• 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!

May Occam's Razor Apply? Can we learn as much from a $199 EARS as from a $41,000 HATS?

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
Or 90% as much? 50% as much? Nothing similar at all? Or?

I look at these things and see two microphones mounted 180-degrees facing outward surrounded by a frame of 'somewhat squooshy stuff' for placing the headphones.

Then add maybe $300 - $700 in hardware. A bunch of processing software. With enough known in the Public Domain to replicate it - approximate it.

What else?

Interested in opinions..any and all..:) Reasons why - really appreciated!
Hopefully some from members who have used one or both.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,250
Likes
17,182
Location
Riverview FL
@amirm:

Do you still have the miniDSP setup with which to make a comparison with the HATS?
 
OP
Cahudson42

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
Going slightly deeper..Occam occurred because I was wondering if all the hoops HATS goes thru to model pinnae and ear canal really makes any difference in the comparative 'listenability' of the headphones evaluated and compared..

That is, if a headphone 'measures flat' with a simple device like EARS, our pinnae and ear canal will effect all those flat headphones similarly. So why bother to try to measure it?

EDIT:
Well, maybe not get exactly the same listenability for each flat measuring headphone... Given each flat headphone likely has different physical drivers and clamping and positioning, you would expect the 'room' (the very small room of your pinnae and ear canal) to interact a bit differently.. requiring slightly different 'room EQ' to compensate
 
Last edited:

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
16,023
Likes
36,359
Location
The Neitherlands
That is, if a headphone 'measures flat' with a simple device like EARS, our pinnae and ear canal will effect all those flat headphones similarly

Measurements show that this is not the case. It also is completely dependent on the used compensation when you want to have a clear relation with how it sounds.
It is true that the raw measurements of fairly similar HATS at least have a good overall correlation. They only differ in narrow bands.
The EARS lacking an earcanal of course is not having good correlation when looking at raw measurements.
When both EARS and HATS are compensated, corrected for the changes they make to the emitted sound of the headphone, one could get results that can be close together on a lot, but certainly not all, headphones.

The supple and more real pinnae and the stiff ones from EARS present their own issues with over-ears having a shallow pad depth or on-ears.
It is a matter of figuring out the correct 'correction' in both cases. SBAF has been at this for the EARS and, IMO are doing a better job than miniDSP themselves.
 

Sgt. Ear Ache

Major Contributor
Joined
Jun 18, 2019
Messages
1,895
Likes
4,162
Location
Winnipeg Canada
Going slightly deeper..Occam occurred because I was wondering if all the hoops HATS goes thru to model pinnae and ear canal really makes any difference in the comparative 'listenability' of the headphones evaluated and compared..

That is, if a headphone 'measures flat' with a simple device like EARS, our pinnae and ear canal will effect all those flat headphones similarly. So why bother to try to measure it?

EDIT:
Well, maybe not get exactly the same listenability for each flat measuring headphone... Given each flat headphone likely has different physical drivers and clamping and positioning, you would expect the 'room' (the very small room of your pinnae and ear canal) to interact a bit differently.. requiring slightly different 'room EQ' to compensate

I agree with you. I don't get it either. I don't see how there's any benefit to measuring headphones through a fake ear canal unless it happens to be a perfect match for my own ear canal...and also includes the "room EQ" my brain applies to every sound that feeds through my ears...
 
OP
Cahudson42

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
Hmmm. Wonder if we would get more consistent and more comparable (even though different) results between headphones if we simply sawed off the EARS fake pinnea?
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
16,023
Likes
36,359
Location
The Neitherlands
It will mostly differ in the 1-6kH range once you have both 'versions' corrected.
Flatplates also need corrections. Most DIY flatplates just shove a mic in a piece of board and expect the outcome to be the reality.
The EARS could be a bit more correct once properly compensated.

Of course the mic/coupler and the pre-amp for it will never reach the same quality as high$ gear can reach.

