• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

General debate thread about audio measurements

sergeauckland

Major Contributor
Forum Donor
Joined
Mar 16, 2016
Messages
3,440
Likes
9,100
Location
Suffolk UK
So, in summary, there is no mechanism by which the sound coming out of the speaker can be changed in any way that may be sensed by some little understood brain/ear phenomenon if there is demonstrably no way in which the electrical wiggle generating the pressure wiggle we sense has been changed - that is the nub of my point and the measurements we can do today do show whether that has been achieved or not.

Exactly so. It is trivially easy to check how much or how little a signal has changed between input and output, no need for any reference to ear/brain. DACs, digital players generally and amplifiers can be and mostly are so linear as to be totally transparent in normal operation. If one prefers the sound of something non-linear, then fine, that's a personal preference, but then it's an effects box not a DAC or an amplifier.

S
 

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
Differences between signals/sounds can be detected and measured. If the difference is below the threshold of audibility it can't be heard. What is difficult about that? All the ifs and buts will not change a thing except help doubters cling to their erroneous claims.

We are right, subjectively, prove us wrong doesn't go anywhere.

One example: https://en.wikipedia.org/wiki/List_of_religions_and_spiritual_traditions Each says: "We are right. Prove we're wrong!".
 
Last edited:

rebbiputzmaker

Major Contributor
Joined
Jan 28, 2018
Messages
1,099
Likes
463
If the difference is below the threshold of audibility it can't be heard.
This is not necessarily correct. It might be nice if you start using "IMO" where applicable. You tend to speak in absolutes, and sorry, but your opinions might absolute for you, but not necessarily for all.
 

rebbiputzmaker

Major Contributor
Joined
Jan 28, 2018
Messages
1,099
Likes
463
Not only is it correct, it's tautological.
Really, so inaudible high-level harmonics issues cannot cause audible issues (instability etc) further down the audio chain? i.e. preamp to power amp. I think you are oversimplifying.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
Yes, that must be it.

/looking for the rolling eyes smiley.
 

rebbiputzmaker

Major Contributor
Joined
Jan 28, 2018
Messages
1,099
Likes
463
Yes, that must be it.

/looking for the rolling eyes smiley.
Glad to know there are so many experts here.

Looking for the flat earth map, I know it is around here somewhere.
 

sergeauckland

Major Contributor
Forum Donor
Joined
Mar 16, 2016
Messages
3,440
Likes
9,100
Location
Suffolk UK
This is not necessarily correct. It might be nice if you start using "IMO" where applicable. You tend to speak in absolutes, and sorry, but your opinions might absolute for you, but not necessarily for all.
Rubbish. If it's below the threshold of audibility, it can't be heard. That's what the threshold of audibility is!

S.
 

sergeauckland

Major Contributor
Forum Donor
Joined
Mar 16, 2016
Messages
3,440
Likes
9,100
Location
Suffolk UK
Really, so inaudible high-level harmonics issues cannot cause audible issues (instability etc) further down the audio chain? i.e. preamp to power amp. I think you are oversimplifying.
No they can't. If they are of sufficiently low level that they're inaudible, then they won't cause audible issues. Instability has nothing to do with harmonic distortion.

In any event, any decent amplifier should be unconditionally stable into any load and at any frequency. There are plenty that are, so there's no excuse these days if they're not. A few manufacturers make a feature that their amps are not unconditionally stable, so they need special loudspeaker cable, but I don't see those as serious products.

S.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,384
Location
Seattle Area
Really, so inaudible high-level harmonics issues cannot cause audible issues (instability etc) further down the audio chain? i.e. preamp to power amp. I think you are oversimplifying.
Those audibility issues will be measurable if they are in audio band.
 

