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

[boxerfan88]

Maybe we need to use 50 shades of grey?

That’ll be nice.

 

[Randyman...]

Are we heading in the Pronoun direction now? Infinite pronouns!

 

[solderdude]

There is only one certainty: if SINAD is better that 115 dB, it guarantees transparency.

If that is really the case all one needs is 2 tiers .. for instance green and red... or 'transparent' and 'may not be transparent'
Problem then is ... people want to know how green or red and want to see a list.
conclusion nothing will be solved even with 2 color bands.... some DAC with 110 may still be transparent but would be red.
The horror. Sliding color scale is basically the same problem as at some point the scale is going to change color.

Darn it ... to turn the ranking of gear into something more useful as a buying guide (what some people think or make it out to be) one would have to invent a new metric (which already has been attempted, PKmetric for instance.
That would have to take frequency response, distortion profile, various load conditions (which would be appropriate ?), noise and IM into consideration including psycho-acoustics. That could make an educated guess and maybe even create various rankings for certain typical applications.

Nah... this can't be solved.

Educating those watching the numbers/plots is the only thing that might help (and can already be found).
You can't force education on people. They would be willing to educate themselves. Therein lies another problem... one would prefer to read that what confirms their line of thoughts. I mean .... there will be plenty of people that hate ASR and they will never believe what is stated and rather build a straw man and use that to discredit those that believe otherwise.

I say ... continue on the chosen path as it is impossible to satisfy all people...

 

[Sokel]

I think we complicate things more than we should.
There's two questions for any individual.
-can you hear noise at your system?
-can you hear distortion at your system?

And I deliberatively use "system" and not "component" cause that's where the juice is.The best of the best linked together does not guarantee an 100% optimal result.
For example we bush AVR's cause of their 90-100dB SINAD,right?

Ok,built me a system consisting of a pre-x-over-various DSP addilities,codecs,a gazillion of connections -power amps and measure it end-to-end and tell me if you can reach the one-box AVR.Let alone the various sub-optimal conditions one faces at home.
I don't want to be misunderstood,I don't even use AVR's.But measuring and watching how a gain structure evolves even at a two-VC system one gets a bitter smile.

Yes,education is what's needed.We see people buying from the top of the rank and say "you know,I listen to a 115dB SINAD system" .
It would be funny if it wasn't sad,as no one counts level and gain structure to it.

So what's left?
You hear noise at YOUR system>no,good.
Your hear distortion>no,good.

It's far more useful to promote articles like the gain structure ones and the rest here and urge people to start measuring (anything will do for basic ones) than anything else.

 

[thewas]

You guys must think a bunch of idiots read my reviews.

Once more, the categorization, as shown in colors, is to get people to think about four broad categories of ratings here, rather than individual values.

There is a contradiction again there, if the readers are experienced they don't need such arbitrary ranking classes and if not they can be influenced by such in a wrong way as it is been seen so many times in discussions. One approach would be to make just a summarizing table with the main measured values from SINAD over IMD to channel imbalance and next to them some approximate limits under which a deviation might be audible and if for some not feasible a min/max range from previously tested devices but without a qualitative comment. As its now it seems that some inexperienced readers seem to concentrate on high SINAD devices which often have other issues which might be more problematic and audible then a lower SINAD device which doesn't have such.

 

[solderdude]

Tiers are for laymen (that don't understand and misinterpret it anyway) and the numbers are for people that understand what the number actually means ?

You can't please all people and not all people understand.
There isn't a single number or ranking that ranks something in the way people want something ranked.

 

[Old_School_Brad]

Then quit saying people are confused when one device has a SINAD of 99 and another 100 because they are two different colors. You are solving nothing with your suggestions and constant complaining across topics.

Constant complaining? That's a mischaracterization.
I recall having a civil disagreement about rating and testing amplifier power, along with this single comment of mine regarding SINAD. In it, I shared a proposal on how it could be improved and explained why I believe people sometimes misinterpret it.

Your belief is in conflict with your own argument as I explained to @thewas. Once more, the categorization, as shown in colors, is to get people to think about four broad categories of ratings here, rather than individual values. It was a solution that was created based on consensus from me and membership. You have made no logical or rational argument to change it, on top of it being totally inconsistent with your own position.

