Is SINAD important? - "Myths" about measurements! [Video YT]

[ahofer]

I am left scratching my head in wonder of whether you are joking or serious.......

Charges?
I merely said out in the open that many value SINAD quite a bit. That is COMMON KNOWLEDGE, and well known by everyone on the forum.
For me to have to prove that something happens, that we all know happens.......................

The forum is openly mocked on videos, and other audio forums, for the SINAD thing quite frequently.
Therefore, no charge was leveled, but just a simple truth, that NEEDS NO proof.......

Many in the forum use SINAD as a line in the sand for Junk/Acceptable/Great products.

Hell, that is what the forum is known for.....

Argumentum ad Populum

If it is so obvious and well-known, you should easily be able to cite a few examples. People say all sorts of stuff, often because they feel threatened by what Amir is doing here. You can facepalm all you like but you are the one who has yet to make a valid point.

More likely, as well illustrated here above, people use the SINAD measurement as a proxy for *engineering* quality, and being a science and engineering-oriented site, we tend to condescend to electronics manufacturers who can’t achieve easy levels of engineering quality. ASR’s detractors see that and conflate it with a judgement of *sound* quality, as you have. So we see the accusation a lot, and it seems like a strawman. Certainly most people I see contributing regularly here wouldn’t make the mistake of confusing those two things.

This seems like a good opportunity to promote the Ideological Turing Test. It’s a useful tool to think about whenever you see yourself saying “they all say” or “so your point is”

https://www.thefire.org/sites/default/files/2019/10/14142634/The-Ideological-Turing-Test.pdf

 

[Chagall]

This one seems like a good article about it but I cannot get into the references to dig deeper.
It seems around the 16-bit mark though with a very rough look,at least for best case young people.

Audibility thresholds for SINAD/THD measurements

Noise perception, detection threshold & dynamic range

Noise audibility threshold & 'subjective dynamic range' of 16/44.1k and some popular formats... Signal-to-noise ratio, segmented signal-to-noise ratio, weighting curves, noise perception...

www.tonestack.net

There is also a 47-page thread on thresholds here on ASR. So unless Rob Watts, use them as guidelines:

Audibility thresholds of amp and DAC measurements

Introduction A recurring theme on ASR is whether or not the various measured qualities of the devices are audible. In this post, I'll present some clear and visual thresholds for when those imperfections can be considered a potentially audible concern. I will not explain the basics of amp/DAC...

www.audiosciencereview.com

 

[Tell]

What is the research to back that?

I got none, but this thread is a reference at least, or you could go with numbers on klippel.de where it seems that people even have a hard time pick out distortion at -30dB. Of course noise is missing there, but it's an indicator that the thresholds are way lower than what people think.
But anyways, SOME kind of disclaimer with also maybe some audibility number is way better than nothing, since people really are misinterpreting SINAD and that ranking no matter what it actually represents.

 

[solderdude]

The conclusion is there is no single audibility threshold.
Also not for SINAD (1kHz at specified level).

The threshold depends on many factors including psychoacoustics.
The latter is the difficult one with the biggest variations in it.

From a technical stand point lowest noise, lowest distortion, lowest phase deviations, flat frequency amplitude response (with cut-off points specified) outside of the audible range is most desirable (highest SINAD with wide/flat frequency response).
From a 'sound quality' stand point .... it depends... both high and low numbers may well sound fine depending on metrics other than SINAD.

The explanation for this was touched upon but not clarified in the video. Instead the focus was on SINAD and Harman target compliance strictness.
Those are the 2 'things' people think about when commenting on ASR posts/reviews.

IMO these aspects were singled out way too much and a broader addressing of the gap between perception and measurements would have been far more educational.

 

[JustJones]

I don't see any need for a disclaimer. If people come here and glance at the SINAD chart then go away saying all these people do is listen to graphs and charts without bothering to read through the reviews and see that's not the case then I really don't care. They're not using good faith but handwaving bunk because that's their agenda.

 

[Tell]

The conclusion is there is no single audibility threshold.
Also not for SINAD (1kHz at specified level).

The threshold depends on many factors including psychoacoustics.
The latter is the difficult one with the biggest variations in it.

From a technical stand point lowest noise, lowest distortion, lowest phase deviations, flat frequency amplitude response (with cut-off points specified) outside of the audible range is most desirable (highest SINAD with wide/flat frequency response).
From a 'sound quality' stand point .... it depends... both high and low numbers may well sound fine depending on metrics other than SINAD.

