What Headphones do not need (or scream for) EQ?

[Postlan]

I find the problem of HD800's narrow 6K peak is possibly due to its transducer placement, because my HD600 6K dip is compensated when I move the HD600 forward a little towards the face. Test it yourself with a sine wave generator. So HD800's infamous huge sound space can't be done with its unhealthy 6K resonance, I guess. I choose narrower sound space than 6K peak, like Bob Kats.

 

[ADU]

Also OT but I don't think that the sound power response of a loudspeaker should be the reference but the in room response (either predicted or measured) which for good typical cone loudspeakers has a less steep slope, usually around -0.5 dB till -1.0 dB per octave:

Fwiw, I agree that neutral loudspeakers in a semi-reflective room is the best model for a neutral in-ear response.

If you had a pair of speakers calibrated to a spectrally flat response at the listening position in a semi-reflective room, and were able to measure their response inside of either your own ears or the ears of a head and torso measurement rig, like the HBK 5128 HATS, and use that to compensate the raw in-ear measurements of headphones, then an in-room response curve similar to what you've described above might be a fairly good target for something like that. This is the approach that I believe Robbo99999 has been experimenting with, for measurements made on the GRAS rig.

Approximating this type of neutral in-ear response from a diffuse field measurement requires a larger correction though. And the diffuse sound power response of a neutral loudspeaker seems to be the better model for something like this. So that's why I'm using a steeper slope or average based on this as targets for my DF compensated graphs.

The in-room slope or response of a loudspeaker just does not seem to be enough to make up the difference between a HATS rig's response to a spectrally flat diffuse sound field, and a neutral in-ear response.

 

[thewas]

Fwiw, I agree that neutral loudspeakers in a semi-reflective room is the best model for a neutral in-ear response.

If you had a pair of speakers calibrated to a spectrally flat response at the listening position in a semi-reflective room, and were able to measure their response inside of either your own ears or the ears of a head and torso measurement rig, like the HBK 5128 HATS, and use that to compensate the raw in-ear measurements of headphones, then an in-room response curve similar to what you've described above might be a fairly good target for something like that. This is the approach that I believe Robbo99999 has been experimenting with, for measurements made on the GRAS rig.

Approximating this type of neutral in-ear response from a diffuse field measurement requires a larger correction though. And sound power seems to be the better model for something like this. So that's why I'm using a steeper slope or average based on this as targets for my DF compensated graphs.

The in-room slope or response of a loudspeaker just isn't enough to make up the difference between a HATS rig's response to a spectrally flat diffuse sound field, and a neutral in-ear response imo.

I understand your thinking, the problem is that the diffuse field approach is not good for such an approximation and adding another flawed input (loudspeaker sound power) does not really compensate/correct it. If you have done such a correction for the HD600 I would like though to listen to it.

 

[ADU]

I understand your thinking, the problem is that the diffuse field approach is not good for such an approximation and adding another flawed input (loudspeaker sound power) does not really compensate/correct it. If you have done such a correction for the HD600 I would like though to listen to it.

I'll see what I can do on this, thewas. I haven't run across any measurements made on the HBK 5128 for the HD600 though yet, either by Amir, Jude, The Sound Guys, or Harman.

It looks like there are some pretty good measurements available for the Sennheiser HD280 Pro, 599, 650, 6XX, and 800S though. And probably also some other popular headphones, like the AKG K371 and K702, AudioTechnica M40x and M50x, Beyer DT 700 Pro X, 900 Pro X, and 80 ohm 770 Pro, Shure SRH 840 and 1540, Sony MDR-7506, Focal Utopia, HFM Sundara, Monolith M1070C, and some wireless and NC headphones by Bose, Sony, and Apple(?). And Jude also has HBK 5128 plots for a few other higher-end headphones, like the Audeze CRBN and LCD-5, DCA Stealth and Expanse, Focal Clear Mg, and Meze Empyrean. So I could probably do some DF compensated plots for most of these.

The HD6XX and K371 were going to be the next two on my list though, after the HD650 and HD800S. And I'll probably be adding some other targets based on different kinds of loudspeakers and their SP responses as well.

 

[ADU]

Imo, it should be possible to determine the best target or adjustment curve for diffuse field measurements by working one's way back to it from the in-ear measurements of neutral loudspeakers in a room, raw in-ear measurements of neutral headphones, and also Harman's own in-ear target response curve. And as far as I can tell, sound power (or something pretty close to it) looks like the best model for this so far.

 

[thewas]

I'll see what I can do on this, thewas. I haven't run across any measurements made on the HBK 5128 for the HD600 though yet, either by Amir, Jude, The Sound Guys, or Harman.

