[ADUHF]

I'm not really that up on the whole SINAD is good/bad thing. But since this seems to tie in well with the generally sciency thrust of this thread/forum, here is another recent video on the subject from Resolve. Feel free to give your opinions below.

 

[ADUHF]

This dates back a few years, but it summarizes some of Steve Temme's and Dr. Sean Olive's findings on the subject of distortion in headphones at the time.



I think this is the AES paper discussed above.

https://www.aes.org/e-lib/browse.cfm?elib=17441

 

[ADUHF]

I have to admit the excerpt you quoted is pretty bad, so you do have a point. I can only promise you (since you probably didn't get any further than that) that the rest of the post is probably no better.



I don't believe you.



Of course it's not about assuming that those looking at measurements are dumb-dumbs. It does seem to me, though, that the moment there's an x and y axis and some squiggly lines, they're taken quite literally.

One headphone manufacturer used to include individual measurements with every headphone they shipped, and people would compare their measurements on the forums. As I recall, these were production test measurements, and not comprehensive, averaged sets. So when one headphone's measurement showed more energy or less energy at, say, 5 kHz than another, there was often concern. People would return and exchange headphones to get one that didn't have a measured dip or peak that they didn't see on another's measurements of the same model on the forums.

It was just as likely (probably more likely) that the differences that were posted were just the kinds of differences you get from one measurement pass to the next, even on the same headphone. As you can see in my measurements above, even two passes on the exact same headphone can result in notable differences.

So it's not about assuming people are doofuses, but knowing that measurements are interpreted very literally by many -- every dip and peak.



We actually do have some planned, and just have to find the time to put them together. The measurement process can be interesting and challenging.

I will occasionally post things like the above long-winded post -- yes, other long-winded posts -- to explain things about measuring that hopefully some people find interesting or surprising.

If you haven't already seen our Apple AirPods Max measurement steps, you should check out these links:

https://www.head-fi.org/threads/airpods-max.949152/page-98#post-16058062​

​

https://www.head-fi.org/threads/airpods-max.949152/page-127#post-16071445​

​

https://www.head-fi.org/threads/airpods-max.949152/page-153#post-16091082​

Here's a post about measuring headphone sensitivity (and headphone sensitivity specs):

https://www.head-fi.org/threads/sennheiser-hd650-massdrop-hd6xx-impressions-thread.197776/post-13803816​

I hope we can get to making videos about headphone measurements in the near future. There's a lot of ground to cover with the fascinating Brüel & Kjær 5128, we did (and are still doing) ANC testing and MOS (mean opinion score) testing with Head Acoustics, I'd like to show the fun that can be had with the Audio Precision audio analyzer, and more. It's just about finding the time.

Thank you for the above links, Jude. And I look forward to seeing more on the subject of measurements, when you get the chance to post something on that. I know the videos take a bit more work. But those are also good because I can listen to them while doing some other things, like reading/plotting graphs, doing dishes, and so forth.

Will you be posting the measurements of the DCA Stealth and other headphones made on the 5128 rig in the Head-Fi Audio Measurement Lab when they become available btw? Because it would be sort of nice if that data was all available and easy to get at in one place. And potentially make it a bit easier to maybe try out some DF and other compensation on some of the graphs.

 

[ADUHF]

Some older articles and videos on headphone measurements by Tyll of Inner Fidelity.

There are other videos on YouTube where Tyll goes into alot more detail on how to interpret his graphs and measurements as well. The videos below mostly just show how some of the measurement equipment is set up and works.

https://www.stereophile.com/content/innerfidelity-headphone-measurement-procedures

https://www.stereophile.com/content/innerfidelity-headphone-measurements-explained







More of his articles from Inner Fidelity: https://www.stereophile.com/writer/15081

 

[someguyontheinternet]

I'm not really that up on the whole SINAD is good/bad thing. But since this seems to tie in well with generally sciency thrust of this thread/forum, here is another recent video on the subject from Resolve. Feel free to give your opinions below.