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
Hi,
I know you moved this debate with me here but I am in England, so most of yesterday here there was nothing for me to look at at all and this morning I wake to a large number of posts which take a long time to digest!
My point is this. Whatever the mental or ear brain response is like may be fascinating but in this context it is not relevant, apart from any expectation bias or placebo effect.
This is why:
Before it gets to the speakers all the sound is is an electrical wiggle.
It is produced first at the microphone, of course, but that is usually manipulated by recording and mixing engineers to produce the LP, CD, file we buy.
Next we play it back at home. We, or at least I :), am fully conversant with all the many ways in which different record players can and do produce different electrical wiggles from the same LP, so lets ignore LPs for now.
In the case of a DAC the "end result" isn't whether it is a ladder dac has different filters or analogue sections it is "what is the electrical wiggle coming out of the analogue outputs, and how close is it to what it should be?"
This is also the case for the preamp and power amp. They can be treated as a black box with a transfer function.
If the transfer function of the "black box" is exactly as it should be, then there is no mechanism by which it could possibly actually sound any different to any other with the same transfer function, is there?
The electrical wiggle is unchanged so the air pressure wiggle from the speakers created by it is unchanged too, so how our brain may intereact with it can only percieve a difference by expectation bias or the placebo effect.
There is, of course a caveat. This is interaction between power amps and speakers. What I wrote above is the case for properly engineered DACs an preamps. Power amps have interaction with speakers which means that changing the power amp into the same speaker may well change the air pressure wiggle. This can be measured too though and show the changes/deviations from how it should be.
It is debateable exactly where distortion and signal to noise starts being audible, so I quite agree that the level at which deviation from perfect transfer function becomes audible is an unknown, so here it is just a case of preferring less deviation to more as a matter of principle IMO.
In the case of speaker/power amp interaction there is clipping and other sorts of overload into the actual speaker rather than an 8 ohm resistor that is obviously relevant to whether a system sounds different or not, but also the effect of a high output impedance of an amp on the complex impedance of actual speakers. This obviously is both amp and speaker dependant but since we know that changes in level of 1dB are definitely detectable, and very possibly much less by practiced listeners, it is probable, even likely, that an amp with a sufficiently high output impedance into speakers with a sufficiently complex impedance will result in sound level and frequency response changes which are audible, in that even though the speaker/room FR won't be flat the change created in it should be audible.
I know the ear has a very high dynamic range, but as a person who has made many recordings over the last 55 years it has never been necessary for me to record all of it at the same time, so having a level control on my recorder to get the recording into the "good bit" of a restricted dynamic range has always been possible (though more difficult on a reel-to-reel tape since it is quite marginal for classical music in a quiet environment).
Personally I agree with you that with music dynamic range and masking may well make most distortion inaudible, but it might not :) and the point is, why, since accurate stuff is available at modest prices, take the risk?

So, in summary, there is no mechanism by which the sound coming out of the speaker can be changed in any way that may be sensed by some little understood brain/ear phenomenon if there is demonstrably no way in which the electrical wiggle generating the pressure wiggle we sense has been changed - that is the nub of my point and the measurements we can do today do show whether that has been achieved or not.
Thanks for your response. I visited England once. I saw the Buckingham palace from outside. :)
I see what you are saying. That's a concept that's given me pause too. We have signals that are already beyond the threshhold of audibility and even if they aren't can impact the ear very little. So if the signals are below audbilty what does it matter. For me that's a strong argument. However even very cheap gear can reach this level. So even though I can't prove it, it's never proved true for me that things sounded the same.

I'm willing to venture something that contradicts what I usually say about audbility. That is technically the two signals aren't the same. They are different. It's hard for me to imagine that would make a difference though due to audibility.
But the standard measurements we make tend to be measuring single tones. Except perhaps the IMD test. Perhaps something changes during music playback. I know the DACs/preamp impedance also can matter. The Yggy first generation didn't do too well with 600 ohm impedance. A lot of dacs now are used as preamps to power amps as well.
I don't see a whole lot about timing in standard measurements. I think maybe we see impulse response.

