[goodyfresh]

Keep in mind that changing the amplitude of a frequency doesn’t change the pitch or timbre of a sound... EQ can help with intensity and balance, but it can’t fix everything if you don’t start with a good foundation

Keep up the good work exploring and sharing tips with others!

Of course, frequency-response is only one small part of the whole picture when it comes to the sound of headphones or speakers! As I mentioned in a previous post, I use both the HD560s and the X2HR's EQ'd to exactly the same target-curve using exactly the same method from exactly the same website that used measurements of the two headphones done on exactly the same rig. Meaning that with EQ I give both headphones a frequency-response curve that is literally IDENTICAL except for a bit of wiggliness in the highest treble. And yet, there are still very, very audible differences between both headphones (stuff like the timbre as you mentioned, soundstage, imaging, resolution, "fun" vs "analytical," etc.) with both EQ'd to exactly the same FR curve like that.

My point here was that in the particular case of testing with the sounds of things like rain and ocean waves, it has become extremely audibly apparent to me that with both the X2HR and the HD560s, the issues that do exist with how natural they sound LARGELY come from "intensity and balance" due to frequency-response as you mentioned. With both headphones, rain hitting a glass window sounds a BIT too much like sleet rather than real-life rain, with too much white-noise-like background to it, but using an EQ for the treble-peaks (and to even out the middle-to-upper mids but that is a FAR bigger concern with the X2HR than the HD560s) makes rain hitting a window sound much, much more natural and closer to real life!

But then beyond that, things like timbre do indeed come into play, and probably also stuff relating to the differences between the time-domain responses of the two headphones (in measurements you can see this through the X2HR and HD560s having different impulse-responses, leading edges to square-waves, and most significantly CSD Waterfall and Spectrum plots) play a role. With both EQ'd to exactly the same frequency-response, the timbral and time-domain differences seem to produce a result where ocean-waves sound SLIGHTLY more natural/real on the X2HR than the HD560s, while rain sounds slightly more natural on the HD560s. Although both are capable, in my experience, of reproducing the sounds of nature very, very closely to real life when EQ'd to a flat tonality like I do with them; with either headphone, I can close my eyes while listening to ocean or rain sounds and feel almost like I'm really there, which is one of the great things about open-back headphones at this level or above. The bigger differences between the two headphones become much more noticeable with material like music.

 

[MayaTlab]

Of course, frequency-response is only one small part of the whole picture when it comes to the sound of headphones or speakers! As I mentioned in a previous post, I use both the HD560s and the X2HR's EQ'd to exactly the same target-curve using exactly the same method from exactly the same website that used measurements of the two headphones done on exactly the same rig. Meaning that with EQ I give both headphones a frequency-response curve that is literally IDENTICAL except for a bit of wiggliness in the highest treble. And yet, there are still very, very audible differences between both headphones (stuff like the timbre as you mentioned, soundstage, imaging, resolution, "fun" vs "analytical," etc.) with both EQ'd to exactly the same FR curve like that.

Oratory1990 has commented on that : https://old.reddit.com/r/oratory199...beats_solo_pro_is_the_best_headphone/fpay3b5/
Personally when using Oratory1990's presets with two headphones I rarely end up with the exact same sound either, but this could be explained by frequency response at my own eardrum anyway. All the more so if the underlying measurements didn't correspond to my subjective impressions (his Bose 700 measurements for example).
Remaining differences in FR at my own eardrum are sometimes quite easy to spot by various means, at least their location across the FR range (the magnitude I find a lot more difficult to assess by ear).

Also, this article may be evidence that FR at the listener's eardrum may vary quite enough because of anatomical variations that it makes measurements less relevant in certain FR ranges for some people : https://www.aes.org/e-lib/browse.cfm?elib=16877

Which is why I never really use any of AutoEQ's or Oratory1990's presets as is, and often will introduce quite significant deviations from them (particularly at the two extremes of the FR range). For the HD560S the PEQ I'm applying is a work in progress but I've already quite significantly departed from Oratory1990's preset in the bass (I only need two bands instead of three, I don't recognise the wiggles he measured, my own subjective impressions align closer to Rtings) or in the trebles (as is the case with most headphones, I find measurements above 5000hz or so highly unreliable).

It's also why I rarely like to EQ headphones which FR is messed up to start with, I'm not confident enough in properly using high Q alterations and not that confident in measurements accurately portraying such high Q peaks or dips, particularly in the upper mids / trebles.

But then beyond that, things like timbre do indeed come into play, and probably also stuff relating to the differences between the time-domain responses of the two headphones (in measurements you can see this through the X2HR and HD560s having different impulse-responses, leading edges to square-waves, and most significantly CSD Waterfall and Spectrum plots) play a role. With both EQ'd to exactly the same frequency-response, the timbral and time-domain differences seem to produce a result

https://www.reddit.com/r/headphones/comments/aw45df/just_discovered_harman_target_curve/

At least with passive headphones fed via cabled DACs / amps, frequency response might actually be all that matters.

