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

[boxerfan88]

Play a 0dB test tone (NOT AT FULL LEVEL-BEWARE!) and attenuate it at -80dB (better with an analog way so noise will go down too if there's any)
Tell us what you hear,either noise or distortion.

Great idea!

Unfortunately, my setup does not have analog volume control.
What I did is to generate -12dBFS signal, and set foobar volume at -65dB, my DAC volume is set at -3dB (insurance), total -80dB.

Results (-80dB):

• 1kHz - the tone is very faint
• Sweep - I could hear faintly only part of the sweep

I re-tested at -75dB foobar volume, total -90dB.

Results (-90dB):

• 1kHz - cannot hear the tone, room noise masking
• Sweep - I could hear very very faintly only small part of the sweep, probably 2k-5k. I really really had to concentrate to hear it.

I think your test is a very good idea! Folks, you should try!


(Please HEED THE WARNING ABOUT VOLUME LEVEL.)


.

 

[Sokel]

Great idea!

Unfortunately, my setup does not have analog volume control.
What I did is to generate -12dBFS signal, and set foobar volume at -65dB, my DAC volume is set at -3dB (insurance), total -80dB.

View attachment 419595


Results (-80dB):

• 1kHz - the tone is very faint
• Sweep - I could hear faintly only part of the sweep

I re-tested at -75dB foobar volume, total -90dB.

Results (-90dB):

• 1kHz - cannot hear the tone, room noise masking
• Sweep - I could hear very very faintly only small part of the sweep, probably 2k-5k. I really really had to concentrate to hear it.

I think your test is a very good idea! Folks, you should try!


(Please HEED THE WARNING ABOUT VOLUME LEVEL.)


.

Good!
Now imagine trying to hear the above while listening to normal level music.
It's liberating,isn't it?

 

[Multicore]

I ask again: why do folks create videos over and over to complain about this?

Such videos are an expression of the creator's learning in process. At least this is how things often seem to me. I have read a number of non-fiction books by respected authors that left me with the same impression (and even some fiction). Technical blogs explaining computer topics are more often written by someone who has just learned the thing than by an old hand to whom it is old hat. Sometimes the creators/authors are open about this but often not. Our culture rewards authority more than innocence or ignorance.

In defense of SINAD​

I think ASR's use of SINAD does little harm and on net more good than harm. If people are to become expert enough to understand all the measurements and how they relate to their individual applications and purposes for audio equipment then they have to learn. A lot. Their choice to learn is strictly their own. All we can do is help if they so choose. I think ASR is a splendid resource for those with the potential and inclination to learn.

And SINAD is one of our best lures to reel people towards learning, it seems to me. Threads like this one are almost always active on ASR. I think they reflects how ASR's use of SINAD is a force for good. They document the learning.

These evergreen arguments over SINAD are especially dear to me because they are an example of the uneasy relationship between objective data and problem solving, e.g. What headphones should I buy? that are deeply and unavoidably subjective. It's dear to me because I recall well how nearly 35 years ago I was disabused of the notion of engineering as ideally a scientific activity. Engineering, i.e. practical problem solving, e.g. Should I add a sub to my desktop audio?, involves questions too diverse and interrelated to develop objective answers. Being good at it requires more than expertise the objective domain.

objectively I do observe people being told that they should buy new equipment or pay more than they would otherwise in order to get something better-scoring (more often than not on sites other than ASR, but with reference to the rankings here), and that is indeed quite annoying to me.

I used to retail wine in an bottle shop. I was expert on the products and considered myself well experienced in how to use them in relation to diverse human purposes (to drink, decorate, seduce, educate, reminisce etc.). The Wine Spectator was indeed quite annoying to me. If someone departed with their WS90+ bottle having rejected my sophisticated council then the commercial purposes are met and the customer will likely be satisfied too. The harm was to my vanity.

When I got over myself I had to acknowledge that WS and its anti-intellectual scores is on net better than nothing. It's a gateway for those who will move on to learn more and it's not devoid of truth.

