tmtomh
Major Contributor
- Joined
- Aug 14, 2018
- Messages
- 3,470
- Likes
- 10,427
No.
Last edited:
-
Welcome to ASR. 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!
"Bias" of some members towards headphone measurements?
- Thread starter Dazerdoreal
- Start date
oleg87
Addicted to Fun and Learning
What?
Distortion measurements should measure distortion (ie, nonlinearities in the response). A measurement protocol to measure soundstage should be designed to measure soundstage. Considering an illusion of soundstage would necessarily depend on your particular head/ears, it's far from obvious how you would measure this in a useful way, RTings' dubiously evidenced attempts aside.
Last edited:
Had not seen this thread before and somehow landed on page 8 where member who used to work at Harman was asking why we use a headphone fixture whereas we don't for speakers.
There is a simple answer: the output of a headphone is very low as to try to measure it in an anechoic chamber or Klippel NFS. The fixture solves that problem by sealing the space between the headphone/IEM and putting the microphone very close to it. This in turn causes acoustic coupling issues due to frequencies involved. To solve this standardized fixtures were developed which while not perfect, is all we have. The fixture also removes the external aspects of the headphone from interfering with the measurements.
Given what we are stuck with, then research was performed to see what measurements correlate with preference. To the extent that was successful, then we have nullified the negative effects of using a fixture.
Same thing happened with speakers where an anechoic measurement succeeded as reference as opposed to say, always measuring a speaker in certain room.
There is a simple answer: the output of a headphone is very low as to try to measure it in an anechoic chamber or Klippel NFS. The fixture solves that problem by sealing the space between the headphone/IEM and putting the microphone very close to it. This in turn causes acoustic coupling issues due to frequencies involved. To solve this standardized fixtures were developed which while not perfect, is all we have. The fixture also removes the external aspects of the headphone from interfering with the measurements.
Given what we are stuck with, then research was performed to see what measurements correlate with preference. To the extent that was successful, then we have nullified the negative effects of using a fixture.
Same thing happened with speakers where an anechoic measurement succeeded as reference as opposed to say, always measuring a speaker in certain room.
Jim Taylor
Major Contributor
- Joined
- Apr 7, 2024
- Messages
- 1,235
- Likes
- 4,042
I'm not so sure of that. Many people associate lateral spread with soundstage, and I agree that can be "created". A good recording of an orchestra will certainly exhibit lateral spread that is natural, and not "created".
I also associate soundstage with depth, and that is more difficult to recreate. Please listen to this selection and tell me what you think:
I would also listen to OneMic recordings, and see whether you hear depth in addition to lateral spread.
For depth that is created rather than recreated, just listen to Holly Cook's "Angel Fire", or Roger Waters' "Amused to Death", made with QSound. They are entertaining, but to me they are very unnatural.
P.S. - my 'phones are Audio Technica ATH-M40x ... not the best but not the worst, either.
Last edited:
The measurements are meaningful for a certain purpose. That purpose is not to help with a personal choice. Hence the curves are not useful to sell the product to yourself.
Other than with speakers the interpretation of headphone measurements takes several steps. In some the 'absolute truth' is dismissed deliberately. Example given when deciding to tune to 'diffuse field'. The latter even pulls the validity of stereo into question (besides the missing crossfeed).
It is not easy to dig into the topic, and it may become a bit frustrating also, not to have an ideal. That's why, and I confirm your observations.
Many times I asked myself, if people actually agree to a preference rating scheme, as there is a personal preference needed to begin with. It should be as easy as can be, but not with real enthusiasts it seems.
WOW!!!!! Those recordings are amazing!
Jim Taylor
Major Contributor
- Joined
- Apr 7, 2024
- Messages
- 1,235
- Likes
- 4,042
I would express these points differently.
1) Preference is personal. Learn how your preference deviates from a standard, such as the Harman curve. This is called "correlation".
2) You can then apply this correlation to measurements to predict with more accuracy whether the headphone under consideration is right for you or not. This way, you have at least one reproducible facet of your search (measurements) so that you won't need to "trust your ears". Trusting to sensory input is notoriously unreliable, because it is not reproducible and is subject to many biases.
Last edited:
I decidedly won't pick on you, but did you already try to optimize your headphones towards your preference? I tried and, for the time being, failed.
A flat coupler came to use, not an HATS. Adjusted the frequency response to a smoothly falling line as we expect from speakers. Correlated to white noise to be smooth. First obstacle was, it depends a lot on sound volume! Second, with music it didn't sound to good. Third, after some time not listening to the headphones the impression changed again.
Maybe white noise is to sensitive to serve as a sensible benchmark. Also a smooth line isn't a goof target either, as it doesn't reflect my directional HRTF for what I prefer the sound to come from.
Pink noise is less revealing at the top end. I see that the topic gets heated once the universal validity of "the Harman" is put to doubt. Let's see how the scene develops once the other, more detailed Harman is officially released. People seem to need curves, and then they will have a new one.
Edit, there's already a new one, more details: https://crinacle.com/2025/02/05/the-new-2025-ief-target/
The whole article is worth the time reading for the fun of it.