I'm not qualified to comment on that but from my own experience owning various headphones / DAC / amps at various price point, and never really finding anything north of the HD600 to be a straight up upgrade but rather a side-step, and finally "breaking" through that barrier when I started to use PEQ more seriously, at least count me in the camp that considers that whether FR is the only thing that matters or not, it's by far, far, far the most important thing, and that a lot of people are missing the forest for the trees when they're wasting time and energy fiddling with cables or switching DACs or using 16/44 vs 24/96 instead of taking the time to start learning how to properly use PEQs .

Talking of EQ, I'm really quite enjoying how fairly easy it is to EQ the HD560S's sub bass frequencies up compared to other truly open headphones. I'm not 100% sure but as it could be predicted looking at measurements alone, I may indeed find it easier than with the HD650. Conversely, while the HD650s have always been for me and my own anatomy / tastes as close it gets to "I don't need to EQ the trebles" and simply apply a simple shelf up or down, the HD560S's default trebles response is something I'd definitely like to fine tune and fiddle with with EQ (albeit less so than most other headphones, it's quite benign to me as is already).
One thing I really like a lot is how insensitive these seem to sealing variations, at least compared to some other headphones I own(ed).

 

[goodyfresh]

Oratory1990 has commented on that : https://old.reddit.com/r/oratory199...beats_solo_pro_is_the_best_headphone/fpay3b5/
Personally when using Oratory1990's presets with two headphones I rarely end up with the exact same sound either, but this could be explained by frequency response at my own eardrum anyway. All the more so if the underlying measurements didn't correspond to my subjective impressions (his Bose 700 measurements for example).
Remaining differences in FR at my own eardrum are sometimes quite easy to spot by various means, at least their location across the FR range (the magnitude I find a lot more difficult to assess by ear).

Also, this article may be evidence that FR at the listener's eardrum may vary quite enough because of anatomical variations that it makes measurements less relevant in certain FR ranges for some people : https://www.aes.org/e-lib/browse.cfm?elib=16877

Which is why I never really use any of AutoEQ's or Oratory1990's presets as is, and often will introduce quite significant deviations from them (particularly at the two extremes of the FR range). For the HD560S the PEQ I'm applying is a work in progress but I've already quite significantly departed from Oratory1990's preset in the bass (I only need two bands instead of three, I don't recognise the wiggles he measured, my own subjective impressions align closer to Rtings) or in the trebles (as is the case with most headphones, I find measurements above 5000hz or so highly unreliable).

It's also why I rarely like to EQ headphones which FR is messed up to start with, I'm not confident enough in properly using high Q alterations and not that confident in measurements accurately portraying such high Q peaks or dips, particularly in the upper mids / trebles.



https://www.reddit.com/r/headphones/comments/aw45df/just_discovered_harman_target_curve/

At least with passive headphones fed via cabled DACs / amps, frequency response might actually be all that matters.

I'm not qualified to comment on that but from my own experience owning various headphones / DAC / amps at various price point, and never really finding anything north of the HD600 to be a straight up upgrade but rather a side-step, and finally "breaking" through that barrier when I started to use PEQ more seriously, at least count me in the camp that considers that whether FR is the only thing that matters or not, it's by far, far, far the most important thing, and that a lot of people are missing the forest for the trees when they're wasting time and energy fiddling with cables or switching DACs or using 16/44 vs 24/96 instead of taking the time to start learning how to properly use PEQs .

Talking of EQ, I'm really quite enjoying how fairly easy it is to EQ the HD560S's sub bass frequencies up compared to other truly open headphones. I'm not 100% sure but as it could be predicted looking at measurements alone, I may indeed find it easier than with the HD650. Conversely, while the HD650s have always been for me and my own anatomy / tastes as close it gets to "I don't need to EQ the trebles" and simply apply a simple shelf up or down, the HD560S's default trebles response is something I'd definitely like to fine tune and fiddle with with EQ (albeit less so than most other headphones, it's quite benign to me as is already).
One thing I really like a lot is how insensitive these seem to sealing variations, at least compared to some other headphones I own(ed).