Ofcourse, if you really want to measure beyond what others do you would also need a completely silent chamber and perhaps also measure sound leakage and sound isolation as well. Which also needs to be calibrated and corrected for.
Let's just say that if one wants state of the art measurements, only a HATS isn't going to cut it.

The EARS, once properly compensated, can be used for frequency response measurements, looking for substantial (alarming) distortion numbers and to check what modifications for headphones do.
 
OP
Cahudson42

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
The EARS, once properly compensated, can be used for frequency response measurements, looking for substantial (alarming) distortion numbers and to check what modifications for headphones do.
Sounds pretty good to me for $199. ! Do we really need more at ASR? Is there a real need to have ASR measurements comparable to HATS and industry standards?

I think not.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
16,023
Likes
36,359
Location
The Neitherlands
Knowing Amir bought an AP555 (also screamingly expensive) to get more meaningful measurements compared to his old analyzer I don't think he'll consider a soundcard or something like that.

Likewise with speaker measurements he went for a Klippel which is a logical choice.

I would expect Amir to only work with professional equipment which measurements he would trust without a shadow of a doubt and using standardized corrections. The EARS is not that and reckon why he tried the latest and greatest to see how far he can get that.
So Amir will go for scientific grade or nothing. I can certainly get that.

The EARS belongs in the hands of youtube reviewers, hobbyists etc.
The upside of these experiments is that a cheap DIY rig, that is completely wrong in so many ways, still isn't that far 'off' compared to the HATS Amir is testing and is good enough to modify and EQ on. That, however, is not Amir's goal.

I guess he is only interested in official measurements using calibrated and compliant equipment, do measurements, give a short impression and move on. There will be people like Jaakko that will use the measurements to have a program calculate EQ on it and that's as far as Amir will probably take it. Don't know if that extra workload, and money spent is worth the costs of a HATS + deadroom and other essential stuff.
Considering there are plenty of headphone websites around, like Rtings, that already do this. The Delta between all the different HATS is not inspiring confidence for me.
 

Rusty Shackleford

Active Member
Joined
May 16, 2018
Messages
255
Likes
550
Measurements show that this is not the case. It also is completely dependent on the used compensation when you want to have a clear relation with how it sounds.
It is true that the raw measurements of fairly similar HATS at least have a good overall correlation. They only differ in narrow bands.
The EARS lacking an earcanal of course is not having good correlation when looking at raw measurements.
When both EARS and HATS are compensated, corrected for the changes they make to the emitted sound of the headphone, one could get results that can be close together on a lot, but certainly not all, headphones.

The supple and more real pinnae and the stiff ones from EARS present their own issues with over-ears having a shallow pad depth or on-ears.
It is a matter of figuring out the correct 'correction' in both cases. SBAF has been at this for the EARS and, IMO are doing a better job than miniDSP themselves.

Yes, with careful calibration, it’s definitely possible to get FR results from EARS that match much more expensive rigs.

Careful placement, multiple measurements, and averaging, as even Tyll did with his expensive InnerFidelity rig, is also crucial. I think the Butterworth article is much too pessimistic about EARs in that regard. Getting distortion and CSDs from EARS is a bit more up in the air.

Regardless of the rig, though, I think Butterworth is right that basically any measurement is useless on its own because of variance between rigs, compensation, etc. One needs graphs that compare a headphone to known competitors measured on the same rig with the same compensation to provide context.
 
OP
Cahudson42

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
basically any measurement is useless on its own because of variance between rigs, compensation, etc.

Ok, there is variation between various expensive HATS systems. But how much variation between multiple instances of an EARS, and say a $325 HA-DSP? (Plus REW). A simple <$1000 system, but one easily replicated for multiple user comparisons.

I'm reminded that some 'gifted users' can provide exceptional results from the same simple tool. For you woodworkers out there - think Chris Becksvoorts dovetails - with simple chisels.

My point is, given use by 'gifted users', a simple EARS/HA-DSP $1000-type system can produce excellent comparable results.

