• 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!

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

Is that science? I'm not supporting any side here, I have no dog in this race. I admire the S in ASR.
We are in applied science here. Everything has to have results and not just for the sake of science!
 
That is orthogonal to mission and purpose of my reviews. I only test for performance, not reliability, design unrelated to performance, etc. Those factors by definition cannot enter its performance rating. A Porsche is very fast and handles great. That it is also unreliable doesn't stop auto reviewers from giving high marks for its great "engineering."
I haven’t been a car nerd for about five years, but back when I was paying attention Porsche had a pretty good reputation for reliability—and a great one if you just consider sports cars.

But for me I always looked at the lack of reliability as not orthogonal to the question of engineering performance as such, simply beyond the abilities of the car magazines to measure. Only Consumer Reports had the size and resources (by polling from a much larger base than car nerds) to get remotely accurate reliability data.

For me, ignoring reliability is missing a huge and critical piece of the engineering puzzle. I’ve owned Audis and Volkswagens, and the replacement parts I’ve constantly had to purchase for them have been the most beautiful and substantial Bosch parts you could imagine. Meanwhile the Japanese cars I’ve had look like they’ve cut every corner they could, and yet they keep on going, and going, and going. And when I do have to replace a part, it’s humble, homely, cheap, and i never have to replace it again. There’s real engineering that goes into that.
 
Last edited:
Problem with SINAD is the distortion part.That's what people question all the time.
Noise is obvious,it's either there or it isn't.

And it gets worst as these new chips have the ability to alter THD at your "liking" so some use it to create what they call "Sound color" or similar nonsense.
Top of the chart DACs too.
These usually have a SINAD of 40-50-60dB cause of these "profiles"

Problem is that people keep testing them and no one can tell them apart from the usual 120dB SINAD.
And that invalidates the metric in their eyes (ears better),I'll say it again,the vast majority of the small percentage around traditional forums who try to put SINAD in their lives link it directly with SQ.
 
For me, ignoring reliability is missing a huge and critical piece of the engineering puzzle.
Nobody is asking you to ignore reliability. My point with the post was that no one is saying Porsche has lousy engineering because their cars break down. Yes, reliability is important especially when you pay so much for these cars. But as I said, it is orthogonal to the mission and charter of this site.
 
Problem with SINAD is the distortion part.That's what people question all the time.
You guys keep talking about this "problem." As soon as I review anything, half a dozen people say performance is more than good enough. That anything over 100 dB SINAD doesn't matter, etc. All of this gets said in first page or two of every review. So there is no problem here but people wanting to make some argument disguised as "what is wrong with SINAD."

SINAD is what it is. First order and important indication of audio fidelity and transparency. It is a gift in the way it can do that. Let's move on for heaven's sake and focus on solving the million things that are wrong with marketing and reviews of products outside.
 
And that invalidates the metric in their eyes (ears better),I'll say it again,the vast majority of the small percentage around traditional forums who try to put SINAD in their lives link it directly with SQ.
I have never seen this. But so what if it happens? World of audio is broken at the core in so many levels that you all want to worry about this?
 
Let's move on for heaven's sake and focus on solving the million things that are wrong with marketing and reviews of products outside.
Ok,let's see it this way.
Is it OK to stuff a DAC with distortion and call it "sound color"?Isn't that some of the worst marketing around?
Plus it makes distortion seem like something nice,not a problem.
Topping and SMSL have it at their flagships.
 
Is it OK to stuff a DAC with distortion and call it "sound color"?Isn't that some of the worst marketing around?
Who cares? It is an obscure option in a menu that someone here and there plays with.
 
Who cares? It is an obscure option in a menu that someone here and there plays with.
Judging by myself (I sure don't) a few.But vocal few who make a case out of it.
I care 1000% more for something that looks nice or feels reliable than anything else.
So...
 
That is orthogonal to mission and purpose of my reviews. I only test for performance, not reliability, design unrelated to performance, etc. Those factors by definition cannot enter its performance rating. A Porsche is very fast and handles great. That it is also unreliable doesn't stop auto reviewers from giving high marks for its great "engineering."
Yep, all ASR can do is check technical performance and maybe some engineering aspects.

In case of the Porsche the design is good, even the engineering is good but the build quality of parts is lacking or even the engineering of the failing parts was lacking.
The same can be the case for audio gear. Good design (low noise, low distortion, good FR response (bandwidth), does not fall short on load drive capabilities, designing proper protection circuits.
It can even be well engineered but the quality of components may fail or overall engineering is fine but some components over-heat or vibrate (inductors/transformers) and cause failures. This is engineering and/or component choice/quality.