I don’t believe there are any conflicts or inconsistencies in my comment, so it seems that either you’re not interpreting it correctly, or I’m not expressing it clearly.
There seems to be some defensiveness when polite suggestions or thoughtss about your work are shared. This obviously make it harder to have constructive discussions on some topics.

I considered letting it go, but then I thought perhaps you didn’t fully understand what I was trying to convey.

So, let me try explaining it differently.
Grading products by tiers influences how people perceive them. Adding colors to these tiers amplifies this effect, especially when the colors evoke emotional responses or are associated with good (green) and bad (red). This kind of framing and introducing of interpretation bias doesn’t align well with the principles of science, where data should be presented objectively.

With this in mind, I suggested that the pushback on the SINAD value might stem from how powerful the chart is due to these interpretive mechanics.

I then proposed that the effect and potential misinterpretation could be reduced by making the chart more neutral -perhaps removing the colors and showing only the values. Additionally, I recalled the use of "strict" and "lenient" terms to describe two thresholds, where "strict" represents absolutely inaudible and "lenient" represents inaudible in most real-world scenarios. (I haven’t verified if these descriptors are completely accurate, but that’s how I remember it. Feel free to correct me if I’m mistaken.)

These descriptions could be incorporated into the chart, perhaps as dividers or dashed lines, adding clarity and usability while grounding it in science. This approach could also help educate users in a meaningful way.

 

[JSmith]

Adding colors

SINAD table colour brackets

As a slight introduction: I'm a new member. I've been following along every review since the Google Pixel adapters which, looking back, was a only on a ti-ny SINAD list! Thanks @amir for all your hard work. Before ASR, I had little to no proper foundation on what DAC to buy other then pricing...

www.audiosciencereview.com

JSmith

 

[Old_School_Brad]

SINAD table colour brackets

As a slight introduction: I'm a new member. I've been following along every review since the Google Pixel adapters which, looking back, was a only on a ti-ny SINAD list! Thanks @amir for all your hard work. Before ASR, I had little to no proper foundation on what DAC to buy other then pricing...

www.audiosciencereview.com

JSmith

Thanks, I understand the reasoning behind adding them. I’m simply suggesting that they might also contribute to the misinterpretations we’ve observed, for the reasons I outlined in my previous comment.

 

[JSmith]

Thanks, I understand the reasoning behind adding them. I’m simply suggesting that they might also contribute to the misinterpretations we’ve observed, for the reasons I outlined in my previous comment.

All good mate, was just pointing out the thread from 6 years ago where it was also discussed. If you look closely you can even see a post from @amirm where he expressed a possibility of having to defend the categories and a desire not to... but in the end everyone decided what was proposed was good.

If the membership wished to suggest something new and discuss same, that might be the thread to necro? This one was more about that Cameron fella's video.


JSmith

 

[boxerfan88]

SINAD table colour brackets

As a slight introduction: I'm a new member. I've been following along every review since the Google Pixel adapters which, looking back, was a only on a ti-ny SINAD list! Thanks @amir for all your hard work. Before ASR, I had little to no proper foundation on what DAC to buy other then pricing...

www.audiosciencereview.com

JSmith

I went to take a look at the old thread, glad to note there is no rounding error in the original proposal.

The current chart has bugs. Example the SINAD 90 devices straddling both POOR & FAIR categories.

 

[JSmith]

Example the SINAD 90 devices straddling both POOR & FAIR categories.

So this one was 89.4 & 89.8, midpoint between the two is 89.6.

xDuoo XA-10 Review (DAC and Headphone Amp)

This is a review and detailed measurements of the xDuoo XA-10 combination DAC, Headphone Amplifier and Bluetooth receiver. It was kindly sent to me by one of their sellers, Shenzhenaudio audio. It costs US $479 from their store. The XA-10 breaks the mold for desktop DACs with one attractive...

www.audiosciencereview.com

This one was 88.9 & 90.3, so midpoint 89.6.