"There have been a number of studies that demonstrate that even at fairly high levels of THD that did not have any correlation with listener preference."
This is just completely misleadingly worded. It's not technically wrong as he is referring to 1% distortions (which is only added as a sidenote and not clearly articulated in the argument) that did not correlate. In fact the paper shows that THD does correlate to subjective listener impressions! For the headphones that showed highest THD below 100Hz and above 1kHz the subjective commentary matched the measurement. This is however not addressed at all in the video. There are bits and pieces taken from the paper that support the claim, but the bits and pieces that oppose the claim in the same paper are conveniently left out. To me that is just not acceptable.

Another argument he brings up is the relevance of noise. The claim he makes is that the ability to drive different sensitivity/impedance headphones should be questioned.
In this hobby I'm gonna try different headphones which of course will have variance in their sensitivity. Having a single amplifier that can drive all headphones without noise as opposed to one that can only drive headphones without noise starting from some sensitivity/impedance level is a complete no-brainer decision to me. Having a single amplifier that gives me the flexibility to try any headphone I want without having to worry about noise is just better to me. If someone is sure they will not buy any additional headphones ever in their life and only has low sensitivity/ high impedance headphones, sure they can go ahead and ignore a higher noise floor. But even budget option amplifiers can reach a low noise floor paired with high power output that can drive most headphones. So high noise floor is not really justifiable in my opinion.

The thing I do agree with is that a single THD measurement value is useless when evaluating headphones. There should be multiple measurements across the frequency range at ex. 70/80/90db to see if the driver can handle the entire dynamic range that may occur in a song without distorting at any particular frequency area.
I also do agree that ranking SINAD can be misleading. SINAD matters when it's in the audible range. Beyond that it can only satisfy the worries of people with high levels of neuroticism like myself.

However that does not mean that THD and SINAD measurements are useless. They have a purpose and context, but if taken out of that context (as many people tend to do), they will be useless and/or misleading. But this is true of most useful things.

 

[bigshot]

I might be wrong, but isn't distortion usually measured at 1kHz?

 

[ADUHF]

I might be wrong, but isn't distortion usually measured at 1kHz?

There are others here who can probably answer this better than I can, bigshot. I think you could be right though, when it's given simply as a percentage in a piece of gear's specs.

I think distortion can vary, sometimes quite a bit, across the frequency range on a pair of headphones though. Which is why headphone reviews will often show the different levels on a graph between 20 Hz and 20 kHz, in much the same way as a headphone's frequency response, or impedance.

 

[ADUHF]

Descriptions of SINAD and THD+N from Wikipedia, for those who may not be familiar with some of the differences:

https://en.wikipedia.org/wiki/SINAD
https://en.wikipedia.org/wiki/Total_harmonic_distortion#THD.2BN

 

[pila405]

I might be wrong, but isn't distortion usually measured at 1kHz?

In what context? Many provide THD vs frequency for the entire range. (20-10kHz or more), which gives a much better picture of course.

Regarding Resolve's video, I find it insincere. Assuming one is provided with SINAD over a certain threshold (110db or whatever), knowing the system with which you take the measurement has a -160db noise floor, it actually tells you quite a lot already. Besides that, one can find a lot of information on the internet regarding many devices, providing measurements for IMD vs frequency, THD+N vs frequency, impedance and more, all for one device, which gives you the full picture. That's why good brands which provide well-engineered products also supply the full picture in terms of specs for each of their products and not only 1 meaningless number.

Also, Andrew himself finds it very convenient to talk about and give his subjective impression regarding different equipment which measures well below audibility and is completely transparent and yet he miraculously finds all these differences which he conjures up in his mind.