As for the HD600: That supports what I said about time-domain stuff creating differences. Time-domain decay (as seen in CSD plots) tremendously affects the sound of headphones. The HD600's CSD Waterfall plot is practically IDEAL with even, fast decay. Oratory1990 claims that EVERYTHING about sound comes entirely from FR alone. That claim is nonsense from a mathematical standpoint; things like resolution and soundstage do NOT arise solely from FR. Decay as seen in CSD plots can vary between two systems with the exact same FR regardless of what Oratory claims. It's not like the guy knows everything. He outright admits that he doesn't know the math behind the technical details of every aspect of sound. He's right that impulse-response and frequency-response are essentially the Fourier Transform of each other, but CSD plots capture WAY more time-domain info than impulse-response. CSD plots PROVE that Oratory is wrong: Only their top/back edge is the FR-curve; beyond that, the decay at a given frequency has little relation to initial amplitude. As for the impulse response "just" being the Fourier Transform of the FR: Even that isn't as simple as Oratory says. He claims that when you see an impulse-response for a headphone, it was obtained by taking the transform of the FR. But that's a lie: Just look at Solderdude's measurements on DIY Audio Heaven. The guy measures impulse-response completely separately from FR, using an actual 100-microsecond pulse for impulse response. I don't know if he's ignorant or outright lying, but Oratory claims "The impulse response is purely mathematical - it is not measured by sending an actual impulse through the loudspeaker and recording the output. It is measured by sending a sweep, recording the output and correlating the output against the original sweep signal." LIES. As if people like Solderdude haven't been using ACTUAL impulses for years. If you want the TRUTH, go here: https://diyaudioheaven.wordpress.com/tutorials/how-to-interpret-graphs/squarewave-and-impulse/

I'm speaking about this as someone with a M.S. in Math who took courses up to PhD-level on Fourier Analysis: Reconstructing a wave from superimposed sine-waves (basically a Fourier Series) is only ONE part of sound. Oratory1990 claims that things like time-domain decay come ENTIRELY from frequency-response alone, but he's unequivocally wrong about that. By his logic, what would even be the difference between an open-back headphone and a closed-back aside from FR? Oratory is full of it. By his logic, you could take a $30 pair of no-name headphones from Walmart and with perfect EQ, make them sound just like the HD560s or even Focal Clears or HD800's. Ridiculous. The HD560s don't sound so damn good JUST because of their nearly-truly-flat FR. They sound so amazing for anything under $350 for other important reasons, like interaction with the pinna of the human ear (what creates soundstage and which Oratory claims comes entirely from FR because he has no clue what he's talking about; if you look at Rtings' measurements of pinna-interaction you can see how wrong he is) and how well-damped they are in the time-domain leading to clean decay as seen in CSD plots (which is responsible for their resolution). Once again, if you want the TRUTH rather than Oratory's half-truths and outright nonsense, go to DIY Audio Heaven and read about interpreting measurements, especially the time-domain aspects: https://diyaudioheaven.wordpress.com/tutorials/how-to-interpret-graphs/. For example, about CSD plots, Solderdude explains: "In theory the ‘line’ you see should be closely the same to that of a normal frequency plot. A perfect headphone (and perfect CSD plot program) would look like a sharp falling edge only in the back of the plot (0 ms after the signal stopped) and should look like a sharp drop of a waterfall. In reality this isn’t the case [...] Long resonances usually indicate an issue in that specific part of the frequency range and may mean that the headphone doesn’t resolve very well in that part of the frequency range."

Yes, the HD560s are pretty much completely insensitive to seal-breaches, as are the X2HR's. That's part of why I chose those two headphones, because I wear glasses. Being a glasses-wearer rules out MANY headphones as being viable options, but those two lose NO sub-bass with a seal-breach like mere glasses, so they're perfect for me in that regard. And yes, they respond extremely well to EQ, which is great.

 

[Evshrug]

Glasses wearer + Bald or balding hair = pick— I mean, a very discerning headphone reviewer! At least that’s what I like to tell myself, haha!

 

[VNandor]

Both the impulse response and the system function (also known as the Fourier-Transform or Laplace-Transform of the impulse response) can completely describe a set of systems called Linear Time Invariant systems. This information can be found in any article that discusses the fundamentals of LTI systems theory, even Wikipedia has this.

If headphones were LTI systems, they could be characterized perfectly just by finding the headpones' impulse response or its system function. However, headphones are not perfectly linear systems, even the best ones have a tiny bit of nonlinear distortion. In practice, they aren't time invariant either, if anyone tried to measure the same headphone 10 times, it would measure slightly differently every single time. The scale of innaccuracies that comes from considering headphones as LTI systems are directly proportional to these factors so if a headphone is linear "enough" and time invariant "enough" then the headphone can be characterized fairly well by using the impulse response or its transform just like any any other LTI system.

 

[goodyfresh]

Both the impulse response and the system function (also known as the Fourier-Transform or Laplace-Transform of the impulse response) can completely describe a set of systems called Linear Time Invariant systems. This information can be found in any article that discusses the fundamentals of LTI systems theory, even Wikipedia has this.

If headphones were LTI systems, they could be characterized perfectly just by finding the headpones' impulse response or its system function. However, headphones are not perfectly linear systems, even the best ones have a tiny bit of nonlinear distortion. In practice, they aren't time invariant either, if anyone tried to measure the same headphone 10 times, it would measure slightly differently every single time. The scale of innaccuracies that comes from considering headphones as LTI systems are directly proportional to these factors so if a headphone is linear "enough" and time invariant "enough" then the headphone can be characterized fairly well by using the impulse response or its transform just like any any other LTI system.

