- Joined
- Oct 25, 2019
- Messages
- 11,111
- Likes
- 14,774
Great! I thought you'd invested a great deal of time digesting the reviews here yet you missed the not infrequent posts linking to the YouTube channel.
Just how much spoon feeding do you need?
-
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!
Master Complaint Thread About Headphone Measurements
- Thread starter amirm
- Start date
Just how much spoon feeding do you need?
- Joined
- Oct 25, 2019
- Messages
- 11,111
- Likes
- 14,774
FIFY
solderdude
Grand Contributor
I know I can't read (you told me so must be true) but can you point me to the line where Amir said: "oh screw him, readers should just read more until they understand me" in those exact words or is it your interpretation ?
Correct ... it only seems that way and most likely not only to you.
- Joined
- Oct 25, 2019
- Messages
- 11,111
- Likes
- 14,774
No, no- apparently I cant read either as quote marks dont necessarily mean its a quote. Apparently.
- Joined
- Sep 21, 2018
- Messages
- 507
- Likes
- 747
At the risk of rubbing multiple people the wrong way, I wish we weren't debating the merits of @amirm 's methodology in reviewing the Celestee, but were instead figuring out how to account for how two headphones with low distortion and EQd to the same requency response target can subjectively sound so different to the same person.
We talk a lot about methodology and objectivism and what not, but to me the essence of science is curiosity, followed by experimentation in order to gain new insights. From the initial review through the ensuing debate, I've not seen much curiosity, which bodes poorly for our ability to learn and advance our understanding in an area that we seem to agree is far from fully understood.
We talk a lot about methodology and objectivism and what not, but to me the essence of science is curiosity, followed by experimentation in order to gain new insights. From the initial review through the ensuing debate, I've not seen much curiosity, which bodes poorly for our ability to learn and advance our understanding in an area that we seem to agree is far from fully understood.
- Joined
- Oct 25, 2019
- Messages
- 11,111
- Likes
- 14,774
I agree. I've said, as have others, that the celestee review raises more questions than it answers and perfectly highlights the challenges in both creating and using these (or any other ) HP reviews. Also highlights the importance of listening to the cans as well as the graphs.
I think that a sound starting point is to assess whether they truly had the same FR on someone’s head or not, which I think likely isn’t the case, albeit I wouldn’t know to which degree.
https://www.head-fi.org/threads/how...guarantee-a-better-sound.958201/post-16405751
Don't worry so much about it man, people on these forums have different levels of understanding and there can often be looseness involved in communication that can muddy the waters. Amir's answer to you here was a good one: https://www.audiosciencereview.com/...bout-headphone-measurements.18451/post-816970Ok, let's review what happened here:
1. In the Focal Celestee thread, I said that it would be unscientific to have great measurements, but conclude with a not recommended based on subjective opinion.
link
2. Helicopter says that headphone measurements only tell 2/3 of the story.
link
3. So I argue, if we can only tell 2/3rd of the story with measurements, and that the conclusion could override that entirely. Then perhaps we shouldn't be reviewing headphones scientifically until we have properly solved the complexities.
link
4. Helicopter says that everyone has their own ears and tastes, so we cannot be so scientific about it.
link
5. I argue that it is contradictory to mix scientific measurements using a target that is aimed at a general consensus, with subjective ears and tastes. (Either we should agree that the target is good, and we stick with the measurements. Or we should agree that the target is not good enough due to subjective ears and tastes, and let this part of the site be non-scientific.)
link
6. Then several members of this community tells me that Amir's reviews are ofcource inaccurate and headphone reviews should be subjective.
link link link
7. I argue that this site was meant to be about science reviews. Emphasizing science and objectivity.
link
8. I describe the specific issue i have, and give recommendations
link link
9. Other members insist that Amir's reviews must be subjective, and cannot be based on science.
link link
10. I explain that the way I interpret Amir's goal with this site is to be as scientific as possible.
link link
11. Other members tell me I should just leave this place, or to ignore the conclusions, or to just essentially shutup and stop complaining
link link link
12. I explain that I have no ill-will towards this place or Amir, and that I'm only trying to give suggestions in hopes that this place could be more scientific, because I thought that was Amir's goals.
link link link
13. People start hand waving me away (I say people because several members liked these posts form Jimbob54)
link link
14. Several members explains to me that my original interpretation of the Focal Celestee review was wrong, and that in fact the measurements was actually poor.
link link
15. I accept that I misinterpreted it.
link
16. I'm accused of being the one that is unscientific, the one who thinks Amir's reviews are not meant to be viewed as scientific, and that I think "we are all different".
link link
You guys have great listening skills, but maybe you could use a little work on your reading skills.
Spend some time here on ASR reading the reviews and maybe reading some snippets from papers....Amir has done a few posts here on ASR outlining the process of creation of the Harman Curve Target as well as some of the problems & inaccuracies that can occur, and over on reddit there's Oratory and I've seen him explain the same points, he's here: https://www.reddit.com/r/oratory1990/
So just be a sponge & try to absorb the right information.....internet is about seperating noise from good info....but it takes a bit of time hanging around here on ASR or Oratory's reddit.
After solderdude urge me to move my say here, here it is:
solderdude claimed on the AKG K240 report thread, that the plot of Fr. above 10kHz is not applicable (as it fluctuates by 20 dB or more and has an enormous drop. So why test and post such plot? go till 10 kHz and that's it.
I also have a bad feeling, it doesn't end there, and there are more "mines" in HP testing. In other words, they are done for me.
I rather read other stuff than this report.
solderdude claimed on the AKG K240 report thread, that the plot of Fr. above 10kHz is not applicable (as it fluctuates by 20 dB or more and has an enormous drop. So why test and post such plot? go till 10 kHz and that's it.