I find our new target a bit embarrassing in that it replicates, or shall I say predicts, my preference for a sharper peak, lower in frequency for the main ear gain. The "10k" dip comes to the rescue, as it is defined (a pun) at 11k now, but I like it a lower in frequency. Not so normal.
Edit, there's already a new one, more details: https://crinacle.com/2025/02/05/the-new-2025-ief-target/
The whole article is worth the time reading for the fun of it.
I find our new target a bit embarrassing in that it replicates, or shall I say predicts, my preference for a sharper peak, lower in frequency for the main ear gain. The "10k" dip comes to the rescue, as it is defined (a pun) at 11k now, but I like it a lower in frequency. Not so normal.
Last edited:
Thanks Jim, what an excellent recording. I was trying to listen to Tchaikovskys Violin concerto last night on Tidal and the orchestra was very left and right with the soloist appearing and disappearing into the sound stage. My hearing isn't very good and the dynamics were causing me problems. I didn't know if it was the performer the recording or my DAC headphone combo but the contrapunction between soloist and first violin seemed to be awry I thought it might be exaggerated separation or perhaps the conductor. Would you mind having a listen if you have Tidal? I'll try what you have suggested but I think you might have saved me a packet. I'm using the RME ADI-2 DAC fs through a Topping A70 Pro into Anandas. I was going to try a MOJO into HD800s or Audeze LCD-X at a local hifi shop out of curiosity following Amirms review. Thanks for your help.I'm not so sure of that. Many people associate lateral spread with soundstage, and I agree that can be "created". A good recording of an orchestra will certainly exhibit lateral spread that is natural, and not "created".
I also associate soundstage with depth, and that is more difficult to recreate. Please listen to this selection and tell me what you think:
I would also listen to OneMic recordings, and see whether you hear depth in addition to lateral spread.
For depth that is created rather than recreated, just listen to Holly Cook's "Angel Fire", or Roger Waters' "Amused to Death", made with QSound. They are entertaining, but to me they are very unnatural.
P.S. - my 'phones are Audio Technica ATH-M40x ... not the best but not the worst, either.
Jim Taylor
Major Contributor
- Joined
- Apr 7, 2024
- Messages
- 1,235
- Likes
- 4,042
I would love to help you, but I do not subscribe to Tidal ... or any other streaming service. I have a small (very small) library on-site.
I have the Julia Fischer recording, with the Russian National Orchestra under Yakov Kreizberg. Other than Ms. Fischer being miked a bit too close, I don't hear anything strange. However ... there is another Fischer recording with the Orchestre Philharmonique de Radio France under under Vasily Petrenko. That recording seems to be miked differently.
There are several other recordings of the Violin concerto (Baeve and, I believe, Hahn) but I've not heard them. Sorry that I can't be of further help.
Indeed. This seems yet another bait thread inspired by another shape-shifting troll.I don't know how we can endlessly debate some of these issues.
Using modern measurement tools we can measure speakers and closely determine how well it can reproduce a waveform
both anechoic and into some known room parameters.
Preference is preference but when we discuss a tool for High Fidelity Music REPRODUCTION, a speaker is little different from any other source.
It's output should mirror it's input as accurately as possible.
Unfortunately speakers are as yet a long way from being as accurate as an amp, and the room interface is a huge unknown factor.
But still, within small margins we are able to measure and determine which of those speakers come quite close, and which miss by a mile.
If accuracy is important to the listener, we then have to choose the speakers from within the "very good" measurement group, which ones will best suite
our personal preferences and needs for room size, spl, etc.
If accurate reproduction isn't what you want, you can even use the spec's to find those that will please the boom and sizzle crowds.
We've come a long way baby.
It is actually dead easy. Measurements tell you how true to the original signal a component (or chain thereof) can get.
Nobody forces anyone to declare that feat of engineering their favorite. People can prefer whatever they want. And due to many possible factors, subjective hearing may respond better (or worse) to certain deviations from that ideal linear response. Plus let's keep in mind many recordings do not deserve nor care about utter accuracy.
I KNOW my personal preference deviates some from. total linear accuracy. I also know how to read that in measurements. ASR is unique in offering that training as well as data. While Amir has his preferences, he will never shoot you down for drawing your own conclusions from the measurements he provides.
Last edited:
Old_School_Brad
Major Contributor
- Joined
- Sep 10, 2024
- Messages
- 1,864
- Likes
- 2,616
Exactly.
But to see that, you need to start with the right foundation. You need something solid to build on.
That foundation is objective measurements -neutrality, flat response, and linearity. Once you’ve got that down, you can test yourself and figure out your own hearing thresholds. Then, choosing gear becomes much easier because you’ll know you just need equipment that meets those thresholds.
After that, you’re free to explore and fine-tune things according to your own taste, without second-guessing the basics.
Simple.
PEQ friendly I call it.
What's funny is people thinking slight differences in frequency response are the only thing that matters, misinterpreting what Sean Olive actually ever claimed.