IMO, we are already seeing that at @solderdude diyaudioheaven. And we would likely see same here at ASR after enough experience, just as is happening with the speaker testing methodology.
 

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
Yes, with careful calibration, it’s definitely possible to get FR results from EARS that match much more expensive rigs.
The calibration required to approximate a frequency response on a HATS - or a human - differs for different headphones, however. This has been separately documented by myself and other independent users, and is the basic problem with any system that doesn't approximate an actual human head (or at least ear).

If people here are really interested I can try to find time to compare 10-15 headphones on my 4128 and EARS sometime this week to highlight the effect I'm describing.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
16,023
Likes
36,359
Location
The Neitherlands
That would be interesting but would like to see the following:

Measure, for instance a HD650, with 4128. Then apply the correction you normally use, let's say Harman,.
Then measure the HD650 with the ears. Apply a custom correction (so not the supplied one or SBAF one) that makes the corrected plot similar in corrected response as the HD650 on the 4128 with correction.
Then, using that custom correction for EARS compare a few other headphones. There will of course be differences still but the question (my question) is how close can these possibly get.
Don't know how feasible this is.

I don't feel showing differences between raw or even compensated plots from those devices is educational as it is clear raw and corrected will measure very different.

The question could be how close can cheap and expensive be when the playing field (the corrected plots) is leveled before the test.
A precondition could be (but realize is a bigger problem in EARS) is that the measured headphones have good seal and the pinna doesn't prevent a good seal.
 

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
That would be interesting but would like to see the following:

Measure, for instance a HD650, with 4128. Then apply the correction you normally use, let's say Harman,.
Then measure the HD650 with the ears. Apply a custom correction (so not the supplied one or SBAF one) that makes the corrected plot similar in corrected response as the HD650 on the 4128 with correction.
Then, using that custom correction for EARS compare a few other headphones. There will of course be differences still but the question (my question) is how close can these possibly get.
Don't know how feasible this is.

I don't feel showing differences between raw or even compensated plots from those devices is educational as it is clear raw and corrected will measure very different.

The question could be how close can cheap and expensive be when the playing field (the corrected plots) is leveled before the test.
A precondition could be (but realize is a bigger problem in EARS) is that the measured headphones have good seal and the pinna doesn't prevent a good seal.
Bear in mind, the comparisons I have done in past are the variation between the delta of a given headphone on system 1 and system 2. This is agnostic to compensation, so long as the compensation would be constant between both systems; it is a measure of how much variation there is in the difference between the systems across a range of different headphones, which is itself the determinant of how reliable a compensation to "turn one into the other" can be.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
16,023
Likes
36,359
Location
The Neitherlands
Yes, the delta between the measurements would yield the same results, didn't think of that one and is easier to do. :)
 

JeffS7444

Major Contributor
Forum Donor
Joined
Jul 21, 2019
Messages
2,365
Likes
3,551
If Amir were to generate data for my specific model of headphone using HATS, that would be great, provided that there wasn't much variation between his set of headphones and my own. Otherwise, EARS it is because I already own it, and the difference between doing nothing and using EARS + HA-DSP has been one of my bigger sonic bargains, IMO.
 
OP
Cahudson42

Cahudson42

Major Contributor
Joined
Sep 21, 2019
Messages
1,083
Likes
1,557
If people here are really interested I can try to find time to compare 10-15 headphones on my 4128 and EARS sometime this week to highlight the effect I'm describing.

Comparisons within grouping different types - on ear, over ear, in ear - planar, similar size drivers, whatever.. most interesting.

Apologize for stating the obvious..:)
 

Rusty Shackleford

Active Member
Joined
May 16, 2018
Messages
255
Likes
550
The calibration required to approximate a frequency response on a HATS - or a human - differs for different headphones, however. This has been separately documented by myself and other independent users, and is the basic problem with any system that doesn't approximate an actual human head (or at least ear).

If people here are really interested I can try to find time to compare 10-15 headphones on my 4128 and EARS sometime this week to highlight the effect I'm describing.

I think this is mainly an issue with IEMs, which the Reddit link seems to support. For over-ear headphones, the EARS seems to do quite well.

I created a compensation curve for my EARS using a handful of common open-back headphones with the most consistent quality control, including the HD650. I compared my measurements to Tyll's from InnerFidelity, then essentially created a compensation that was the average of the differences between the EARS' and Tyll's measurements. After applying the compensation, I went on to measure another 10 or 15 pairs of over-ear headphones, and the EARS measurements continued to match Tyll's quite well without needing to further tweak the compensation curve.

Now, there're obviously going to be discrepancies between any measurement rig. But I think the within-rig variation for most headphones is larger than the variation between EARS and pricier rigs. Things like seal, placement, etc. matter so much that, with the exception of a few almost foolproof headphones, one can easily get two results from the same rig that are many orders of magnitude larger than the difference between a well-done EARS measurement and a well-done GRAS, for example, measurement.

IMHO, the proper metrics for judging the EARS is price to performance. On that count, I think it's hard to fault it. Can it do everything as well as a rig that costs as much as a new car? Of course not. But for the price, it's great.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,496
Doesn't matter to me. So for example, when I used LCD2C's, and use Oratory's EQ, they're like completely new headphones. Like not even remotely close to stock, and are so much better.

I use his HD58X or 6XX EQ, and it's alright (I'm not saying it's bad, but the change isn't as big, nor is the sound much at all as pleasing compared to the LCD, but still very pleasing nonetheless).

Now obviously this virtually worthless anecdote means nothing since you can't really compare headphones in this manner from my perspective (pads, drivers, dimensions, everything is completely different, and who knows what's going on to the sound before it reaches my mind for processing). But these days I generally prefer to EQ on my own (I start with a FR sweep, and try to smooth out any stark peeks by doing multiple passes until things sound okay, and then I come back another day and do the same just to be sure my mind isn't adapting or tuning things out after multiple passes).

As an on-looker to this whole recent headphone measurements ordeal, I'm just somewhat shocked (and somewhat not consider its audio, and the amount of half-assed nonsense that passes through this industry), that there isn't a far more strict validation means of objective data gathering and processing. I again stick to the idea in my head, that most headphones are slapped together with all these unknowns, and they're simply measuring with whatever they've got using basically just FR, and a few subjective listens just to make sure nothing is insanely bad (who knows, since everything in this industry is so hush, no one has a clue on what anyone's doing unless you're an insider I suppose).

With all the basic glance-level reading I've been doing, I still need someone to convince me there's some actual science going on behind the scenes of successful headphone production other than this somewhat-fudging and proceeding to release simply due to the industry and consumers knowing no better anyway..

I still take it to be the case, that even basic approximations of frequency response are good enough, and paired with THD metrics, that's all you really need to tailor your phones to your preference. It's seemingly the case that manufacturers aren't moving much in this realm, so might as well do it yourself in the same way you would tailor your room treatment for speakers or something. Provided again, that THD isn't too high where EQ would just ruin it. You can see active speakers somewhat realizing this, and going this route, trying to fix issues with DSP, because seemingly native transducer characteristics can't be dialed in by whoever is producing them for the majority of audio device makers.

One area that leaves me speechless is Custom-IEMs. If all this mucking around with headphones, speakers, and universal IEM's is bad due to whole sorts of pinna and canal aspects. I cannot imagine A SINGLE manufacturer somehow being able to produce 2 custom IEM's for two different customers, and holding to the idea they're actually producing the same sound as one another. I would LOVE to see how custom IEM's are treated in the development phase (I imagine just getting manufacturing of shells down, and just basic FR sweeps to make sure nothing deviates from the "house sound" is good enough). Can one even produce proper customer IEM's faithfully without being provided HRTF data from a customer (as inaccessible as that is?)
 
Top Bottom