I also have a bad feeling, it doesn't end there, and there are more "mines" in HP testing. In other words, they are done for me.
I rather read other stuff than this report.
solderdude
Grand Contributor
I did not claim 'not applicable'. I mentioned that above 8kHz there is an uncertainty and specifically around 10kHz where measurements are less reliable.
One can still plot above 20kHz and mark the uncertainty area (Oratory1999, Crinacle, Rtings) for instance.
When you want an answer from Amir himself you should include @amirm in your question.
That's what shows it isn't easy to measure headphones 'exact' and the possible variances are not only due to the used pinna which only is certified within a certain bandwidth but also seal, pad wear, clamping force, positioning, unit tolerances, target curves present, Pinna and ear canal shape, driver-ear distance, driver angle, analysis, test types and reporting are technical hurdles and we are not even talking about driver, membrane, porting, and headphone enclosure issues.
Then we want to 'match' results to percieved sound with all its extra 'hurdles' that are too much to mention and you might get an idea how complex this is. MUCH, MUCH, MUCH more complex than a rather simple electrical voltage (and current) vaying over time.
You can choose to ignore headphone measurements or those that make them or you use your brain and try to understand and integrate that knowledge into the published tests.
Last edited:
It is what it is.
I did measurements in my life and know a thing or two about it. Wrote TRD's, ATD's, done ATP, Qual environmental and more.
Designed ATE, built ATE, used ATE (including AP).
That to include MIL STD, for airborne (F-16, APCHI and more), tanks (MERKAVA-IV) and VOA broadcast stations.
A test criteria is coming with a test. If the results exceed the criteria, the test fails. If it is within that criteria, the test pass. This is absolute.
Testing comes with a responsibility. It may disqualify a product, an entire batch or be published on a site like this. Who ever put the test to work,
practiced it and made that decision to qualify (a bad product) or disqualify a perfectly good one, comes with a responsibility, that this site's
results publishing is compromised.
Many tests are made over products, in development or production, to approve that they are good enough to be introduced and sold.
As far as I'm concerned in Audio, a 20 dB drop (or peak) is a failure. It also contradicts some very reputed brands in the HP market.
None would allow such a response in it's product. EQ is not a cure for everything (on this site). Just as Aspirin is not a medicine for every headache.
When you suffer on and on with a headache, Aspirin does not helps, you done take an other one (EQ). You go to a doctor to get checked out.
The same anomaly (of huge spikes) repeating in tests, is a bad sign, that the UUT may not be the cause, but the test fixture / procedure does.
As so, until it is revised thoroughly, it should not be published. As you publish what you want, the option left for me, to read what I want.
A thorough test, proving that it is what it is, or finding the real cause, but get it solved and explained, once and forever, is a must.
I don't see any of that is done.
I use my brain.
Sorry for those who don't, and that includes you.
Last edited:
solderdude
Grand Contributor
In (acoustic) audio a sharp drop is most likely caused by an acoustic null and doesn't have to come from the driver (it can though).
In this case it comes from the measurement gear combined with little/no smoothing so is not a fail on 10kHz.
It only is a fail in your eyes because of a profound lack of understanding these measurements.
You may have measured a lot but you clearly have no experience with acoustic headphone measurements at all.
No need to feel sorry for me.
I don't use that type of test fixture for exactly that reason, nor my brain so I am told, as I am mostly interested in the 6-15kHz range for which these test fixtures can not be used. They are, however, quite useful for other parts of the frequency band.
It's like this guy has never seen acoustic measurements before...
Your assumptions need a proof. A nule, frome it comes from (if not the driver, but the test fixture) and it need to be fixed.
20dB is an amplitude of 100 smaller or larger than the ref. Fr. (usually 1kHz). It is enormous. On a DAC Fr. response a 0.2 is noticed. a -1dB or -3dB
would get EQ or NO recommandation. 20 dB for you, is kind of a pass. Good for you. Not good and NO pass for me.
20dB is an amplitude of 100 smaller or larger than the ref. Fr. (usually 1kHz). It is enormous. On a DAC Fr. response a 0.2 is noticed. a -1dB or -3dB
would get EQ or NO recommandation. 20 dB for you, is kind of a pass. Good for you. Not good and NO pass for me.
The problem is you don't understand the background to acoustic measurements of headphones as well as what happens if you were to put those same headphones on your head & measure them somehow at your eardrum.....you have to understand that in order to not be alarmed by what happens in the measurements above 10kHz. You just need to learn more to put it into perspective. It's fine.
solderdude
Grand Contributor
Where is your proof I am making assumptions ?
- Joined
- Jan 15, 2020
- Messages
- 6,897
- Likes
- 16,900
Here is a nice explanatory article with simulations which shows how such nulls are resulting from reflections
https://www.tonmeister.ca/wordpress/2017/09/11/bo-tech-directivity-and-reflections/
https://www.tonmeister.ca/wordpress/2017/09/11/bo-tech-directivity-and-reflections/
If Amir is doing the testing, what should we call what you are doing?, when posting that the Fr. to rely on, is up to 10kHz, but no more?
Thanks
The article of B&o is about speakers. We gathered here about headphones. Find some tiny differences...
The B&o is explaining response vs. room behavior (reflections and off axis).
Testing is done in a room with isolation and walls reflection is avoided by special sound absorbing material. Repeating the same test,
in a non acoustic room, will have different results.
The same is done with EMI / RFI testing. It is not done in a room in a house, or on a test desk in a work room
It is done in a specific EMI / RFI chamber. So calls the MIL STD 461.
HP seems to have some less restricted testing methods, so we can interpret the results in more than one way. Not all are right.
Similar threads
- Poll
