[jude]

@jude what's your take on Hi-Res thing and do you think 40kHz extension matter for IEMs at all?...

@piotrus-g, I'm just trying to get meaningful headphone measurements out to 20 kHz, and haven't given this as much thought! Do I personally look for the "Hi-Res" label as part of my buying decision? No. I'm not clear on precisely how headphones are measured to 40 kHz to earn the "Hi-Res" seal -- no idea what the procedure is, how it's done. If you've had your IEMs measured for that certification, how does that work?

At the 2016 AES International Conference on Headphone Technology, Steve Temme from Listen, Inc. presented a poster covering this. From my notes and photos:

• His poster quoted David Griesinger: "The phase relationships of harmonics from a complex tone contain more information about the sound than the fundamentals."
  
  
• Temme asked, "So, why exceed 20 kHz?" His answers:
  • "For a transducer to have a flat phase response out to 20 kHz, a flat magnitude out to 40 kHz is required (by definition, at resonance, there is a 90° phase shift)"
  • "Inaudible high frequency components, e.g. 30 & 33 kHz resonances, can produce intermodulation products that 'beat' down into the audio band of frequencies"

EDIT 2017-11-15 1900 EST: I should also add these things from his poster:

• "A sampling rate of 44.1 kHz @ 16 bits adequately covers the range of human hearing and musical instruments"
• "To achieve a flat frequency response (mag & phase) out to 20 kHz requires a bandwidth beyond 20 kHz"
• "Loudspeakers and microphones that measure beyond 20 kHz can exhibit trade-offs in performance, e.g. audible intermodulation distortion"
• "Try down-sampling a hi-res audio file to CD quality and do an ABX listening comparison. Can you hear the difference?"

All that said, I've been enjoying my favorite IEMs without any knowledge or evidence of their performance up to 40 kHz.

 

[ThomasHK]

@jude what's your take on Hi-Res thing and do you think 40kHz extension matter for IEMs at all?

I know this might be even more controversial than discussing measuring rigs or 10kHz accuracy. I'm just wondering what is your view.

Way to open a whole other can of worms

 

[twister6]

...
I know one of the learning experiences for me was with the newer qJays where I couldn't understand why people were calling them sibilant, and others weren't (I heard no sibilance). It was only afterwards when it was explained to me that for such a small IEM, insertion depth, angle, and canal physiology could have significant effects on the higher frequencies that I had that "aha" moment. Until then I'd been a bit of a prat about it - afterwards was able to appreciate both sides, and completely modify my stance from that time forward. One of the benefits of some of the expertise on the forum. You're always learning, and slowly modifying your thoughts and understanding with reviewing. An enlightening and often humbling journey.

add CFA VEGA to this list. The discrepancy of people hearing/not-hearing sibilance bothered me too because I don't hear it, and I even listened to suggested test tracks which suppose to reveal it. Btw, I received the same explanation about variables you listed above, which all contribute to the perception of the sound around 6k peak.

 

[piotrus-g]

@piotrus-g, I'm just trying to get meaningful headphone measurements out to 20 kHz, and haven't given this as much thought! Do I personally look for the "Hi-Res" label as part of my buying decision? No. I'm not clear on precisely how headphones are measured to 40 kHz to earn the "Hi-Res" seal -- no idea what the procedure is, how it's done. If you've had your IEMs measured for that certification, how does that work?

At the 2016 AES International Conference on Headphone Technology, Steve Temme from Listen, Inc. presented a poster covering this. From my notes and photos:

• His poster quoted David Griesinger: "The phase relationships of harmonics from a complex tone contain more information about the sound than the fundamentals."
  
  
• Temme asked, "So, why exceed 20 kHz?" His answers:
  • "For a transducer to have a flat phase response out to 20 kHz, a flat magnitude out to 40 kHz is required (by definition, at resonance, there is a 90° phase shift)"
  • "Inaudible high frequency components, e.g. 30 & 33 kHz resonances, can produce intermodulation products that 'beat' down into the audio band of frequencies"

EDIT 2017-11-15 1900 EST: I should also add these things from his poster:

• "A sampling rate of 44.1 kHz @ 16 bits adequately covers the range of human hearing and musical instruments"
• "To achieve a flat frequency response (mag & phase) out to 20 kHz requires a bandwidth beyond 20 kHz"
• "Loudspeakers and microphones that measure beyond 20 kHz can exhibit trade-offs in performance, e.g. audible intermodulation distortion"
• "Try down-sampling a hi-res audio file to CD quality and do an ABX listening comparison. Can you hear the difference?"