I see many reviewers now in the weeds about tiny differences in tonal balance, but within a certain difference threshold, frequency response differences are relatively meaningless for most people. And then they won't cover any other measurement!
Humans are way more sensitive to timing and other phase distortion, even tiny amounts. Furthermore, non-linear distortion like IMD is far worse sounding than total harmonic distortion, and in speakers and headphones you also need impulse response and CSD waterfall plots along with the aforementioned measurements to identify resonances and ringing to really say anything definitive *after* all the other measurements. Square waves can also be useful, but only if you know what they're supposed to look like for the frequency response target, and understanding they're composite 'slices' of the other measurements. QA/QC is something often ignored, e.g. I just tried a Truthear Gate, and there's no way mine measures like the reviewer data I've seen for it.
But back to my main point: you cannot simply EQ any two headphone models to sound the same, and sometimes that's even the case for two units within the same model! Since most reviewers do not actually measure *all* the stuff I just covered, you will indeed have to use your ears, and the measurements are just a canary in the coal mine. Just because the canary is still awake, doesn't mean everything in the mine is perfect. They are just a tool, and just because a few measurements look good, doesn't mean the gear is kosher. If all these measurements *are* covered, then, yes, my personal confidence would be quite high for a given headphone, IEM, and speaker. For other gear, btw, you might need stereo crosstalk, jitter, etc, and I have no doubt someday a super nerd will figure out even more obscure measurements.
I see many reviewers now in the weeds about tiny differences in tonal balance, but within a certain difference threshold, frequency response differences are relatively meaningless for most people. And then they won't cover any other measurement!
Humans are way more sensitive to timing and other phase distortion, even tiny amounts. Furthermore, non-linear distortion like IMD is far worse sounding than total harmonic distortion, and in speakers and headphones you also need impulse response and CSD waterfall plots along with the aforementioned measurements to identify resonances and ringing to really say anything definitive *after* all the other measurements. Square waves can also be useful, but only if you know what they're supposed to look like for the frequency response target, and understanding they're composite 'slices' of the other measurements. QA/QC is something often ignored, e.g. I just tried a Truthear Gate, and there's no way mine measures like the reviewer data I've seen for it.
But back to my main point: you cannot simply EQ any two headphone models to sound the same, and sometimes that's even the case for two units within the same model! Since most reviewers do not actually measure *all* the stuff I just covered, you will indeed have to use your ears, and the measurements are just a canary in the coal mine. Just because the canary is still awake, doesn't mean everything in the mine is perfect. They are just a tool, and just because a few measurements look good, doesn't mean the gear is kosher. If all these measurements *are* covered, then, yes, my personal confidence would be quite high for a given headphone, IEM, and speaker. For other gear, btw, you might need stereo crosstalk, jitter, etc, and I have no doubt someday a super nerd will figure out even more obscure measurements.
- Joined
- Jun 24, 2024
- Messages
- 2,714
- Likes
- 4,694
What's your source for this, as everything I've seen from the research says otherwise as far as I can recall.
Why?
Resonances and that don't show up in the frequency response aren't a concern. As Amir has noted many times, what a waterfall shows depends heavily on what value you put at the bottom of the y-axis, and there's no objective means to determine where it should be. Impulse response is also of questionable value, there's little objective way to evaluate it.
For engineering purposes, sure. For the end customer it's meaningless.
That's not something a reviewer can reasonably evaluate.
Unless you've taken a measurement, there's no way for you to know that. Ears are not measurement instruments. For IEMs, this could easily just be a fit issue anyway.
Crosstalk is regularly measured here, but it's pretty much never an audible concern outside of turntables. Jitter is also a non-issue.
Discrimination of low-frequency tones employs temporal fine structure - PubMed
An auditory neuron can preserve the temporal fine structure of a low-frequency tone by phase-locking its response to the stimulus. Apart from sound localization, however, much about the role of this temporal information for signal processing in the brain remains unknown. Through psychoacoustic...
pubmed.ncbi.nlm.nih.gov
Correct. You'd have to get several of each model from different vendors to figure QA/QC, which is why even sufficient measurements on one unit only take you so far. We already have to usually cross-reference multiple reviewers' various measurements to get a fairly good breadth of measurements, and even then, there's variation from unit to unit even if you assume you've got enough data. Since frequency response is usually the only measurement all the measuring-reviewers do, that's often the only way to check for unit-to-unit consistency, but that still leaves the possibility of set-up differences. Early units are sometimes made with closer tolerances, while later units are more relaxed in quality as quantity is ramped up. One of the big manufactures knowingly had this issue comparing lots over 25 years ago to more recent ones, which led us to recommending they do versions made to closer tolerances, which they did, including signed measurements for each unit at a premium price (and extra letters on the model name).
As for the Truthear Gate I just got likely deviating from what people measured on other examples of that model, this is in comparison to eight other cheap IEMs I just got. From cross-checked measurements of multiple reviewers, encompassing many different measurements, and some duplicates like frequency response, all the IEMs I just got should have been reasonably good. The Gate, however, is a substantial deviation from the others. Everything from 7hz, by the way, had very good channel matching.
Last edited:
