Unless I’m misunderstanding, this is just compensated frequency response. So why not just call it that?
No. This is headphone behavior variation as a function of the head/ears that it’s placed on. So the effect of the head and ears on the headphone can be identified this way.
Think of it like this:
HpTF: The response of a headphone (a typical FR measurement)
HRTF: the response of an ear (in a given condition)
The HpTF will look different when on a different head, however you can only identify the variation when you subtract the HRTF of each head from the response. So it would be compensated/calibrated, but the point here is that you’re identifying the variation in the headphone’s behavior as a function of the effects of a different head and ears - be that positional effects or the effects of a different pinna in close proximity, that sort of thing.
1 Like
Thank you for taking some time to clarify this, Resolve.
I think I understand what you’re referring to above. And maybe it’s just me who’s having some trouble/confusion with this. But when you refer to the above as “non-HRTF related HpTF variation”, my 60+ year old brain wants to check out.
To my more old-school way of thinking about these things, this is just about differences in the compensated frequency response on different heads or rigs. And I think it might help to use less complicated language for things like this, when its possible. Just my 2c, fwiw.
1 Like
Depends what the compensation is. If that compensation is the subtraction of the individual’s (or rig’s) HRTF, then yes. But if it’s something like Harman, or filter-adjusted DF, or any other ‘target’, then… while it’ll show differences, that’s not exactly what I’m referring to with this.
It’s what I’m referring to in this video here: https://youtu.be/Km2DNLbB-6o?si=HQDzRDSwjiGkv2NV
Just to clarify, the reason to designate it as “non-HRTF related”, is because two HpTFs will look different from one another, solely because the sound is being impacted by different anatomical effects. That’s not the difference we’re interested in so much as we’re interested in the effect on the headphone behavior itself in those conditions.
1 Like
Understood, and I agree.
“HRTF” is somewhat vague as well, since there are different kinds of sound fields/sources that can be used for measuring a head’s HRTF, such as diffuse, free, or semi-reflective sound fields. And these sources interact with a head’s geometry and acoustic impedance differently.
Understood. Fwiw, “non-HRTF related” just = HRTF compensated, in my book. But I understand what you are driving at. In my parlance, you’re interested in the differences in compensated response between heads. And also how those differences in compensated response may vary or not be the same for every headphone.
I don’t know if I agree that the differences in compensated response between measurement rigs or HATS really gives us useful insights on the behavior of actual human heads though. A measurement rig is different than an actual head. And the 711 systems are not as accurate at the extremes of the frequency range, or in terms of acoustic impedance as the newer 5128-style rigs.
Yup, this is it.
Yup, this is also true. And it really is a key point that needs to be reinforced. We show HpTF variation between rigs, but those are all aiming at a given standard. They’re trying to be average of some kind, based on a particular dataset. Real human ears will likely vary more substantially, and it’s worth also getting a sense of that.
However with that said, we do see some HpTF variation across the different HATS pinnae, and even that should be an initial eye opener to people - that what this head/ears ‘hears’ isn’t identical to or precisely indicative of what I personally hear, at least with respect to information above 3khz.
And we’re already starting to see some interesting outcomes, especially when it comes to low acoustic Z headphones that are also mostly immune to positional variation effects - like the egg shaped HiFiMAN headphones for example. Like we just see the HpTF results at the DRP, and that could be from any number of factors, like positional variation and acoustic Z differences. But as you remove those other factors, you get closer to understanding the effects of a given pinna in close proximity on a pair of headphones, and how that differs from a different pinna.
The next step is of course to start modeling real human pinnae and getting the HRTFs for these.
1 Like
Understood. Between the 5128 and 711 though, I put more stock in the 5128 measurements for the reasons I stated above, namely more accurate/anthropomorphic ear canal and acoustic Z, and more accurate performance outside of the midrange frequencies. So including the results for both systems in your compensated plots isn’t that helpful to me.
Some of your other users seem to like it though. And I still have some issues with the current pref and DF curves as well. So I’m still mostly relying on the raw 5128 measurements, and usin them with my own compensation at this point.
I would expect to see differences in the raw measurements on the same rig with different pinnae. And also differences in the HRTF compensated measurements, if you’re using the same HRTF data for both pinnae (which is a no-no).
The compensated measurements will be less meaningful and accurate if the different pinnae are not incorporated in the HRTF measurements. And I suspect that most HATS vendors won’t supply that kind of HRTF data with their rigs. Just HRTF data for their standard configuration, coupler, and pinnae.