All that said, I've been enjoying my favorite IEMs without any knowledge or evidence of their performance up to 40 kHz.

Thanks for your reply.
Those are some interesting points.
The 40kHz output it measured as open baffle mounted speaker at some distance from microphone. SPLs are typically measured around -60dB or something like that - again I'm speaking from memory of papers I saw some time ago which I cannot find at the moment, so don't take my words for granted.
In IEMs I think intermodultion and phase shifts caused by crossovers in multi-BA set ups would be (and should be) more of a concern than what happens after 20kHz, as those have much more audible consequences.
If you think about it planar and electrostatic headphones would be benchmarks for High-Res certification then
We don't certify our IEMs for Hi-Res.
That said I'm definitely not dismissing the idea, it's certainly controversial ATM, but if it pushes the industry to create better drivers etc. I'm happy with that!

Way to open a whole other can of worms

 

[TheGiantHogweed]

Hope this stays on topic with this thread. I think I have always been really sensitive to high frequency sounds for my age. But my headphones that I use are the AKG K550s Audio Tecknica ATH-AD700 (had these for about 7 years I think) and more recently, the AKG K702. All of these have some strong peak in the treble. Especially the AD700. I love this type of sound. But doing a test from this website here down below, I can just about hear 19kHz on a moderate volume.
http://www.szynalski.com/tone-generator/
The video posted earlier in this thread went up and down and all over the place with strange other affects, not just the frequency.
19kHz is very quiet, but really high pitched and painful to hear. But I'll be honest, it sounds like a lot CTR monitors / TV's do when you turn them on, or the whole time with certain ones. I often notice that in a lot of shops, they still use CRT screens for the to display the images from the CCTV. And I very often notice this high pitched noise from the monitor and it drives me insane! Enough that I sometimes just can't stand being in there. Is this unusual for my age to notice these things? I'm only just over 20. With my headphones, I'm very keen on a lot of treble, but sometimes, with certain recordings, you can hear really high pitched squeaks in the background. This makes me wonder if I'm so sensitive, I can hear some buzzing or something from the amplifiers they are using in the studio? This is a particular track that I like the music of, but I just can't work out what causes these high pitches squeaks:

I hear a bit between 0:28 and 0:30, then stops, then 0:33. Then one of the most obvious ones is at 0:44 to 0:48. I just can't tell if these are intended to be there or not. They are certainly not pleasant to hear. But it is something I only really notice on my headphones. But some other people I know don't notice it. It sounds even worse on Spotify as the recording is more clear there. Would be interesting if any others notice it on this forum. If not, maybe I just have super sensitive hearing. As something else that often really bothers me is phone chargers. Unless they are of really good quality, they often buzz or make a high pitched squeak. If they are like that, I can't sleep in the same room, no matter how far away they are. The only people who seem to notice it easily are those who are younger than me.

 

[ubs28]

According to my ears, the measurements of Tyll are correct. The Z1R for example has a clear peak. How else can a dark headphone sound bright at some sections of a song? I listened to the same sections with other headphones and the tonal balance remained stable. Not with the Z1R which moved from dark to bright and back to dark again depending which sections of the song it was.

If Jude his measurements are correct, then I don’t know to say. Because what I’m hearing does not align with his measurements at all. There is something funny happening with some headphones he measured which do not show up in his graphs.

About why the 10khz is so different? That is impossible to say without having both Tyll and Jude their machines and start doing some analysis. It’s probably due to the shape of the ear canal or an other parameter that is different between both setups.

But both setups could be wrong in the end, because who says that the modelled ear that Tyll and Jude use is the ear that most people have in general?