So when we talk about definitions we can say that a high SINAD is a sign of proper design and could (per there is no 100% correlation) to engineering. That may be crappy.
The potting of the PA5 amps, the lack of potmeter grounding and PCB grounding definitely falls under engineering.
Then there is something like insufficient cooling, EMC sensitivity etc.
This too is engineering.
It all has to do with longevity, which arguably you cannot test for without adding other (EMC) tests and accelerated aging tests and operation in climate chambers.
Clearly outside of the scope of testing for signal fidelity and certain basic load testing.

One could also build (engineer) a superb device that might function according to design specs for over 100 years and never fail and safe to use, yet be designed to have a certain distortion profile that affects the D in SINAD. That design follows the design criteria, is sells well and gets positive reviews from owners (and of course press as they rarely are disappointed) and have excellent build quality using top quality components and has an extremely low failure rate.
That would have excellent engineering, have the designed technical properties and is independently verified. It could even be priced really friendly (unlikely though)>
That device would have great engineering (but together really well), perform exactly (or even better) than specified but ... has poor SINAD (because of the D).
Such a device could well be highly recommended for people that do not care about the D, despite having poor signal fidelity.
So... SINAD is a good indicator of signal fidelity and circuit design but not the engineering.

I agree that this isn't the scope of ASR testing which is checking technical performance (i.e. signal fidelity).
Maybe during testing you noticed the temperature becomes too high in which case you mention this and post FLIR pics.
Still SINAD could be very high, even at the top of the chart but still engineering (build longevity) is not great despite the build of the internal PCB and casing, switches, controls itself being top notch, crappy or anything in between.
Something that falls outside of the scope (and mission) of ASR.

So IMO SINAD is a good indicator of circuit design and signal fidelity. Just not of engineering. Chances are it is well engineered but may not be is my point.
Signal fidelity is more than SINAD though. One can have great SINAD (at 1kHz) but still have not so great performance outside of the (very easy for audio gear) 1kHz sine wave.
Multi-tone could be not that great (we've seen those devices, thanks to you not just measuring SINAD and calling it a day :) ) or other aspects could not be great (under load conditions for instance) .
This could lead to You not recommending the tested device but it still can have high SINAD.

So my definition of SINAD is that it certainly is a good indicator for proper electronic design and signal fidelity but not particularly on engineering.
Engineering (at least for me) is saying something about how well a design is turned into a practical device including heat management and component quality.
SINAD cannot possibly be an indicator for that (build quailty).

It's about defining what SINAD is and isn't a good indicator for. Definitions should be well defined (proper word choice).

Is SINAD important ? sure... it is a good indicator of signal fidelity and circuit design.
In the end it is just one number that is part of an entire measurement suite that indicates noise and distortion at 1kHz at a certain level.

The whole issue with stating that SINAD is important is that people that hate ASR and what it stands for and its mission use that as a reason to whack the dog with.
Simply because the reason for the ranking chart isn't always clarified.
All that could take is a single sentence embedded in the chart explaining what SINAD can show (not even what it can't) as ASR is the only site that ranks it.
Problem solved and haters can't use the chart anymore for that.... unfortunately they will use other things instead...
 
Last edited:
Reluctant about what? Putting silly disclaimers in reviews? Where are those disclaimers in other people's reviews? I have for long put disclaimers in reviews, with net result being cluttering them to no good effect. I have done tutorials which talk about all aspects of the review and that is that. I am not writing a legal agreement to have 12 page fine print on them.
In what way is that silly? If anything I'd say it's silly not recommending stuff because they have "low" SINAD when in reality the noise and distortion is most likely not audible. And I don't see why it would be silly adding Strict and Lenient auditability threshold lines just to be a bit clearer, tbh it's just as silly/unsilly as ranking with those words and colours you have on there now. Because again, that SINAD ranking is _clearly_ a problem for lots of people, of course not for you, me and half of ASR, but for many people it is, so being a bit clearer with a "silly" disclaimer will help everyone.
And it's not about writing a 12 page fine printed legal agreement, don't know why you need to exaggerate to an extreme making some kind of point? But yeah sure, if you don't want to then don't, doesn't really matter for me personally, but don't keep on looking surprised when more people will come asking/complaining/raging about SINAD, because they surely will.
 
Honestly, I think a lot of people don't understand that SINAD is just THD+N. The same shit they have been seeing on spec sheets for literal decades now. The only difference is that its value at 1kHz is what is used to give a quick insight into whether or not the device has a glaring issue.

The buckets are just to make it easy to see how it performs relative to the other devices tested, and yes sometimes a device is just on the edge but that is just how quartiles work.

If the measurement was changed to the minimum SINAD value between 20Hz-20kHz? Would that make people feel better? I'd argue you'd probably want to stop at like 15kHz, because most people can't hear 20kHz anyway, but you get the point.
 
