intervals much smaller than 1/44100 sec and from a given point in time are otherwise identical, and would be sampled at 44.1Khz with exactly the same values. The brain would be able to perceive the time differences as different positions in space for the original signals, and since there would be just one pair of sampled signals, no reconstruction filter would be able to differentiate them).

But everything the ears can hear can be measured. We just need to want to do it.

Well you apparently don't wish to know, but everything you wrote there is incorrect. And can be demonstrated to be so. Want to learn something new?

 

[Duke]

In terms of what @CMOT is saying about pitch or even arguably frequency being more perceptive than measurable at times, I think I am pretty sympathetic to that point of view.

Yeah psychoacoustics includes a bunch of little counter-intuitive twists and turns. One of them has your name on it, or at least part of your name on it:

Steven's Rule says that the perceived pitch of a high frequency tone goes UP a little bit as SPL is increased, and that the perceived pitch of a low frequency tone goes DOWN a little bit as the SPL is increased. I have heard that this implies "correct" playback levels may exist for classical music. The correct playback level for Spinal Tap is, of course, Eleven.

(I used to think I knew how to pronounce the word "Steven", and I used to think I knew how to pronounce the word "Eleven". When I try to read your username out loud, I'm not so sure anymore.)

 

[jonfitch]

Well a spinorama is never going to show you how dynamic a speaker is. So it's not so much that there are no measurements, rather the standard measurement suite doesn't give you a full picture. For example, if you just look at the headphone/IEM space, some driver typographies displace a ton of air while others do much less--BA basically displaces no air, whereas dynamics displace the most. A FR graph really won't tell you anything about a dynamics and punchiness in this regard.

 

[mansr]

In the context of sound reproduction, it is important to remember that everything on a recording _has been_ measured, by definition.

 

[mansr]

Steven's Rule says that the percieved pitch of a high frequency tone goes UP a little bit as SPL is increased, and that the perceived pitch of a low frequency tone goes DOWN a little bit as the SPL is increased.

That is unsurprising if you think about it.

 

[Blumlein 88]

Yeah psychoacoustics includes a bunch of little counter-intuitive twists and turns. One of them has your name on it, or at least part of your name on it:

Steven's Rule says that the percieved pitch of a high frequency tone goes UP a little bit as SPL is increased, and that the perceived pitch of a low frequency tone goes DOWN a little bit as the SPL is increased. I have heard that this implies "correct" playback levels may exist for classical music. The correct playback level for Spinal Tap is, of course, Eleven.

(I used to think I knew how to pronounce the word "Steven", and I used to think I knew how to pronounce the word "Eleven". When I try to read your username out loud, I'm not so sure anymore.)

Well, as long as nobody used compression on your classical music recording. There is also a bit of a connection between how loud something sounds and the size of the space it is heard in. Which would interact with the Steven Rule if you are playing back your big orchestral recording in a domestic living room sized space.

 

[BinkieHuckerback]

Surely measurements don't say what is 'right' or 'wrong' but will tell you whether there's actually a difference that you (a 'normal' - fits in the Bell curve of what's 'usual') ought to be able to discern?

 

[Frontwave]

If you are talking audio waveforms, then yes.

Then any differences in perceived sound quality say..from blind testes on DACs.....is all in peoples heads? It's all subjective? OR maybe there are some other components in the DAC chain that might make one "sound" better the the other?

 

[symphara]

Listening preference does not fit here, I don't believe in using measurements to tell people what they should like. If you compare the speakers with a SET amp on one and a Benchmark on the other some pretty easy to do measurements can show that in all likely hood you are not doing a fair comparison. of the speakers.

I wasn't disputing that at all. This thread presents a trivial question, with a trivial answer. Yes there are things that cannot be measured, at least not very well. And measurements we haven't thought of.

 

[Jimbob54]

Then any differences in perceived sound quality say..from blind testes on DACs.....is all in peoples heads? It's all subjective? OR maybe there are some other components in the DAC chain that might make one "sound" better the the other?

In which case it's measurable

 

[Emlin]

I wasn't disputing that at all. This thread presents a trivial question, with a trivial answer. Yes there are things that cannot be measured, at least not very well. And measurements we haven't thought of.

Name one measurement that we haven't thought of!

 

[BinkieHuckerback]

I wasn't disputing that at all. This thread presents a trivial question, with a trivial answer. Yes there are things that cannot be measured, at least not very well. And measurements we haven't thought of.

What metrics might there be that 'we haven't thought of'?

 

[mansr]

What metrics might there be that 'we haven't thought of'?

Synergy.

 

[BinkieHuckerback]

Synergy.

Then we've thought of it. How would you measure it? (If something has/is 'synergy', then surely it can be distinguished from something that has/hasn't got 'synergy')?