Yup but then Oratory1990 claims that the impulse-response is ALWAYS just the Fourier Transform of the frequency-response and tells a blatant lie that nobody even actually measures impulse-responses. Yes, all headphones have some non-linearities, which can be seen as distortion in the impulse-response. And as for time-variance and impulse-response: As we know, treble frequencies are the ones that cause the largest visible variation/effect on impulse-response, but are also the most time-variant aspect of headphone measurements since it's the treble frequencies whose response tends to vary WIDELY across different re-seats/positions. I don't think that ANY headphone is "time invariant enough" in the treble range to be entirely characterized by either impulse-response or by frequency-response alone. You have to look at both separately, as well as a CSD or spectrum plot to see what's happening in the time-domain.

Back to the topic of this thread and the headphone I like to compare it to the most often (since it's the biggest competitor in the price-bracket): The HD560s is ALWAYS more detail-resolving than the X2HR no matter HOW I equalize the FR of either headphone, while the X2HR is ALWAYS more smoothed-out, forgiving, and "fun" (subjective term as I myself tend to find a resolving, unforgiving sound more fun with most material). The reasons for this can be VERY clearly seen by looking at more than just the FR: If you look at Solderdude's measurements on DIY Audio Heaven, you'll notice that no matter how he EQ's/filters the two headphones, the impulse-response for the HD560s has less ringing than that for the X2HR, which has to do with nonlinearities from distortion like you mentioned (it's worth noting that the X2HR's distortion is mostly second-harmonic, so not BAD sounding, but still distortion nonetheless and notably higher than the HD560s' distortion through the entire frequency range).

Meanwhile, if you look at the CSD Waterfall plots, you'll notice that the HD560s has cleaner, faster, more "ideal" (if what you strive for is unforgiving accuracy) time-domain decay than the X2HR, which is largely what results in the X2HR being more smooth (forgiving and "fun") while the HD560s is more unforgiving and resolving.

Now soundstage is a whole different and VERY complex issue I'm not going to go into beyond what I already said about pinna-interaction, and how I noted that while Oratory1990 claims that pinna-interaction is entirely represented in frequency-response, the measurements of headphones on Rtings conclusively prove that he is completely wrong (Rtings are the only site that measures pinna-interaction, and looking at those results it's clear that they have nothing to do with the FR).

Oratory1990 may use a really good measurement rig and compensation for frequency-responses, but the guy is full of crap when it comes to the math behind all of it and knows far less about it than Solderdude does.

The bottom line, and what's relevant for this thread, is this: No amount of EQ for either headphone is going to make the HD560s sound identical to the X2HR or vice-versa, or to the HD600, or anything else for that matter. Claiming that frequency-response is the be-all-and-end-all and that impulse-response is the "same" as FR is a bunch of baloney. No matter what, the HD560s will have a deeper and taller but also narrower stage with more precise imaging than the X2HR, and is going to resolve more detail but as a consequence be much less forgiving than the X2HR of poor recordings and sound less "fun" for those into that kinda thing.

 

[MayaTlab]

Yup but then Oratory1990 claims that the impulse-response is ALWAYS just the Fourier Transform of the frequency-response and tells a blatant lie that nobody even actually measures impulse-responses. Yes, all headphones have some non-linearities, which can be seen as distortion in the impulse-response. And as for time-variance and impulse-response: As we know, treble frequencies are the ones that cause the largest visible variation/effect on impulse-response, but are also the most time-variant aspect of headphone measurements since it's the treble frequencies whose response tends to vary WIDELY across different re-seats/positions. I don't think that ANY headphone is "time invariant enough" in the treble range to be entirely characterized by either impulse-response or by frequency-response alone. You have to look at both separately, as well as a CSD or spectrum plot to see what's happening in the time-domain.

The above paragraph seems very confused to me.
When DIYaudio modifies the FR curve of the HD560S with a filter, his own "impulse" measurements change as well.

If there is another cause for SQ differences than frequency response it means that there are situations where we can control for the FR curve variable to be invariant and witness differences in other variables.

As long as several variables interact with each others we can't come to any conclusion (and some would say that it's just because in terms of headphones measurements some of them are inherently intricate with each others, but I'll leave more qualified people to comment).

Back to the topic of this thread and the headphone I like to compare it to the most often (since it's the biggest competitor in the price-bracket): The HD560s is ALWAYS more detail-resolving than the X2HR no matter HOW I equalize the FR of either headphone, while the X2HR is ALWAYS more smoothed-out, forgiving, and "fun" (subjective term as I myself tend to find a resolving, unforgiving sound more fun with most material).