Problem with SINAD is the distortion part.That's what people question all the time.
Personally I would argue it the other way…
As the noise part seems way more obvious.

-60 dB of pure white noise may have the system hissing like a basket of vipers at a Pentecostal revival.
Whereas 50dB SINAD, mainly of some 2nd harmonic, is something that actually can sound very nice.
Noice to the point where it might be preferred by a bunch of people compared to 120dB of SINAD.
But almost no one prefers a hissing speaker/system.

A system with noise around -110 and THD around <= -70 is generally not bad in an amplifier.
Basically SNR is more important to me than SINAD.
Unless the SINAD number has the FFT next to it, which is often the case, it is hard to determine whether the noise or the distortion is the culprit.

And since most speakers have distortion that is way higher than -60 dB it gets to be a bit of stretch to believe that the distortion will be audible.
But the noise is super audible.
 
Personally I would argue it the other way…
As the noise part seems way more obvious.

-60 dB of pure white noise may have the system hissing like a basket of vipers at a Pentecostal revival.
Whereas 50dB SINAD, mainly of some 2nd harmonic, is something that actually can sound very nice.
Noice to the point where it might be preferred by a bunch of people compared to 120dB of SINAD.
But almost no one prefers a hissing speaker/system.

A system with noise around -110 and THD around <= -70 is generally not bad in an amplifier.
Basically SNR is more important to me than SINAD.
Unless the SINAD number has the FFT next to it, which is often the case, it is hard to determine whether the noise or the distortion is the culprit.

And since most speakers have distortion that is way higher than -60 dB it gets to be a bit of stretch to believe that the distortion will be audible.
But the noise is super audible.
We're saying the same thing,no one argues noise.SNR is a far clearer and useful metric.
Of course is there for all to see at the reviews.But we don't rank devices by it.

I have tested every distortion possible,at many rigs.It has to go high to be noticeable to ME,and I mean high.Even IMD has to go fairly high,depending it's structure of course.
All that both with speakers and headphones.As for if I can discern things I have posted ABX's here which I think I did well so I'm not deaf.

So noise,yes,that's undeniable.
 
As SINAD is only ranked by ASR it would be preferable to either have a link to an explanation/rationalization of using this metric or a clear definition under or in the chart clearly stating what the number shows (signal fidelity/quality for 1kHz at a specified level) so even noobs can understand what it says.

Still, even then people will think or say that ASR believes 'signal quality = sound quality' and misuse that in order to discredit Amir and ASR.
That can be clarified with a link to a page explaining SINAD with examples and what it can and cannot show and why the metric is not silly to use.
Part of the all important education.
 
As SINAD is only ranked by ASR it would be preferable to either have a link to an explanation/rationalization of using this metric or a clear definition under or in the chart clearly stating what the number shows (signal fidelity/quality for 1kHz at a specified level) so even noobs can understand what it says.

Still, even then people will think or say that ASR believes 'signal quality = sound quality' and misuse that in order to discredit Amir and ASR.
That can be clarified with a link to a page explaining SINAD with examples and what it can and cannot show and why the metric is not silly to use.
Part of the all important education.
A sum can be also very useful for every user to pin-point potential problems without having to interpret the charts.
Something at the RMAA fashion,like the one below:

RME ADI-2-4 Pro SE.jpg


It's even going down to preferences if one gets to choose a metric that suits its gear.

Edit:Sums can also be dynamic and comparable,a link to them at the review index for example,like spinorama does.
 
Last edited:
I have 4 excellent Topping Dac's throughout the home in different systems and all purchased at a relatively low cost

Without ASR and SINAD, I would still be wasting time on DIY Dacs or purchasing over priced commercial dacs with poor noise performance that were given great reviews by dishonest reviewers.
 
Last edited:
Problem is that people keep testing them and no one can tell them apart from the usual 120dB SINAD.
And that invalidates the metric in their eyes (ears better),I'll say it again,the vast majority of the small percentage around traditional forums who try to put SINAD in their lives link it directly with SQ.
Exactly. They invest in a DAC with an 80 dB SINAD, either don’t hear a difference or think they do because it’s more expensive, and then conclude that the SINAD chart is meaningless and ASR users are misguided.

This type of comment isn’t uncommon outside this here small corner of the internet, and I believe it’s something that could be easily addressed by small tweaks to the chart itself.
 
Exactly. They invest in a DAC with an 80 dB SINAD, either don’t hear a difference or think they do because it’s more expensive, and then conclude that the SINAD chart is meaningless and ASR users are misguided.
Nobody reading ASR does that because they can get a far better DAC at any price point and feature set. You continue to invent non problems.
 
Back
Top Bottom