Yeah.....this throws his credibility out the window when talking about DACs\amps. Would say the same for headphones, but I guess that's a different topic

One can also hear the effect of THD for himself and see whether they can perceive it or now and to what levels (assuming your system outputs a clean enough signal):
https://www.audiocheck.net/testtones_thd.php
For example, using my rig with Ether CX, I can easily hear down to 0.01% THD at 125Hz (though I doubt the headphones produce less than 0.01% harmonic distortion at 125Hz, so I think there is some additional problem here) and 0.05% at 1kHz (although harder)

 

[bigshot]

Most spec sheets list a single figure for THD. I was under the impression that this figure represents distortion at 1kHz, which sits at the most sensitive spot of human hearing. THD can be expressed either as a percentage or an amount of attenuation in dB. 1% THD is -40dB, .5% is -46dB and .1 is -60dB. That rating is established with test tones, but distortion in test tones is not really reflective of the real world impact of distortion on music. In the experiment by Ethan Winer (see the link in my sig file) he takes a horrible buzzing sound- the absolute worst thing you could mix with music- and drops it a bit at a time under the music until it is inaudible. By the time it gets to -40dB, there is no way it can be heard except in the total silence between tracks, which would be totally silent if the buzzing sound was distortion and not noise.

Audiophiles often overstate the importance of distortion when it's a number on a page, yet they can't hear it in practice. I see numbers like .1% being thrown around in audiophile circles as being the minimum acceptable THD, while they wax rhapsodic over the sound of LP records that have distortion levels of as much as 3-4% (-30dB) at the inner groove. The truth is, with digital audio and solid state electronics, distortion isn't really an issue. Even cheap electronics has distortion levels far below audibility. And that's true for just about all specs... response, timing, crosstalk, etc. Distortion isn't much of an issue with most headphones either. It is relatively simple to minimize distortion with the small drivers in headphones. Are there any headphones with THD over 1%? I don't think I've ever run across any. Response is MUCH more important there. Distortion might be more of an issue with speakers, but again, that is dwarfed by the importance of a balanced response.

I might be wrong, but it seems that an awful lot of discussion in audiophile circles is about theoretical things that really don't matter a heck of a lot when someone sits down to listen to Mozart in their living room.

 

[ADUHF]

Again, take anything I have to say on this subject with a grain of salt. Because this is not something I've looked into that deeply.

But I'm not that sure about the 1 kHz thing, when it comes to distortion. (Which is why I've struck that comment in my previous post above.) Based on this article, it looks like the THD and THD+N ratings can potentially cover a range of frequencies (from 20 Hz to 20 kHz?), even when given as just a single value or percentage...

https://www.ap.com/blog/thd-and-thdn-similar-but-not-the-same/

This may (or may not) apply mostly to devices with a fairly flat/linear response though, like an amp or DAC, where the distortion or THD+N does not vary significantly over the audible frequency range.

How a single distortion or THD+N value would be handled on speakers or headphones, I'm not really sure. I think distortion could be more of an issue in the lower frequencies on speakers though. So it probably wouldn't make sense to measure them at 1 kHz. And would be better to show the results on a graph, like it's often done for headphones.

I believe that the sensitivity or efficiency in dB/V or dB/mW SPL on headphones are often, or generally measured at 1 kHz (or thereabouts) though. Which is a different topic.

 

[ADUHF]

A couple other data points that might be useful to some on the issue of distortion...

This is a PDF that was referenced in the Wikipedia article on SINAD that explains a little more about how SINAD, THD+N, and various distortion and noise ratings relate to one another...

Understand SINAD, ENOB, SNR, THD, THD + N, and SFDR so You Don't Get Lost in the Noise Floor
https://www.analog.com/media/en/training-seminars/tutorials/MT-003.pdf

 

[ADUHF]

And this is the video by Julian Krause that Resolve referenced in has SINAD discussion that shows some of the differences between various types of distortion. And how masking can effect your perception of some of them. This was fairly informative imo, as a newbie to this topic.



I believe the levels of distortion that are used for this demonstration are much higher than what you'd typically find in most audio gear btw.

 

[ADUHF]

Another discussion on this topic with Earl Geddes.

 

[bigshot]

Perhaps it is measured from 20 to 20. But the distortion around Fletcher Munson would be the most critical I would think. Distortion in the upper mids is always what makes my teeth scrape.