That may simply be because they still don't reach the exact same FR curve at your eardrum. It is simply impossible to EQ headphones to reach the same FR for everyone. It's already hard enough to do on a test rig, probably harder still on your own ears.
EQ headphones A and B to reach the same FR on test rig X, put them on test rig Y, and you'll get two more or less different curves. EQ them to reach the same FR curve on Y, and you'll get two different curves on X.

Same applies to humans. I find this paper quite interesting :
https://www.aes.org/e-lib/browse.cfm?elib=16877

There isn't quite enough variability to make EQ presets irrelevant, far from it, but they can only get you to a certain ballpark. Just like you noticed, if I EQ two headphones with Oratory1990's presets, they'll still sound different to me (albeit less so if they diverged a lot before). Improving further their "matching" by ear is entirely possible (I do it all the time and sometimes manage it to a point where I'm not so sure I'd be able to reliably A/B them if it weren't for their physical format), but I wouldn't presume that I have the capacity to reach the exact same FR at my own eardrums, and neither should you think that you do have it.

And as even fairly small dB variations, even at somewhat high Q, can be A/B blind by humans, reaching the exact same FR curve matters a lot if you really want to control the FR curve variable.

As far as I'm concerned with the HD560S, the left plot below is Oratory1990's preset, the right one where I'm at ATM (it will continue to change further) so that they sound "right" to me, which in general means that I can translate fairly well from my pair of near-fields + Dirac to various headphones (in general probably tuned somewhere in the region of Harman's target but possibly not quite match it exactly, I have no way of knowing for sure). This isn't the point anyway, the point is that the difference between both is fairly benign in terms of dBs, the general "gist" of it is similar, and the overall tonality at the "macro" level the same, and yet I'm very confident that I could A/B them any day of the week. If the EQ you've applied to the HD560S and X2HR leave small differences of this kind at your eardrum, that might just be enough for you to find differences in your listening impressions.

The reasons for this can be VERY clearly seen by looking at more than just the FR: If you look at Solderdude's measurements on DIY Audio Heaven, you'll notice that no matter how he EQ's/filters the two headphones, the impulse-response for the HD560s has less ringing than that for the X2HR, which has to do with nonlinearities from distortion like you mentioned (it's worth noting that the X2HR's distortion is mostly second-harmonic, so not BAD sounding, but still distortion nonetheless and notably higher than the HD560s' distortion through the entire frequency range).

Meanwhile, if you look at the CSD Waterfall plots, you'll notice that the HD560s has cleaner, faster, more "ideal" (if what you strive for is unforgiving accuracy) time-domain decay than the X2HR, which is largely what results in the X2HR being more smooth (forgiving and "fun") while the HD560s is more unforgiving and resolving.

As long as you don't control variables you simply can't attribute the cause of what you're hearing to any one of them.

The X2HR's THD measures below the threshold THD is audible for the most part so that isn't it (there are countless articles on AES on the audibility of THD).

BTW, quoting DIYaudio's X2HR review since you like Solderdude's work :
"small but short lived resonance is seen at 5.5kHz (also visible in the distortion plot).
The ringing that is there is very short lived and well damped. I don’t think this has any audible consequences."
"The 40Hz square-wave shows there is roll-off in the lower frequencies but not severe.
The 440Hz squarewave is excellent and almost exemplary. A near perfect square wave indicates excellent and realistic mids. The ringing is short lived and low in amplitude.
The measured signal closely follows the applied signal.
The impulse shows excellent impulse response. The measured signal reaches the target level indicating excellent impulse response. In the horizontal part the 5.5kHz resonance is visible."
https://diyaudioheaven.wordpress.com/headphones/measurements/brands-philips/fidelio-x2hr/
He certainly doesn't seem to be quite as concerned as you in regards to the X2HR's measurements in these two aspects. I have a feeling that you're over-interpreting his graphs.

Now soundstage is a whole different and VERY complex issue I'm not going to go into beyond what I already said about pinna-interaction, and how I noted that while Oratory1990 claims that pinna-interaction is entirely represented in frequency-response, the measurements of headphones on Rtings conclusively prove that he is completely wrong (Rtings are the only site that measures pinna-interaction, and looking at those results it's clear that they have nothing to do with the FR).

Rtings's PRTF measurement is based on frequency response differential, so ??? I don't understand your point. Rtings' approach may be interesting but would maybe benefit from more extensive research before being considered a paragon of "soundstage" (whatever that means with headphones) evaluation.

The entire talk about soundstage on a pair of headphones where only one aspect of how humans localise sound cues (FR curve) is present while the others aren't (inter-aural time difference for example) is completely perplexing to me anyway.

 

[goodyfresh]

The above paragraph seems very confused to me.
When DIYaudio modifies the FR curve of the HD560S with a filter, his own "impulse" measurements change as well.

If there is another cause for SQ differences than frequency response it means that there are situations where we can control for the FR curve variable to be invariant and witness differences in other variables.