Review and Measurements of Sony STR-ZA1100ES AVR

This is a review and detailed measurements of the Sony STR-ZA1100ES Audio/Video Receiver. It is on kind loan from a local member. It costs US $999. It must have price protection as Amazon price is only $1 cheaper. Being part of the ES series from Sony, one expects good look and feel and...

www.audiosciencereview.com

This was was 89.6 & 90.2, so midpoint 89.9.

Hifi Forum TDA-1541A DAC Review

This is a review and detailed measurements of the "Finished Philips hifi forum tenth anniversary TDA1541 DAC Coaxial input L17-53." It was purchased on ebay by a member for $388. Owner swapped out the fake TDA-1541A for a real one. This was a DAC design created for the tenth anniversary of...

www.audiosciencereview.com

89.6 & 89.9 round up to 90, but are technically still in the 89.* grouping when looking at the more exact data in each review.

The others in orange that start at SINAD 90 had figures that started in the 90's, thus are in the other category.


JSmith

 

[boxerfan88]

89.6 & 89.9 round up to 90, but are technically still in the 89.* grouping when looking at the more exact data in each review.

The others in orange that start at SINAD 90 had figures that started in the 90's, thus are in the other category.

JSmith

Yup yup. The chart with integer SINAD is misleading, isn't it?
Since I reported the bug, nothing done to fix it.
The fix is so simple - round up/down (or trunc) SINAD to full integer - then categorize/bracket/group to the color codes.
(or change to color shades as suggested by others).

 

[Geert]

So what's left?
You hear noise at YOUR system>no,good.
Your hear distortion>no,good.

You missed an option, actually the most common one:

You hear no noise at your system, but you buy an upgrade and you hear the noise floor is lowered.

 

[Sokel]

You missed an option, actually the most common one:

You hear no noise at your system, but you buy an upgrade and you hear the noise floor is lowered.

Blacker black!

 

[boxerfan88]

Blacker black!

That and "darker background".

I have never figured out what "dark background" really means... relating to equipment changes.

 

[Tell]

Grading products by tiers influences how people perceive them. Adding colors to these tiers amplifies this effect, especially when the colors evoke emotional responses or are associated with good (green) and bad (red). This kind of framing and introducing of interpretation bias doesn’t align well with the principles of science, where data should be presented objectively.

With this in mind, I suggested that the pushback on the SINAD value might stem from how powerful the chart is due to these interpretive mechanics.

I then proposed that the effect and potential misinterpretation could be reduced by making the chart more neutral -perhaps removing the colors and showing only the values. Additionally, I recalled the use of "strict" and "lenient" terms to describe two thresholds, where "strict" represents absolutely inaudible and "lenient" represents inaudible in most real-world scenarios. (I haven’t verified if these descriptors are completely accurate, but that’s how I remember it. Feel free to correct me if I’m mistaken.)

These descriptions could be incorporated into the chart, perhaps as dividers or dashed lines, adding clarity and usability while grounding it in science. This approach could also help educate users in a meaningful way.

Yes, this is exactly what I've been saying as well! Clarification for all those people that don't fully understand all the measurements while not affecting those that do understand it, this won't hurt in any way. Because again, people quite obviously misunderstand that SINAD ranking list so why not help them a bit? This in the end will also help ASR as a whole so it's really a win-win situation

 

[JimBean]

That and "darker background".

I have never figured out what "dark background" really means... relating to equipment changes.

None more black. I don't know, I've never really had an issue with audible noise in my system, so not sure I would notice a lower noise floor. I also don't listen to vinyl, so maybe that's where all this talk of lowered noise floor and blacker backgrounds stems from.

 

[Chagall]

Yes, this is exactly what I've been saying as well! Clarification for all those people that don't fully understand all the measurements while not affecting those that do understand it, this won't hurt in any way. Because again, people quite obviously misunderstand that SINAD ranking list so why not help them a bit? This in the end will also help ASR as a whole so it's really a win-win situation

This has all been said before; there is no need to beat a dead horse. @Amir is aware - it's up to him, but more importantly, it is up to us (speaking for myself also) to educate ourselves.

 

[Tell]

This has all been said before; there is no need to beat a dead horse. @Amir is aware - it's up to him, but more importantly, it is up to us (speaking for myself also) to educate ourselves.

Yeah it's just that he seems so very reluctant which I don't understand why. But yeah, it's been said enough at least from me by now.