And so it is with SINAD. It is 1) a hook and lure towards learning with its inherent provocative questions: What am I, and what do I mean to you? and 2) not often badly misleading.

So when I read things like this

You are absolutely right, 1kHz 5W SINAD can be called as almost pointless number. And its use as a sound quality comparison number confuses crowds of laymen ASR readers.

I'm inclined to think: Good! Confusing the laity is on net probably better than not. Disorientation is the first step in learning:

From a condition of unconscious incompetence
1. become conscious of your incompetence
2. become competent and conscious of it
3. forget about that competence while directing your conscious attention to other issues.

Those at step 2 or on their way there are the ones making the videos, @amirm . They want to share what they learned.

 

[Sokel]

People should heed your warning about volume level to protect their hearing. Blown drivers can be replaced but ears? Not so much.

Have you tried the test with headphones or IEM? I guess it would be more audible then.

In that case I should probably add a parameter that was overlooked at the above.
Amp's gain.Cause if that -80dB is at line level we should add amp's gain to it,don't we?

And that's where the 60dB pro SINAD mark comes from.

So...

 

[Chagall]

I like seeing it the other way around.
Everything that can be used as a selling point cuts both ways.
In case of SINAD is funny,cause the ones using it as a selling point run out of other claims,like SQ so they use strange language ,like Topping for example who says "from extreme to more extreme"
That's implying improved SQ in my book which is not the case by a long shot at 120dB SINAD.
I know,market finds its ways and anything can be used.But no one can seriously claim audible differences (often never implying noise which is the obvious but other,undefined parameters)

The fumiest of all is that sometimes these devices use what they call "Sound color" which is added distortion in the 50's and 60's order and around forums the consensus is that no one can tell them apart switching from the 120dB SINAD mode.

Amir and others are right,education is the only way.

I agree with you.

Topping and other companies will have to adjust their key selling points, otherwise, it just comes off as silly marketing. SINAD chase reminds me of when gaming mice manufacturers were chasing DPI...10000, 20000, 30000. They still do, but their customers have spoken, and more important aspects like ergonomics, weight, and wireless are now in focus while DPI is secondary.

As customers become more educated, other selling points will be more emphasized and SINAD will be just a spec - and I think we are already there!

 

[boxerfan88]

People should heed your warning about volume level to protect their hearing. Blown drivers can be replaced but ears? Not so much.

Have you tried the test with headphones or IEM? I guess it would be more audible then.

Ha! Another great idea!!
I need to go hunt down my headphone and setup a DAC with headphone out.
(My current DAC has no headphone out). KIV for the next few days.

In that case I should probably add a parameter that was overlooked at the above.
Amp's gain.Cause if that -80dB is at line level we should add amp's gain to it,don't we?

And that's where the 60dB pro SINAD mark comes from.

So...

I don't have an external amplifier in my setup.
My speaker gain are set to 100dB @ 0dBu.
My DAC 0dBFS balanced output is rated at 4V.

 

[Old_School_Brad]

I agree with you.

Topping and other companies will have to adjust their key selling points, otherwise, it just comes off as silly marketing. SINAD chase reminds me of when gaming mice manufacturers were chasing DPI...10000, 20000, 30000. They still do, but their customers have spoken, and more important aspects like ergonomics, weight, and wireless are now in focus while DPI is secondary.

As customers become more educated, other selling points will be more emphasized and SINAD will be just a spec - and I think we are already there!

Hey, an analogy that works!

 

[Sokel]

People with very high efficiency speakers are more likely to hear noise that room noise will not cover mask. And using headphones that isolates or are sensitive are similar.

So the the SINAD depends on the use case as well as if it's noise or distortion dominated.

Another aspect not touched upon is the circumstances when the test is done as in a very quiet environment after acclimation the audible threshold is lowered.

There's lots of ways to be able to hear it,sure.Specially at the upper sensitive range.