The explanation for this was touched upon but not clarified in the video. Instead the focus was on SINAD and Harman target compliance strictness.
Those are the 2 'things' people think about when commenting on ASR posts/reviews.

IMO these aspects were singled out way too much and a broader addressing of the gap between perception and measurements would have been far more educational.

Yes of course, but since people CLEARLY are constantly misinterprets SINAD and that blue/green/yellow/red rating then something is better than nothing isn't it? And no maybe we don't need some audibility marker, but at least a short and clear disclaimer of what ranking actually represents.

I don't see any need for a disclaimer. If people come here and glance at the SINAD chart then go away saying all these people do is listen to graphs and charts without bothering to read through the reviews and see that's not the case then I really don't care. They're not using good faith but handwaving bunk because that's their agenda.

People is always going to be people, and if they're not technical enough to understand everything (and don't have the time/energy/whatever to learn) then why not try to be at least a bit clearer to help them which in the long term might help ASR?

 

[beagleman]

I don't see any need for a disclaimer. If people come here and glance at the SINAD chart then go away saying all these people do is listen to graphs and charts without bothering to read through the reviews and see that's not the case then I really don't care. They're not using good faith but handwaving bunk because that's their agenda.

Here is a question I have had.
I asked Amir a few times, quite a while ago, and never got a reply ( I assumed he missed my post or what have you )

I asked Amir, if when measuring product "X" , that had fairly mediocre SINAD, if he had actually listened to the unit in operation, and had "heard" the actual output of the unit or based his assessment strictly on just a measured number?

It may have been A DAC or possibly some random AVR, but too far back to remember specifics.

I questioned him, only cause I did not see any subjective impressions.
I guess I expected to see some mention of "As the SINAD measured mediocre, I was able to hear/not hear the limitations" in his review.

And yeah I get bad is bad, and great is great, but that middle ground, SINAD can be various combinations of distortion and noise, and not all noise, has the exact same audibility, and often when mixed in with an actual signal of varying frequency, it is masked in different ways.

 

[JustJones]

Yes of course, but since people CLEARLY are constantly misinterprets SINAD and that blue/green/yellow/red rating then something is better than nothing isn't it? And no maybe we don't need some audibility marker, but at least a short and clear disclaimer of what ranking actually represents.


People is always going to be people, and if they're not technical enough to understand everything (and don't have the time/energy/whatever to learn) then why not try to be at least a bit clearer to help them which in the long term might help ASR?

They appear to have enough time to park on other sites bashing this site. I don't have a technical background and one of the first things I looked at was DACs. Of course I saw the SINAD bucket list and wondering what it meant I read a few reviews and thought, Oh, it's a list of how well DACs meet basic engineering principles and bit of a price to performance guide if you dig a bit deeper . It took me all of a couple hours to figure this out.

 

[Multicore]

And when they ask what that threshold is, what would I say exactly?

"It depends."

7 years ago we got puppy Lucy, a shelter transport from Georgia. She developed behaviors we had to learn how to cope with. It's a big and badly organized body of knowledge. We consumed books, videos (commercial and free), and consulted anyone with experience. Eventually it became clear that there are very few general answers to questions in dog behavior. A million online dog experts are offering ready answers made me polemical about it: answers should usually start with "It depends."

How much amp noise is too much? It depends. driver sensitivity, listening distance, ambient noise, tinnitus, ...

Threads like this will, I hope, remain a common feature of ASR since they reflect the complexities and uncertainties people encounter as they try to use measurements in their specific applications.

 

[solderdude]

Here is a question I have had.
I asked Amir a few times, quite a while ago, and never got a reply ( I assumed he missed my post or what have you )

I asked Amir, if when measuring product "X" , that had fairly mediocre SINAD, if he had actually listened to the unit in operation, and had "heard" the actual output of the unit or based his assessment strictly on just a measured number?

There is little to no correlation between perceived sound quality and 'mediocre' SINAD (in the 90-100 range ?) so assesments are going to be made based on ALL measurements Amir did, not on SINAD alone.

I questioned him, only cause I did not see any subjective impressions.

Subjective impressions of electronics have no meaning at all.

guess I expected to see some mention of "As the SINAD measured mediocre, I was able to hear/not hear the limitations" in his review.

And yeah I get bad is bad, and great is great, but that middle ground, SINAD can be various combinations of distortion and noise, and not all noise, has the exact same audibility, and often when mixed in with an actual signal of varying frequency, it is masked in different ways.