 

[DivineCurrent]

Contributing to this discussion, when it comes to IEMs, measurements and perceived frequency response are even more complicated. Jude, you already mentioned ear resonances and how many rigs differ, but on top of all that, every person has an individual resonances in their own unique ear. So, what may sound bright to someone, may sound warm to someone else, simply because of the physical differences in their ear. IEMs do not interact with the outer ear reflections, so the natural peaks you usually hear from over ears will not be there. Or, there will be different peaks in different places your ear is not used to. There’s just so, so many factors going on with measurements.

 

[castleofargh]

According to my ears, the measurements of Tyll are correct. The Z1R for example has a clear peak. How else can a dark headphone sound bright at some sections of a song? I listened to the same sections with other headphones and the tonal balance remained stable. Not with the Z1R which moved from dark to bright and back to dark again depending which sections of the song it was.

If Jude his measurements are correct, then I don’t know to say. Because what I’m hearing does not align with his measurements at all. There is something funny happening with some headphones he measured which do not show up in his graphs.

About why the 10khz is so different? That is impossible to say without having both Tyll and Jude their machines and start doing some analysis. It’s probably due to the shape of the ear canal or an other parameter that is different between both setups.

But both setups could be wrong in the end, because who says that the modelled ear that Tyll and Jude use is the ear that most people have in general?

G.R.A.S! G.R.A.S says it. ^_^ they're the ones making statistical models based on so many people. the model is fairly reliable for what it is, a statistical average of human head. but the dummy head always had troubles and limitations when it came to following exactly that simulation. over the years they have logically come up with small changes to increase the similarity between the model and the measured results.
how that statistical model of average human head does or doesn't yield similar results to your ear or my ear, is in effect irrelevant. I don't know for you, but it doesn't take long for anybody looking at me to guess that I'm not an ideal specimen of the statistical average. the dimensions are all wrong(insert dick joke here). so when I don't perceive what I see on a graph, I don't know if the graph is right, but I have relevant reasons to suspect that I'm not(again only in the context of an average reference).

another approach would be to ask my ex mother in law about the right target. she's never wrong.

 

[james444]

@piotrus-g, I'm just trying to get meaningful headphone measurements out to 20 kHz, and haven't given this as much thought! Do I personally look for the "Hi-Res" label as part of my buying decision? No. I'm not clear on precisely how headphones are measured to 40 kHz to earn the "Hi-Res" seal -- no idea what the procedure is, how it's done. If you've had your IEMs measured for that certification, how does that work?

Thanks for your reply.
Those are some interesting points.
The 40kHz output it measured as open baffle mounted speaker at some distance from microphone. SPLs are typically measured around -60dB or something like that - again I'm speaking from memory of papers I saw some time ago which I cannot find at the moment, so don't take my words for granted.

Here's a white paper by Knowles on how they measure their BAs for Hi-Res audio:
http://www.knowles.com/eng/premiums...-Armature-Drivers-for-Hi-Res-Audio-Earphones#

 

[castleofargh]

Here's a white paper by Knowles on how they measure their BAs for Hi-Res audio:
http://www.knowles.com/eng/premiums...-Armature-Drivers-for-Hi-Res-Audio-Earphones#

that pic makes me feel so much better about the way I measure stuff sometimes ^_^.

 

[MacedonianHero]

According to my ears, the measurements of Tyll are correct. The Z1R for example has a clear peak. How else can a dark headphone sound bright at some sections of a song? I listened to the same sections with other headphones and the tonal balance remained stable. Not with the Z1R which moved from dark to bright and back to dark again depending which sections of the song it was.

If Jude his measurements are correct, then I don’t know to say. Because what I’m hearing does not align with his measurements at all. There is something funny happening with some headphones he measured which do not show up in his graphs.

About why the 10khz is so different? That is impossible to say without having both Tyll and Jude their machines and start doing some analysis. It’s probably due to the shape of the ear canal or an other parameter that is different between both setups.

But both setups could be wrong in the end, because who says that the modelled ear that Tyll and Jude use is the ear that most people have in general?

If you watch this video with Paul Barton (founder of the world renowned PSB Speakers), Tyll admits that the peak is likely not really there due to no bend in his dummy head's ear canal:

https://www.innerfidelity.com/content/canjam-rmaf2017-people-i-love-paul-barton

I've heard a half dozen Z1Rs and there isn't really a peak to speak of. If the pair you listened had one, then there was an issue with that pair.