But with normal listening level music on top of it?Even at the quiet passages of classical RTA shows about 60-70dB noise floor at the recording,and that's not live.
Let alone that some of them need this "air" around them to sound natural.

There will always be corner cases.But fighting FOMO is about the broader picture,cause this corner cases will probably know what to do with their gear.

Edit: To show it from a different perceptive with normal levels,just think what an RT60 measurement stands for.It clears the picture big time.

 

[Sokel]

As you brought up SINAD 60 in a reply above: I don't know if you listened to the three (claimed) SINAD 57 clips in the OPs video with the two first there was added distortion and the third had added noise. The increased noise floor was very audible, especially using headphones.

Yes,noise is one of the most telling clues at ABX and general.
That's why SNR and a good system-wide gain structure is important.There are great articles about it here.

 

[beagleman]

Such videos are an expression of the creator's learning in process. At least this is how things often seem to me. I have read a number of non-fiction books by respected authors that left me with the same impression (and even some fiction). Technical blogs explaining computer topics are more often written by someone who has just learned the thing than by an old hand to whom it is old hat. Sometimes the creators/authors are open about this but often not. Our culture rewards authority more than innocence or ignorance.

In defense of SINAD​

I think ASR's use of SINAD does little harm and on net more good than harm. If people are to become expert enough to understand all the measurements and how they relate to their individual applications and purposes for audio equipment then they have to learn. A lot. Their choice to learn is strictly their own. All we can do is help if they so choose. I think ASR is a splendid resource for those with the potential and inclination to learn.

And SINAD is one of our best lures to reel people towards learning, it seems to me. Threads like this one are almost always active on ASR. I think they reflects how ASR's use of SINAD is a force for good. They document the learning.

These evergreen arguments over SINAD are especially dear to me because they are an example of the uneasy relationship between objective data and problem solving, e.g. What headphones should I buy? that are deeply and unavoidably subjective. It's dear to me because I recall well how nearly 35 years ago I was disabused of the notion of engineering as ideally a scientific activity. Engineering, i.e. practical problem solving, e.g. Should I add a sub to my desktop audio?, involves questions too diverse and interrelated to develop objective answers. Being good at it requires more than expertise the objective domain.


I used to retail wine in an bottle shop. I was expert on the products and considered myself well experienced in how to use them in relation to diverse human purposes (to drink, decorate, seduce, educate, reminisce etc.). The Wine Spectator was indeed quite annoying to me. If someone departed with their WS90+ bottle having rejected my sophisticated council then the commercial purposes are met and the customer will likely be satisfied too. The harm was to my vanity.

When I got over myself I had to acknowledge that WS and its anti-intellectual scores is on net better than nothing. It's a gateway for those who will move on to learn more and it's not devoid of truth.

And so it is with SINAD. It is 1) a hook and lure towards learning with its inherent provocative questions: What am I, and what do I mean to you? and 2) not often badly misleading.

So when I read things like this

I'm inclined to think: Good! Confusing the laity is on net probably better than not. Disorientation is the first step in learning:

From a condition of unconscious incompetence
1. become conscious of your incompetence
2. become competent and conscious of it
3. forget about that competence while directing your conscious attention to other issues.

Those at step 2 or on their way there are the ones making the videos, @amirm . They want to share what they learned.

Hmmmmmm.....having just watched the video, it sounds like you are saying the creator of the video, is unaware he has a lot to learn basically??

If anything I see him, as coming off far from clueless, but simply going into more depth about SINAD, and giving some plausible reasons why he feels one simple SINAD figure alone does not quantify, how the end sound signature will be as far as various distortions and various frequencies of noise.

Leaving me a bit confused. By you saying he needs to learn, you are implying he is wrong in most/many/some of his comments, or simply you do not want SINAD and the Harman Curve to be questioned, and not rock the boat??

 

[Multicore]

Hmmmmmm.....having just watched the video, it sounds like you are saying the creator of the video, is unaware he has a lot to learn basically??