As long as several variables interact with each others we can't come to any conclusion (and some would say that it's just because in terms of headphones measurements some of them are inherently intricate with each others, but I'll leave more qualified people to comment).



That may simply be because they still don't reach the exact same FR curve at your eardrum. It is simply impossible to EQ headphones to reach the same FR for everyone. It's already hard enough to do on a test rig, probably harder still on your own ears.
EQ headphones A and B to reach the same FR on test rig X, put them on test rig Y, and you'll get two more or less different curves. EQ them to reach the same FR curve on Y, and you'll get two different curves on X.

Same applies to humans. I find this paper quite interesting :
https://www.aes.org/e-lib/browse.cfm?elib=16877


There isn't quite enough variability to make EQ presets irrelevant, far from it, but they can only get you to a certain ballpark. Just like you noticed, if I EQ two headphones with Oratory1990's presets, they'll still sound different to me (albeit less so if they diverged a lot before). Improving further their "matching" by ear is entirely possible (I do it all the time and sometimes manage it to a point where I'm not so sure I'd be able to reliably A/B them if it weren't for their physical format), but I wouldn't presume that I have the capacity to reach the exact same FR at my own eardrums, and neither should you think that you do have it.

And as even fairly small dB variations, even at somewhat high Q, can be A/B blind by humans, reaching the exact same FR curve matters a lot if you really want to control the FR curve variable.

As far as I'm concerned with the HD560S, the left plot below is Oratory1990's preset, the right one where I'm at ATM (it will continue to change further) so that they sound "right" to me, which in general means that I can translate fairly well from my pair of near-fields + Dirac to various headphones (in general probably tuned somewhere in the region of Harman's target but possibly not quite match it exactly, I have no way of knowing for sure). This isn't the point anyway, the point is that the difference between both is fairly benign in terms of dBs, the general "gist" of it is similar, and the overall tonality at the "macro" level the same, and yet I'm very confident that I could A/B them any day of the week. If the EQ you've applied to the HD560S and X2HR leave small differences of this kind at your eardrum, that might just be enough for you to find differences in your listening impressions.





As long as you don't control variables you simply can't attribute the cause of what you're hearing to any one of them.

The X2HR's THD measures below the threshold THD is audible for the most part so that isn't it (there are countless articles on AES on the audibility of THD).

BTW, quoting DIYaudio's X2HR review since you like Solderdude's work :
"small but short lived resonance is seen at 5.5kHz (also visible in the distortion plot).
The ringing that is there is very short lived and well damped. I don’t think this has any audible consequences."
"The 40Hz square-wave shows there is roll-off in the lower frequencies but not severe.
The 440Hz squarewave is excellent and almost exemplary. A near perfect square wave indicates excellent and realistic mids. The ringing is short lived and low in amplitude.
The measured signal closely follows the applied signal.
The impulse shows excellent impulse response. The measured signal reaches the target level indicating excellent impulse response. In the horizontal part the 5.5kHz resonance is visible."
https://diyaudioheaven.wordpress.com/headphones/measurements/brands-philips/fidelio-x2hr/
He certainly doesn't seem to be quite as concerned as you in regards to the X2HR's measurements in these two aspects. I have a feeling that you're over-interpreting his graphs.



Rtings's PRTF measurement is based on frequency response differential, so ??? I don't understand your point. Rtings' approach may be interesting but would maybe benefit from more extensive research before being considered a paragon of "soundstage" (whatever that means with headphones) evaluation.

The entire talk about soundstage on a pair of headphones where only one aspect of how humans localise sound cues (FR curve) is present while the others aren't (inter-aural time difference for example) is completely perplexing to me anyway.

So if FR at the eardrum is truly all that matters as you and Oratory1990 seem to be claiming, then how do you explain why you even bother to get $200 headphones or to get open-backs? How do you account for the superiority of open-backs over closed-backs, or the fact that full-sized headphones are superior to IEM's? If FR at the eardrum is really all that matters, then just get the flattest/most-neutral sub-$100, or even sub-$50, closed-backs or in-ears available (just make sure to get ones with THD below the audible threshold) and then EQ them as you've been doing with the HD560s to reach your personal ideal. Surely if you fiddle with the EQ enough, you'll be able to get a cheap $50 closed-back headphone from Walmart to sound as good as an open-back audiophile headphone, RIGHT? Surely if you do the EQ perfectly, you can get a pair of closed-backs or IEM's to have soundstage and imaging as good as the HD560s or even the soundstage of the HD800s, right? RIGHT?

