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

[ta240]

Why do these things almost always seem to devolve into being some kind of a competition, or get assigned some nefarious intent?

I don't know that making videos that are geared towards getting views and thus making money is a 'nefarious' intent, however videos about testing vs listening debate seems like a good way to get those. It is similar to the response that videos about Bose used to get or threads about vinyl or tubes get on here. For better or worse it is human nature to want to belong to a side and you do that by building your side up, tearing the other side down or a combination of both. Angry people post a lot and on youtube they have to come back to the video to post and argue.

 

[solderdude]

Not being able to detect spacial enhancement is worrisome to me: that also means that if a device has a flaw which results in a spacial "diminution", we wouldn't detect that

The funny thing is that he DID show the differences between the DAC one with and one without spatial features on using measurements.
It was clearly visible in the shown measurements. Just not in frequency response nor in SINAD (as it does not affect S/N ratio nor distortion at 1kHz).
It is something that cannot be shown in SINAD nor FR nor distortion measurements and is 'stereo signal dependent' so one would need crosstalk or phase measurements (between channels) to show that. The point is all is measurable but one needs the correct measurements for that.

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.

IMO we can do better. Some members can be very dismissive and can't be bothered. I mean we are asking new members did they ABX...they have no clue what that even is. If a new member has a problem with a specific amp, we just say sell that and buy Purifi....what use is that to the member? And get a UMIK, where are your in-room measurements? It's overwhelming...People don't want to get involved and leave nasty comments against ASR.

totally agreed. On the other side it is no different than people on the other side making claims that they can hear it clearly without any proof and stating the differences are so clear a BT is not even needed. Besides blind tests don't need to be double blind, just correctly performed.

What would have to be done is write a single post with pointers HOW to AB devices like DACs, amps, cables and some pointers about speaker and headphone comparisons and in replies simply link to that page. Then kindly make a respectful reply pointing to 'possible pitfalls' in 'testing' with a link.

It should, however, be very obvious that the continuation of 'audiophile myths' is VERY important for cable and certain 'high-end' manufacturers, reviewers, sellers so it is important for that market to continue.

All 'we' (and ASR) can do is point this out and remember ... friendly replies go a long way and 'framing' does not help.

the question is why SINAD is used at all for ranking?

Because it is easy to 'rank' (just like Harman scores b.t.w.) and people love rankings ?

What it shows is technical performance (in other words signal fidelity). I agree there should/could be a disclaimer (or a link to a page with explanation about what SINAD can and cannot say).
It just needs to be written or copied from the many posts from Amir about SINAD, replies I and others have made about SINAD and create a page that can be/is linked to.

It is the readers that don't understand and need to be educated about measurements and this is possible. It is those same readers that make the comments or are uneducated enough to simply buy based on that list. ONE thing is for sure though... the gear in the upper echelon of the SINAD chart will be performing well in general (not guaranteed).

In fact I am of the firm opinion about this for each measurement. Of course I could even do that on my own website (I have such a page but do not link to it under each measurement). Rtings actually does that and did so quite soon after its launch.

On the few positive things of current AI language model algorithms is that they can quickly create some decent text summaries, exemplary for the discussed video:

Importance of Understanding Measurements​

The video begins by emphasizing the need for a balanced perspective on audio measurements, stating that while measurements hold value, they shouldn't be deemed absolute. It encourages viewers to appreciate the limitations of both objective measurements and subjective preferences.

Limitations of SINAD​

GoldenSound delves into SINAD (Signal to Noise and Distortion Ratio), explaining its characteristics and limitations. While it quantifies total harmonic distortion and noise, it cannot predict how devices will sound to listeners as different types of distortion can lead to varying perceptions of sound quality.

Misinterpretation of Frequency Response​

The disparity between a headphone's frequency response measurement and individual listening experiences is discussed. GoldenSound stresses that small deviations from the Harmon Target should not be considered negative, as listener preferences vary widely.

Variation in Headphone Performance​

The discussion includes variations in headphone performance due to differences in measurement rig setups and listener anatomy. GoldenSound points out that headphone measurements can vary significantly among different individuals, making single measurements less reliable.

Encouragement of Healthy Discussion​

The video concludes by encouraging the community to engage in constructive discussions about audio measurements. It stresses the importance of scientific rigor in tests and the interpretation of results, advocating for an open-minded approach to understanding audio quality.

Yep, all very true. Kuddos to Cameron for addressing these points. I whole hardheartedly agree with these statements.

This means the intention was good.
My objections are more about the execution and examples. This could have been done in a much better way, using better examples, using less tooting of his own horn (Wandla, his blind test (still no follow up with real world decent DACs). Sure he does not directly point his finger at ASR (and even is positive about 1 video from Amir) and he is clever enough NOT to do so but ... watching the entire video it is clear what it is about.

There should be more explanations on ASR (there already are but obviously not visible or easy linkable enough) with links to it under specific measurements.

Education is important.

 

[amirm]

Education is important.