It looks like there are some pretty good measurements available for the Sennheiser HD280 Pro, 599, 650, 6XX, and 800S though. And probably also some other popular headphones, like the AKG K371 and K702, AudioTechnica M40x and M50x, Beyer DT 700 Pro X, 900 Pro X, and 80 ohm 770 Pro, Shure SRH 840 and 1540, Sony MDR-7506, Focal Utopia, HFM Sundara, Monolith M1070C, and some wireless and NC headphones by Bose, Sony, and Apple(?). And Jude also has 5128 plots for a few other higher-end headphones, like the Audeze CRBN and LCD-5, DCA Stealth and Expanse, Focal Clear Mg, and Meze Empyrean. So I could probably do some DF compensated plots for most of these.

The HD6XX and K371 were going to be the next two on my list though, after the HD650 and HD800S. And I'll probably be adding some other targets based on different kinds of loudspeakers and their SP responses as well.

The 6XX is very close to 600 with new pads, so you could send that to me.

Imo, it should be possible to determine the best target or adjustment curve for diffuse field measurements by working one's way back to it from the in-ear measurements of neutral loudspeakers in a room, raw in-ear measurements of neutral headphones, and also Harman's own in-ear target response curve. And as far as I can tell, sound power (or something pretty close to it) looks like the best model for this so far.

The problem with both diffuse field and sound power measurements is that they represent all directions and reflections in an equal quantity, which is like listening to a loudspeaker in reverberation chamber or in an large distance in a very reflective room, which is not really the target for good sound reproduction. For high fidelity reproduction approximately 50% of the sound arriving at the ear drum should be direct sound. These reasons and others have made the researchers going away from the DF approach.

 

[ADU]

The 6XX is very close to 600 with new pads, so you could send that to me.

Imo, they are similar, but not really equal headphones. So not sure how well this'll work for you.

Here is the DF compensated graph of the HD 6XX though (with a -1.25 slope in blue). The original raw/uncompensated HBK 5128 plot for this can be found in the Sound Guys guys review here...

Drop x Sennheiser HD 6XX review

Wholesaler Drop is offering a Sennheiser HD 650-like set of headphones: the Sennheiser HD 6XX. Explore the more affordable set here.

www.soundguys.com

This plot shows the difference between the HD6XX with DF comp., and the above -1.25 dB per octave slope...

 

[ADU]

This is the difference between the HD6XX with DF comp. and the average sound power of ten well-extended neutral loudspeakers. The sound power curve is represented by the orange line in this case...

A few peaks (and dips) in the treble are normal on a plot like this btw, especially around resonant frequencies of the ear canal at about 3, 8 and 15 kHz. It looks like there could be a little extra brightness in the mid-treble though.

There also seems to be a slight depression around 5 to 6k. But it does not look as pronounced as on the HD650. And it seems to be less of an issue on the previous graph compared to a slope.

 

[thewas]

Imo, they are similar, but not really equal headphones. So not sure how well this'll work for you.

The main differences are due to the different production year pads as shown by Solderdude in the previous page of this thread.

Here is the DF compensated graph of the HD 6XX though (with a -1.25 slope in blue). The original raw/uncompensated HBK 5128 plot for this can be found in the Sound Guys guys review here...

That shows the problem of this approach, correcting to that target would almost remove 10 dB in the treble region making the already quite dark HD6XX sound even darker, because as said the percentage of the brighter direct sound of a loudspeaker is not considered (only if it was omnidirectional).

 

[ADU]

The main differences are due to the different production year pads as shown by Solderdude in the previous page of this thread.

That shows the problem of this approach, correcting to that target would almost remove 10 dB in the treble region making the already quite dark HD6XX sound even darker, because as said the percentage of the brighter direct sound of a loudspeaker is not considered (only if it was omnidirectional).

I'm not sure how you're getting that from the above graphs, thewas. There are no 10 dB peaks anywhere in the treble on the plots that compare the HD6XX's DF response to a sound power curve, or a similar slope. And certainly not on the basic diffuse field plot in the first graph above. So we are obviously looking at this info from very different perspectives... Or I've just done a completely miserable job of explaining it.

What the above plots suggest to me is that the HD6XX is pretty neutral for an open back dynamic driver headphone.

A good percentage of its frequency response falls within +/-1 dB of the above targets, based on a sound power response. If the measurements and graphs above are correct, then that's pretty good for a headphone with no EQ! And it looks like there are only some minor adjustments on the order of a couple dB here and there that would be needed to bring this into even better compliance.

 

[thewas]

I'm not sure how you're getting that from the above graphs, thewas. There are no 10 dB peaks anywhere in the treble on the plots that compare the HD6XX's DF response to a sound power curve, or a similar slope. And certainly not on the basic diffuse field plot in the first graph above. So we are obviously looking at this info from very different perspectives... Or I've just done a completely miserable job of explaining it.

I am talking about this peak, isn't it almost 10 dB?

What the above plots suggest to me is that the HD6XX is a pretty neutral for an open dynamic driver headphone. A good percentage of its frequency response falls within +/-1 dB of the above targets, based on a sound power response.