The Beats Solo Pro measure as having non-audible distortion and being surprisingly closer to neutral than one would think (I know, surprising to many people on here), and without many peaks or dips to make them difficult to EQ. Surely if soundstage, imaging, etc., are all just purely the frequency-response as you claim, then everybody should just get the Beats Solo Pro, start from Oratory's EQ settings for them (they respond better to EQ according to his measurements than any other headphone he's measured), and then work from there to tweak the EQ to their personal ideal like you do. Surely if someone does so then the Beats Solo Pro will sound just as good as the HD560s despite being closed-back on-ears and will have the added benefit of ANC, right? Want REALLY good noise-cancellation? Just EQ a pair of Bose QC35ii's or Sony 1000XM4's to your ideal. SURELY those closed-back noise-cancelers will have soundstage and imaging, separation, resolution, detail, clarity, etc., as good as a pair of HD800s or Focal Clears as long as you get *the frequency response at the eardrum* to your ideal, right? RIGHT? Oh, wait...

Do you see where this argument breaks down? MOST headphones that are even HALFWAY decent have below-audible THD. So clearly that's not an issue. So why is it IMPOSSIBLE to get a pair of closed-back $50 Sonys or some IEM's, or a pair of well-measuring ANC closed-backs, to sound as good as some nice open-backs no matter WHAT you do? The only possible explanation is that no, frequency-response at the eardrum is NOT all there is to it, otherwise there would literally be no point to things like getting open-back headphones as opposed to clsoed-back, and in fact everyone should just get ANC's and EQ them. But we all know that it's IMPOSSIBLE to get any pair of closed-back ANC's to resolve as much detail or sound as open and clear as good open-backs, and it's impossible to get closed-backs, let alone IEM's, to have soundstage as good as truly good open-backs (like the HD560s) no matter WHAT you do with EQ; closed-backs and IEM's will always sound more congested and unnaturally crowded than good open-backs regardless of frequency-response. If you're going to try to deny that, then tell me this: Why did you even bother to get some good open-back headphones, then?

Also, Rtings' PRTF measurements are the FR-differential between the headphones measured with and without the pinna of the dummy-head present. This is significant. As you pointed out, the same headphone will measure differently on different rigs, and when you EQ two headphones to sound the same on one rig, they'll be different on another. Okay, so... what does it mean to compare the FR with actual ears to the FR with ears with pinnae and take the difference? You're getting the PRTF for the headphone on that measurement rig, which they then compare to the one from a good pair of loudspeakers. That is WAY more complex than just frequency-response-at-the-eardrum and is why two headphones with almost exactly the same FR-curve on Rtings' rig can have vastly different PRTF measurements on their rig. Now, as to whether this is a good method for measuring soundstage, that is very debatable, but Rtings seems to be the only major site with measurements that has made any effort to measure such things.

 

[MayaTlab]

So if FR at the eardrum is truly all that matters as you and Oratory1990 seem to be claiming, then how do you explain why you even bother to get $200 headphones or to get open-backs? How do you account for the superiority of open-backs over closed-backs, or the fact that full-sized headphones are superior to IEM's? If FR at the eardrum is really all that matters, then just get the flattest/most-neutral sub-$100, or even sub-$50, closed-backs or in-ears available (just make sure to get ones with THD below the audible threshold) and then EQ them as you've been doing with the HD560s to reach your personal ideal. Surely if you fiddle with the EQ enough, you'll be able to get a cheap $50 closed-back headphone from Walmart to sound as good as an open-back audiophile headphone, RIGHT? Surely if you do the EQ perfectly, you can get a pair of closed-backs or IEM's to have soundstage and imaging as good as the HD560s or even the soundstage of the HD800s, right? RIGHT?

The Beats Solo Pro measure as having non-audible distortion and being surprisingly closer to neutral than one would think (I know, surprising to many people on here), and without many peaks or dips to make them difficult to EQ. Surely if soundstage, imaging, etc., are all just purely the frequency-response as you claim, then everybody should just get the Beats Solo Pro, start from Oratory's EQ settings for them (they respond better to EQ according to his measurements than any other headphone he's measured), and then work from there to tweak the EQ to their personal ideal like you do. Surely if someone does so then the Beats Solo Pro will sound just as good as the HD560s despite being closed-back on-ears and will have the added benefit of ANC, right? Want REALLY good noise-cancellation? Just EQ a pair of Bose QC35ii's or Sony 1000XM4's to your ideal. SURELY those closed-back noise-cancelers will have soundstage and imaging, separation, resolution, detail, clarity, etc., as good as a pair of HD800s or Focal Clears as long as you get *the frequency response at the eardrum* to your ideal, right? RIGHT? Oh, wait...

Do you see where this argument breaks down? MOST headphones that are even HALFWAY decent have below-audible THD. So clearly that's not an issue. So why is it IMPOSSIBLE to get a pair of closed-back $50 Sonys or some IEM's, or a pair of well-measuring ANC closed-backs, to sound as good as some nice open-backs no matter WHAT you do? The only possible explanation is that no, frequency-response at the eardrum is NOT all there is to it, otherwise there would literally be no point to things like getting open-back headphones as opposed to clsoed-back, and in fact everyone should just get ANC's and EQ them. But we all know that it's IMPOSSIBLE to get any pair of closed-back ANC's to resolve as much detail or sound as open and clear as good open-backs, and it's impossible to get closed-backs, let alone IEM's, to have soundstage as good as truly good open-backs (like the HD560s) no matter WHAT you do with EQ; closed-backs and IEM's will always sound more congested and unnaturally crowded than good open-backs regardless of frequency-response. If you're going to try to deny that, then tell me this: Why did you even bother to get some good open-back headphones, then?