Perhaps this does not apply to the comparisons that you’re doing though.
Just had one other thought on this. And it is that before the arrival of the Harman target, it would’ve been more or less understood that compensated (or corrected, or calibrated) FR measurements were done with some type of HRTF, such as a free or diffuse field. The other types of compensation that you’re referring to above (Harman, etc.) are still a comparatively new thing.
Some specificity is helpful though, because it wasn’t always clear which type of HRTFs were being used for compensated plots in reviews in the past. 
It was always flat DF. At least since we move to showing them with the preference bounds.
Yeah it would need to be each pinna’s DFHRTF. And the point there is that you’re bound to see variation depending on the headphone as a direct result of a change in the pinna in close proximity to it. Like even in best case scenarios with respect to positional consistency and acoustic Z, we still see meaningful differences in the behavior of headphones, particularly above 3khz.
My position on this is that while there are differences, there are still trends, and so while a rig pinna isn’t a guaranteed proxy for a real human pinna, it’s still worth seeing that behavior. At the same time, it should put limitations on what we glean from measurement data above 3khz.
1 Like
I was mostly referring to some other reviews done in the past by other websites. I think you’ve been fairly forthcoming with your compensation approachs here, which is appreciated. Though I had some initial confusion about the sourcing of your 5128 DF curve, which was quickly cleared up by Listener if memory serves.
As I said above, I would not be surprised to see differences in the raw measurements on the same rig. If you’re seeing a noticable difference after compensation though, using appropriate HRTFs for each pinna on the same rig, that might surprise me a little. I suppose it would depend on how different the pinnae are though. And how well or poorly the headphones couple/interface with them.
Yup and this is precisely what’s fascinating about it. Because we do have different pinnae for the GRAS systems that use the same ear sim. So it points to ear proximity alone potentially being one of the factors responsible for this variation. For most headphones we expect positional variation and acoustic impedance differences to be stronger factors.
1 Like
This is for @resolve or any other sys admin person:
Is there any way for me to stop getting notifications on the Grell thread?
I don’t see a way, and I think I get flagged on everything since I created this thread originally. I have ZERO interest in this thread anymore, as the Grell was a bust for me.
Sorry to be a bummer, but the constant notifications are irritating at this point.
Thank you, and happy listening to everyone!
1 Like
Yes, thank you! I was looking up top, and completely neglected the bottom of the page! 
Another video relevant to the design of the OAE1
2 Likes
This will undoubtedly also be of interest to the Grell fans.
I’ve only listened to the first half of the podcast. But had a few thoughts on the discussion. Most will have to wait.
I don’t agree with Grell though that a (modified) free-field is the best target. That may be a little closer to what an engineer might hear in a heavily damped near-field mixing suite or studio. But that is not the best target for listening imo.
A mixing suite isn’t necessarily designed to represent how music will be consumed in a home, car, or elsewhere. It’s designed to allow a mixer to do his job well.
Imo, the target listening space on the user’s end is what’s really most important in terms of closing Floyd Toole’s circle of confusion.
1 Like
Finished listening to the rest of the above podcast. Great convo with both Axel and Jermo. Certainly, one the best I’ve heard so far.
Axel is the real deal. And I wish him and his company the best. And hope I’ll get a chance to audition some of his new gear at some point. I haven’t heard an HD800S or HE-1, so that’s somethin else to look foward to.
There is one question that I don’t think was asked that I’d be curious about. And it’s about their favorite headphones made by other mfrs. I’d be curious if they’ve listened to the new FiiO’s, for example, and other recent releases by other mfrs. And what they think are some of the best sounding newer headphones on the market (other than the OAE-1).
1 Like
They mentioned Dan Clark like 5 times, more than any other manufacturer I think.
1 Like
DCA has done alot to advance the art and science of passive* headphone design. So their appreciation is certainly understandable. The same goes for Audeze in the planar magnetic space.
There are a few DCA headphones I’d be interested in listening to as well. But I think most are outside of my budget. The Drop+Grell OAE-1 is actually very reasonably priced though. I just don’t think it’s quite what I’m lookin for in the way of design and sound signature at this point. That’s why I’d be interested in their perspectives on other new headphones in more or less the same price range.
(*Removed “dynamic”.)
But DCA makes planar headphones?
1 Like
You are absolutely correct, Evan_H. My apologies. I’m not sure where the confusion on that came from. I’ve edited my last post.