It is. That is why I have produced video tutorials explaining all this which I usually include in my reviews. This is the one on amplifiers which explains SINAD in detail and even remarks on audibility of it:

So I don't get people who watch the video in OP and say it has information they did not know but have not watched the equiv. video from me explaining them in far more detail.

 

[Palladium]

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. And the posting style of basically "check this out" seems to also be to for the purpose of hubbub. And I have to say I really like typing hubbub.

Grifters in MY attention seeking economy?

 

[solderdude]

It is. That is why I have produced video tutorials explaining all this which I usually include in my reviews. This is the one on amplifiers which explains SINAD in detail and even remarks on audibility of it:

So I don't get people who watch the video in OP and say it has information they did not know but have not watched the equiv. video from me explaining them in far more detail.

When these tutorials tend to become a bit too technical chances are people (with short attention spans) quickly stop watching but at least they would have been given the opportunity to become more informed in an area they aren't really interested in.

Most people prefer to watch 'drama' instead.

 

[Tell]

It is. That is why I have produced video tutorials explaining all this which I usually include in my reviews. This is the one on amplifiers which explains SINAD in detail and even remarks on audibility of it:

So I don't get people who watch the video in OP and say it has information they did not know but have not watched the equiv. video from me explaining them in far more detail.

This is great video yes, but as said everyone doesn't watch those long things, maybe the same people that probably didn't watch Camerons video either. Also some people won't watch it because it's THAT measurement guy from ASR, while they do watch Camerons because it "debunks" THAT measurement guy from ASR. Yeah people's gonna be people, haters gonna hate etc etc.
So I think a small and short TLDR under that SINAD ranking saying something like "SINAD is not everything, other things matters as well and human audibility threshold is generally way lower than this" or something like that would help, at least a little

 

[boxerfan88]

The SINAD in my measurements are part of a full FFT of the distortion and noise of the product in the dashboard:

You can then easily compare the level of spikes, in this example -130 dB and lower, and SINAD to realize that it is noise, not distortion bound. I routinely make a comment like this in the review:

"Distortion is vanishingly small at -130 dB, with SINAD then being noise dominated. "

If X + Y = Z and I tell you that Y is much smaller than X, then you can approximate X = Y. Sometimes the two numbers are close in which case, you then look at the dynamic range measurements:

You see that the DNR is very close to SINAD. It is 2 dB better because those distortion products do add up to something.

If you don't know these things, then please watch the full tutorial I have given on this, which also follows my AES paper and talk on the same:

Got it. Thank you.

No DAC is rated precisely on SINAD either as I always put them in one of four buckets:

Now, if someone wants the best of the best SINAD value and the gear is still cheap enough for them to afford it, then that is cool for them to purchase. I personally use a DAC that is in the middle of the blue range.

I have not watched his video but if this is all about, all the information and knowledge is here.

"No DAC is rated precisely on SINAD either" ... huh??

Not quite following ... isn't the chart above showing each product/device having precisely just one SINAD figure?


Oh by the way, who decides some devices rated 90 SINAD is "poor", and some other devices rated 90 SINAD is "fair"?

Ditto at 99 SINAD ...

These are arbitrary banding yes?

What is considered the threshold of audibility? 90dB SINAD? 96dB SINAD? 100dB SINAD?

I also wonder ... if the SINAD test is run at 500Hz or 2kHz ... will the SINAD be materially different or pretty much the same?

 

[Purité Audio]

Personally I find Sinad an extremely useful metric, it allows me to discount products at a glance.
Has Amir measured a component that has excellent Sinad but measures poorly in other respects?
Keith

 

[Mart68]

A dealer once said to me that most of his customers want to be told what to buy.

This is why there is these ranking systems like Stereophile classes, What Hi-fi stars, and crazy multi-faceted systems where the device is ranked out of ten for things like 'bass extension' and 'tone and timbre.'

So people take the SINAD chart as the same thing when it isn't.

I watched first ten minutes, he takes a strawman and knocks it over. He's a good presenter, he's wasted doing this sort of crap.

 

[Tell]

Personally I find Sinad an extremely useful metric, it allows me to discount products at a glance.
Has Amir measured a component that has excellent Sinad but measures poorly in other respects?
Keith

So at what SINAD do you discount a product?

 

[Chagall]

Got it. Thank you.



"No DAC is rated precisely on SINAD either" ... huh??

Not quite following ... isn't the chart above showing each product/device having precisely just one SINAD figure?


Oh by the way, who decides some devices rated 90 SINAD is "poor", and some other devices rated 90 SINAD is "fair"?
View attachment 419299

Ditto at 99 SINAD ...

View attachment 419300


These are arbitrary banding yes?

What is considered the threshold of audibility? 90dB SINAD? 96dB SINAD? 100dB SINAD?

I also wonder ... if the SINAD test is run at 500Hz or 2kHz ... will the SINAD be materially different or pretty much the same?

This video explains all DAC measurements:

SINAD chart is just that a SINAD chart, it's not a DAC rating chart. But as explained in the video, it's a great indication of how DAC will perform.