 

[jude]

If you watch this video with Paul Barton (founder of the world renowned PSB Speakers), Tyll admits that the peak is likely not really there due to no bend in his dummy head's ear canal:

https://www.innerfidelity.com/content/canjam-rmaf2017-people-i-love-paul-barton

I've heard a half dozen Z1Rs and there isn't really a peak to speak of. If the pair you listened had one, then there was an issue with that pair.

I don't know if the canal is the only reason (or the reason at all) in this case. I'm not saying it isn't (I do think it has some role here), only that there are other factors involved that we didn't systematically rule out (or rule in). Different manikins, different pinnae/canals, different ear simulators, and more. Again, different setups, different results.

Also, while you, I, and others don't feel like the 10 kHz peak in Tyll's measurements correlate to what we're hearing, there are others who feel they do. Based on the experiences we've had with some simple experiments (one very simple one of which involves the HD650 that we'll talk about after we update the Buying Guide), there is interesting discussion to be had in comparing experiences different people have with the same headphone. We'll start with one example that involves one headphone (the HD650) and two people. (Don't expect a rigorous Sean Olive type of study -- again, it's just a super easy-to-understand yet impactful example.) In this specific case, do I think most will hear the Z1R as measured in our office, or as measured by InnerFidelity? I suspect the former -- and I won't speak for Tyll but his subsequent posts suggest to me that he might also.

Even if there was such a thing as being able to perfectly capture the average human (in all relevant respects) in the form of a measurement manikin or other test fixture, it would still churn out measurements and results that would be dissatisfying to some. I shared some thoughts about endeavoring to have measurements that make sense to most people (independent of preference) in a post at the following link: Headphone Measurements -- Is Average The Best We Can Do?

 

[MacedonianHero]

I don't know if the canal is the only reason (or the reason at all) in this case. I'm not saying it isn't (I do think it has some role here), only that there are other factors involved that we didn't systematically rule out (or rule in). Different manikins, different pinnae/canals, different ear simulators, and more. Again, different setups, different results.

Also, while you, I, and others don't feel like the 10 kHz peak in Tyll's measurements correlate to what we're hearing, there are others who feel they do. Based on the experiences we've had with some simple experiments (one very simple one of which involves the HD650 that we'll talk about after we update the Buying Guide), there is interesting discussion to be had in comparing experiences different people have with the same headphone. We'll start with one example that involves one headphone (the HD650) and two people. (Don't expect a rigorous Sean Olive type of study -- again, it's just a super easy-to-understand yet impactful example.) In this specific case, do I think most will hear the Z1R as measured in our office, or as measured by InnerFidelity? I suspect the former -- and I won't speak for Tyll but his subsequent posts suggest to me that he might also.

Even if there was such a thing as being able to perfectly capture the average human (in all relevant respects) in the form of a measurement manikin or other test fixture, it would still churn out measurements and results that would be dissatisfying to some. I shared some thoughts about endeavoring to have measurements that make sense to most people (independent of preference) in a post at the following link: Headphone Measurements -- Is Average The Best We Can Do?

I'm curious to see how repeatable your treble responses would be if you remove the headphones and replace them on the dummy head. Have you guys performed a repeatability study based on this?
Thanks Jude.

 

[jude]

I'm curious to see how repeatable your treble responses would be if you remove the headphones and replace them on the dummy head. If you remember the work I did a few years back, that showed that this also had a great effect on measurements north of 6kHz. Have you guys performed a repeatability study based on this?
Thanks Jude.

Yes, we always do multiple seatings -- more with circumaurals and supra-aurals than in-ears (which are going into a canal with dimensions that are more limiting with respect to placement). We always do both tightening and loosening re-seats, and headphone-off-and-back-on reseats. At some point after we finish the Buying Guide update, we'll post the results from four different Z1R's.

 

[MacedonianHero]

Yes, we always do multiple seatings -- more with circumaurals and supra-aurals than in-ears (which are going into a canal with dimensions that are more limiting with respect to placement). We always do both tightening and loosening re-seats, and headphone-off-and-back-on reseats. At some point after we finish the Buying Guide update, we'll post the results from four different Z1R's.

Thanks Jude!