SINAD and perceived sound quality have no relation. It merely says something about noise and distortion at 1kHz only at a specified level.

why not try to be at least a bit clearer to help them which in the long term might help ASR?

you mean like this for instance ?

Starting with balanced output, we immediately see signs of competent implementation

or this:

Averaging the two the S3 Pro lands in competent design category for all DACs tested:

in: https://www.audiosciencereview.com/...hidizs-s3-pro-review-headphone-adapter.25876/

 

[Vacceo]

I love heavy distorotion and saturation. I love low-fi music, super crude recording, heavy effects and all sorts of "unnatural" sounds.

But I like them on a very clean plate and fork. Keep the artifacts to the artists, let them use as much artificial sound as they possible can. But reproduction should not add or detract anything.

This forum does not question tastes, what questions is the cleanliness of the silverware.

 

[Wesseling Audio Lab]

JustJones comments are on the mark!

and much of the conversation here will complicate it more for some, and they might be saying "see, the regulars at ASR don't even know the purpose, meaning or importance of SINAD score"

 

[Multicore]

I don't see any need for a disclaimer.

Agree. I think that choice of the word "disclaimer" is peculiar and telling. It's legalistic and suggests that ASR is at risk of being responsible for misleading.

There is ample evidence that ASR does not promote simplistic ideas about SINAD as a metric of SQ. When people straw man that it does then the right response isn't disclaimers, in my opinion. Either robustly denounce the BS for what it is or ignore it. Adding disclaimers concedes the point to the BSers.

 

[solderdude]

disclaimers and explanations on what a measurement can tell and can not tell are different things.

When the idea is that ASR is for the masses too and and not just for measurement geeks that understand it then it makes sense to have some links to explanations of measurements in every post.
Rtings does this for instance.

 

[Multicore]

They appear to have enough time to park on other sites bashing this site. I don't have a technical background and one of the first things I looked at was DACs. Of course I saw the SINAD bucket list and wondering what it meant I read a few reviews and thought, Oh, it's a list of how well DACs meet basic engineering principles and bit of a price to performance guide if you dig a bit deeper . It took me all of a couple hours to figure this out.

Bingo! That's an example of what I said above (#226) about SINAD being a hook into learning.

 

[danadam]

Agree. I think that choice of the word "disclaimer" is peculiar and telling.

I often find it is mistaken with disclosure. Maybe that one would be better?

 

[Tell]

They appear to have enough time to park on other sites bashing this site. I don't have a technical background and one of the first things I looked at was DACs. Of course I saw the SINAD bucket list and wondering what it meant I read a few reviews and thought, Oh, it's a list of how well DACs meet basic engineering principles and bit of a price to performance guide if you dig a bit deeper . It took me all of a couple hours to figure this out.

Yes that's you, but not everyone is you..

you mean like this for instance ?
in: https://www.audiosciencereview.com/...hidizs-s3-pro-review-headphone-adapter.25876/

Yeah there's one example where it's written somewhere in the text.
But lets instead on the top of the SINAD ranking have it say "DAC Distortion and Noise (SINAD) Higher is Better (engineering wise, for human ears 90dBish is more than enough).
Tbh I don't see why this is even a discussion, people _clearly_ are misunderstanding this so why would it hurt to at least try and be a bit clearer?

 

[Chagall]

Subjective impressions of electronics have no meaning at all.

Do you think this will need to change if electronics keep adding "sound-enhancing" features? In other words, is the review complete if it doesn't give an impression (if not a test) of a key selling point?

 

[solderdude]

If they are intentional features then all that needs to be done is show what the feature does (phase, time delay or whatever it does, even forms of EQ)
The problem is that this could require a LOT of extra measurements and reporting.
Amir may well be of the opinion that only core measurements are sufficient.

Describing intentional effects like stereo expansion or crossfeed, volume dependent FR corrections or whatever trick one can think of can (and probably will) not be interpreted in the same way for everyone as 'subjective impressions' are just that.

 

[Old_School_Brad]

If they are intentional features then all that needs to be done is show what the feature does (phase, time delay or whatever it does, even forms of EQ)
The problem is that this could require a LOT of extra measurements and reporting.
Amir may well be of the opinion that only core measurements are sufficient.

Describing intentional effects like stereo expansion or crossfeed, volume dependent FR corrections or whatever trick one can think of can (and probably will) not be interpreted in the same way for everyone as 'subjective impressions' are just that.

If Amir were to test every feature of every device, we likely wouldn’t see many more reviews.
Just consider AVRs with their DSP settings, various inputs, and so on. He’d probably nod off and end up spilling coffee into the AP.