No, I'm not saying that at all. I wrote that very often those who make youtubes like this are at the "2. become competent and conscious of it" stage. I didn't watch much of the video in question so idk about that one.

At level 3. "unconscious competence" you need some reason to turn your attention to the question.

At stage 2. you are conscious of your competence.

 

[Mauro]

Why is that a bother to people?

I don’t understand all the fuss anyway a line for threshold of audibility with a research paper reference would be a nice addition.

 

[shuppatsu]

Why is that a bother to people?

Speaking only for myself, it's not. But to me it looks like you're trying to have it both ways. It's a ranking and you go out of your way to emphasize it as such. This is not a bad thing, but people can and do draw the wrong conclusions from it. It's fair to critique it as such, even if on balance it's a good thing.

In general, I find many people on ASR fetishize measurements in ways that are not helpful to the average consumer. I noticed this most recently in the Wiim Amp Pro comments, where many people dismissed the amp as a "toy" for supposedly poor IMD with no reference to audibility.

 

[ahofer]

I think where I find "Issue", is many posters I have seen, Value SINAD over everything, which is their prerogative. But to me it hints at an issue, when said posters, put out blunt statements about "I would never even CONSIDER that product based on SINAD alone it will sound unlistenable".........but NEVER reference that they are simply reading a number that was measured, and have never even HEARD the item in question.

Their comments make those less in the know, believe, that one can ascertain HOW IT SOUNDS, without having to hear it at all, and that there is some line of audibility that is not really an audibility line, but a point some simply will not go below, based on their interpretation of what they "Want to believe" is acceptable numbers or not acceptable numbers.

My point, is the audibility matters, when one starts making "Sound quality" comments, mixed WITH numbers but has not even bothered to give it a spin.

In other words, one should not say "It sounds horrible or unacceptable", oh yeah but I never really heard it, I just saw a number..........Kinda wanting it both ways.

I think this is mostly a strawman. Can you point me to a few regular posters sincerely making this argument?

 

[ahofer]

I don’t understand all the fuzz anyway a line for threshold of audibility with a research paper reference would be a nice addition.

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

 

[Aerith Gainsborough]

Interesting data point.

According to the chart ... >80 to 89/90 is red (poor/trash). Fair begins at 90.

Sure and rightly so, because it indicates crappy design. It doesn't take much more effort to manage at least -90dB.

My old Yamaha AVR had something around -80dB and the only problem I've ever ran into was an audible noise floor when trying to pump up very small signals like from the Garritan piano VST.

Increasing digital volume at the source or via my RME as a pre-amp completely eliminated the issue.

Naturally, if your use-case demands the heavy amplification of puny signals in a very quiet, specialized room, -80dB will not be satisfactory. But if we're talking run of the mill domestic setups with the usual line levels or typically loud mastered digital stuff, I doubt it will be noticed.

Like with any measurement, you have to evaluate it in context of your actual real life use-case.

Play a 0dB test tone (NOT AT FULL LEVEL-BEWARE!)

 

[krabapple]

I didn't watch more than a minute of the video but I'm guessing the purpose of making the video was to create hubbub.

Of course. As such I neither 'like' nor 'subscribe'. ;>

 

[krabapple]

In short: Not all required measurements to show performance are always done, not all measurements are always done correctly, not all measurements are always interpreted correctly.

Which would make a far more credible video than the FUD that the lad promotes.

 

[antcollinet]

Obviously I must have seen this numerous times, to bother bringing it up.

Or you have just decieded that must be the case to fit your own biases. Or you've seen lots of other people claim what you claim (with a similar lack of evidence)


I also have never seen any of the regulars state what you are saying is a common trope. It is of course a common straw man.

 

[antcollinet]

Sorry, not willing to join in on your personal crusade. You have made several semi rude comments, now, please move on.

Don't post misinformation, and I wont feel the need to comment on it.