 

[Wesseling Audio Lab]

I watched the first 10 minutes "SINAD doesn't account for preference" he states! He played 2 tracks, both having 57db SINAD but with different distortion products and people can hear the difference! "people still love tube amps" Then I watched a little further.. So much wrong with this Video!

I'm not sure if he even talks about SINAD figures below threshold of hearing

 

[NTK]

So at what SINAD do you discount a product?

For DAC, 99 dB. When Apple can do one small enough to hide inside a Lightning/USB-C connector and retail it for <USD$10, what is your excuse?

 

[Tell]

For DAC, 99 dB. When Apple can do one small enough to hide inside a Lightning/USB-C connector and retail it for <USD$10, what is your excuse?

Features is the excuse. So if there's a DAC with inputs/outputs and maybe DSP or whatever below 99dB you won't buy it? Why not?

 

[NTK]

Features is the excuse. So if there's a DAC with inputs/outputs and maybe DSP or whatever below 99dB you won't buy it? Why not?

As Apple has shown, it can be done with negligible cost. Why can't you do it? What is your excuse?
[Edit]
Alright. I will give some leeway to features that are compelling. But when the features are defeated/not used, can they do at least as well as Apple?

 

[Tell]

As Apple has shown, it can be done with negligible cost. Why can't you do it? What is your excuse?

Yeah sure, but it's not like it makes an audible difference so for us consumers it doesn't really matter now does it?

 

[solderdude]

Oh by the way, who decides some devices rated 90 SINAD is "poor", and some other devices rated 90 SINAD is "fair"?

These are arbitrary banding yes?

Round-off errors.

p.e.:
90.35 becomes 90, so does 89.66
90.83 becomes 100, so does 100.42

 

[Sokel]

Features is the excuse. So if there's a DAC with inputs/outputs and maybe DSP or whatever below 99dB you won't buy it? Why not?

DACs do 100dB 40 years now.It has to be deliberate to go lower than that.And I don't mean at silly 4-5V levels where no DAC will ever operate for normal listening,but even at 500mV to 1 perfectly usable V.

 

[boxerfan88]

SINAD chart is just that a SINAD chart, it's not a DAC rating chart. But as explained in the video, it's a great indication of how DAC will perform.

Well, if it is indeed what you assert, then why group into 4 bins, and label them poor, fair, very good, excellent???

If it is indeed “just a SINAD chart” why the need to give labels like “excellent” or “poor”.

I would expect such a pure SINAD chart that is just a SINAD chart to just sort ascending or descending with no labels.

User can go scan the chart and shortlist DACs of interest. Kind of like spinorama website where one can scan and choose any 2 speakers to compare the charts.

 

[Multicore]

In general, tilting at a strawman or an extremist is a really robust rhetorical method - I wasn't directly involved in this video, but I've explicitly used (and referenced) that fact in my own related presentations. When your audience is probably at least half-wrong on a topic¹, a way to avoid making them defensive is to say "well you good people, I know you're better than this, but there are folks out there who are just truly incorrect", and then you crucify the strawman and hope that people shift their incorrect impressions without ever having felt they were the target of an attack, because you're mixing in correcting their misimpressions with indictments of someone who they've never been and will never be.

I wouldn't call this a particularly charitable method of rhetoric, since it's basically assuming that the audience can't stand to be told they're incorrect about something, but I've also observed it to be very effective with large, mixed groups. I probably wouldn't use this style when talking to my peers, but...I mean, it is a youtube video, that's about as low of common denominator as one is likely to get.

1: My baseline assumption is that the youtube audience is going to be >90% wrong on pretty much any arbitrarily chosen topic

...

Humans are political creatures, and viewed through a lens of tribal politics, something is usually either "with us" or "against us"

This is more cynical than I feel comfortable with.

In my own thinking I try to adopt a stance of independence as opposed to tribalism. I try to apply critical thinking to ideas relative to my values and avoid identity with my tribe in opposition to the Lacanian big other.

What are the stats, really? Your >90% assumption might be right or a post hoc rationalization of your choice of rhetorical style. Idk. If you were to use a style of teaching suitable for an audience of independent critical thinkers, addresses them as peers who, like you or me, might be insulted by presumptions of being easily manipulated by flattery and tribal rhetoric, would you not be able to find that audience?

Ignorance or being wrong (incorrect beliefs) are the two possible preconditions to learning. Pandering to peoples' sensitivity and insecurity about being ignorant or wrong is infantilizing and should not need to be a precondition to teaching. We can be more confident than that. Lots of us enjoy learning and seek it. I don't think I need to conjurer an imaginary tribe (big other) of more stupid people to belittle in every lesson, as in each of the bullet points on the PP slide that triggered me.

If it's the case that the stats are as bad as you say then we have a problem. Is this the inevitable natural order? I don't think so. In my experience people learn independence and critical thinking skills. So I have a choice: do I want to contribute to this problem or to resolving it?

Otoh, if we're just in the business of maximizing metrics by making the most tasty morsels of media to fill the spaces between ads then encouraging tribalism could fit the bill. Worked for cable news, until it didn't.