That is definitely the case, although I see a better matching to the other established targets.

 

[ADU]

I am talking about this peak, isn't it almost 10 dB?
View attachment 238597

Nope. The light horizontal grid lines on the graph represent intervals of 1 dB. And the darker lines are 5 dB intervals. Guess I should have explained that earlier.

Fwiw, the circled peak is only about 2 or 2.5 dB brighter than the other ear canal resonances at 3 and 15 kHz. The bandwidth of the peak looks a bit too wide though, which suggests to me that there is probably a small bright spot or resonance somewhere in that ~10k range that might need a little attenuation.

 

[ADU]

If there is any darkness in the 6XX treble, I suspect it is probably in the vicinity of that dip at around 5-6k. The HD650 also seems to have a depression in the same spot, though it looks a bit more pronounced.

A small dip in between the ear canal resonances at 3 and 8k is probably ok though,

 

[Postlan]

If there is any darkness in the 6XX treble, I suspect it is probably in the vicinity of that dip at around 5-6k. The HD650 also seems to have a depression in the same spot, though it looks a bit more pronounced.

A small dip there, in between the ear canal resonances at 3 and 8k, is probably ok though,

My HD600 has the same dip around 4-6K. It would depend on the shape of the ears, I guess. This dip can be acoustically compensated moving the headphones a little bit forward. I currently use HD600 that way. Soundstage moves to forehead, and it's another positive effect.

 

[ADU]

I find the problem of HD800's narrow 6K peak is possibly due to its transducer placement, because my HD600 6K dip is compensated when I move the HD600 forward a little towards the face. Test it yourself with a sine wave generator. So HD800's infamous huge sound space can't be done with its unhealthy 6K resonance, I guess. I choose narrower sound space than 6K peak, like Bob Kats.

My HD600 has the same dip around 4-6K. It would depend on the shape of the ears, I guess. This dip can be acoustically compensated moving the headphones a little bit forward. I currently use HD600 that way. Soundstage moves to forehead, and it's another positive effect.

Interesting. Thank you for posting these additional observations, Postlan.

 

[Postlan]

Interesting. Thank you for posting these additional observations, Postlan.

I'm curious if the other person would hear the same effect moving headphone a little forward, so I wish you would try it some day. Send sine wave sweep, then I think you'll hear the difference.

 

[ADU]

I'm curious if the other person would hear the same effect moving headphone a little forward, so I wish you would try it some day. Send sine wave sweep, then I think you'll hear the difference.

Can't help with this at the moment, because I don't have either of the headphones.

I don't doubt that adjusting the position is changing the sound a bit though. This is why many graphers will sample a headphone's frequency response at several different positions on their measurement rigs, and then computer the average FR from that.

 

[ADU]

Speaking of averaging...

These are by no means definitive, But I averaged the diffuse field compensated plots of the Sennheiser HD800S, HD6XX and HD650 together, and then compared that average to a -1.25 dB per octave slope, and the sound power of 10 well-extended speakers, and this was the result...

The dB intervals are slightly weird on these, because of how they were computed. The darker horizontal grid lines are still 5 dBs apart though (and they are further divided into thirds by the lighter grid lines).

More DF plots are needed for a more definitive result. But I think you can probably already begin to see how well the sound power model is probably going to work, with only these three headphones. It will be interesting to see how these curves evolve though, as even more neutral headphones get added to the sampling!

All three of these headphones are open-backs btw, so that's why they are all coming up rather short in the sub-bass range.

 

[MayaTlab]

My HD600 has the same dip around 4-6K. It would depend on the shape of the ears, I guess.

Also varies with pad compression - and in extenso pad wear and / or pad sample variation and models.
This is varying levels of pad compression with in ear mics on my own head (please don't compare the absolute values with other graphs) :

While the shape of one's ear could play a role, this general trend under pad compression can be reproduced on flat plate measurement rigs which don't include any :

Earpads

home back to headphones published: July-09-2020, updated: Nov-25-2022 Ear-pads or ear-cushions are an important part of the headphone. They are consumables (in most cases) and wear. Pads can be on-…

diyaudioheaven.wordpress.com

The HD650 is among the few headphones that I don't mind too much without EQ.

 

[Postlan]

Also varies with pad compression - and in extenso pad wear and / or pad sample variation and models.
This is varying levels of pad compression with in ear mics on my own head (please don't compare the absolute values with other graphs) :
View attachment 239097
While the shape of one's ear could play a role, this general trend under pad compression can be reproduced on flat plate measurement rigs which don't include any :

Earpads

home back to headphones published: July-09-2020, updated: Nov-25-2022 Ear-pads or ear-cushions are an important part of the headphone. They are consumables (in most cases) and wear. Pads can be on-…

diyaudioheaven.wordpress.com

The HD650 is among the few headphones that I don't mind too much without EQ.

Ha, what a difference. Thank you for the link.