I had a question since you seem to have been involved in record players. I notice that LP's do not have ringing or a digital filter they have to go through while (most) digital audio does. The LP players do have an RIAA filter and other distortions but they don't have a digital filter and ringing. Also unlike DS DACs they don't have things like taps and ways to approximate the waveform produced. What are your thoughts on that?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,384
Location
Seattle Area
The brain is a very complex organ and one we know much less than we think, and one we are still learning about.
Perception of audio has been subject of intense investigation for decades. Back in 1944 Fletcher and Munson for example ran tests to detect the frequency response of the ear:

robinson-dadson-curve.png


That bottom line for example shows the lowest levels we can hear with respect to frequency.

As noted, we don't need to know the "why" (even though we do), to know that the hearing threshold is what it is.

After all, if it is "what we hear that matters" then that is what the above research is all about. :) We test listeners and look for patterns that are reliable and durable across the population.

Also, it is important to read and understand the research in this area before declaring that we don't know enough. How credible would it be for me to declare that we know less than we think about some disease. Without qualification in the field, of course we "don't know" what we are saying. That is no fault of the science.

Let's remember that we can reduce the size of a music file to less than 10% of it which maintaining transparency for vast swath of the population. That is a pretty neat trick if we have so little idea of how we hear!
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
Let's remember that we can reduce the size of a music file to less than 10% of it which maintaining transparency for vast swath of the population. That is a pretty neat trick if we have so little idea of how we hear!
Not necessarily. The simple fact of viable data reduction could just be to do with the way music only partially occupies 'the information space' and finding a more efficient way to store it. After all, FLAC manages 2:1 losslessly without any knowledge whatsoever of how human hearing works. Allowing a little bit of degradation might conceivably increase that ratio, still without needing any knowledge of how hearing works.
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
Differences between signals/sounds can be detected and measured. If the difference is below the threshold of audibility it can't be heard. What is difficult about that? All the ifs and buts will not change a thing except help doubters cling to their erroneous claims.

We are right, subjectively, prove us wrong doesn't go anywhere.

One example: https://en.wikipedia.org/wiki/List_of_religions_and_spiritual_traditions Each says: "We are right. Prove we're wrong!".

You should see the irony as your own " belief framework " (partly cited above) let you express exactly the same.

The argument "if the difference is below the threshold of audibility it can´t be heard" is of course correct, but i´m sure you didn´t mean the individual thresholds auf audibility that you most likely don´t know, but the usual list for thresholds of audibility.

You are asking "what is difficult about that" and the answer is that your categorical statement isn´t warranted.

It is just a hypothesis that "it" then "can´t be heard" and it is based on premises that might be incorrect.

Even if you think it is a plausible hypothesis (which it is obviously) it nevertheless is already from a statistical point of view questionable.
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
@amirm,

the graph illustrates already why knowledge about the "why and how" would be/can be so important. See the differences between the blue and the red curve/line.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,384
Location
Seattle Area
Not necessarily. The simple fact of viable data reduction could just be to do with the way music only partially occupies 'the information space' and finding a more efficient way to store it. After all, FLAC manages 2:1 losslessly without any knowledge whatsoever of how human hearing works.
That is roughly the limit of entropy reducing compression (2.5:1 at the extreme). All lossy codecs actually have a back-end that does the same thing. You just can't get 10:1 compression that way without perceptual analysis and removing what is thought to not be inaudible.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,384
Location
Seattle Area
@amirm,

the graph illustrates already why knowledge about the "why and how" would be/can be so important. See the differences between the blue and the red curve/line.
How so?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,384
Location
Seattle Area
Sorry but i don´t understand to which part of my post the "How so?" question is related.
That the difference between blue and red lines is due to differing understanding of the brain function.
 

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
This is not necessarily correct. It might be nice if you start using "IMO" where applicable. You tend to speak in absolutes, and sorry, but your opinions might absolute for you, but not necessarily for all.

"not necessarily", "might be nice", "start using 'IMO', and précised 'put your views to suit me'. Rebbi(p), be more specific and less garrulous.
 
Top Bottom