Also, Rtings' PRTF measurements are the FR-differential between the headphones measured with and without the pinna of the dummy-head present. This is significant. As you pointed out, the same headphone will measure differently on different rigs, and when you EQ two headphones to sound the same on one rig, they'll be different on another. Okay, so... what does it mean to compare the FR with actual ears to the FR with ears with pinnae and take the difference? You're getting the PRTF for the headphone on that measurement rig, which they then compare to the one from a good pair of loudspeakers. That is WAY more complex than just frequency-response-at-the-eardrum and is why two headphones with almost exactly the same FR-curve on Rtings' rig can have vastly different PRTF measurements on their rig. Now, as to whether this is a good method for measuring soundstage, that is very debatable, but Rtings seems to be the only major site with measurements that has made any effort to measure such things.

Take a chill pill.

Passive closed back headphones tend to have less smooth FR curves to start with across the entire FR curve range - including below 1khz, which will make them more difficult to EQ. In addition to that, headphones with a closed front volume (closed backs, "semi" open headphones like most planars or Stax) are a lot more sensitive to seal than open back headphones, which will make them more difficult to EQ and will increase divergences from measurements. And finally on-ear headphones, like the Beats Solo Pro you mentioned, may sound quite different from how they measure because of how your pinnae deforms vs. the test rig, which may make them more difficult to EQ if the resulting FR curve is a lot less smooth than it measures.

It's a lot, lot easier to EQ a pair of headphones that measures as smoothly as a pair of HD650 than, let's say, an AKG K371, particularly when you can presume that this smooth FR curve is very likely to be quite close to what you'll actually experience when mounted on your head.

That being said, active ANC closed backs can have a feedback mechanism that can ensure that they reach a specific FR curve target at your eardrum below a certain frequency (around 1khz in general), and can be designed to avoid all of the FR curve messiness that is quite typical of most closed backs below 1khz. These should be just as easy to EQ in that range as any pair of truly open headphones if the FR curve is designed to be smooth to begin with. As it turns out that's the case of the AirPods Max, which I have, and which I consider the easiest closed back headphones to EQ I've ever used, by a mile.
But then you're also introducing a number of uncontrolled variables into the mix. While wireless audio is IMO far less of a problem than what some people make it out to be, it's difficult to know for sure whether this or that emitter / receiver combo is truly audibly transparent or not. The noise floor of most ANC / BT headphones can be quite elevated. It's unknown whether or not the DSP treatment is fully transparent or not (it probably is more so than most audiophiles would like to admit).

Now as I already wrote, I already find it difficult enough with headphones which already have a very smooth FR curve to begin with, which already are quite invariant to seal or position, and for which my subjective impressions already reasonably match how they measure, to EQ them to match well enough that I would find them hard to distinguish. Running sweeps, playing tones, pink noise, there are still quite audible differences in terms of hills / valleys, sometimes peaks, that I'm not competent enough to EQ by ear on my own. If you add all the above variables into the mix, i.e. seal sensitivity (closed or semi open HPs), very coarse FR curve to begin with (passive closed backs) - requiring a lot more bands and hair-pulling to address, large deviations from how they measure (particularly on-ears), then I already know that I'm just nowhere near competent enough to ever reach the same FR curve at my eardrum by EQing by ear. Simple as that.

As it turns out I prefer the K371 or AirPods Max to quite a few open headphones (and find them easier to EQ than some, particularly the Max below 3-4khz). And there are pretty good reasons I got rid of my HD800, Stax system, and actually most of my "summit-fi" equipment (well, at least as far as headphones are concerned, for speakers that's another matter) and only kept a pair of HD650. I don't know WTH people mean by "clarity", "soundstage", or whatever, as none of these words have ever received widely accepted operational definition, but a pair of EQed HD650 is still what sounds the most accurate to me so far (mostly because they're the easiest for me, on my ears - and it will be different for other people - to get close to ideal above 1khz).

While not that important to the subject matter, there are plain, practically reasons to prefer open backs, such as breathability or lack of self-noise from your body, something you'll sometimes experience with closed back if they're truly well closed.

I would really like to see Rtings furthering their PRTF measurements but for a start they'd need to confirm the assumption they make. They assume that because headphones A on their pinna X excites it in a way that is similar to a speaker that headphones A would excite pinna Y (and by extension your own ears) in the same way. But that isn't confirmed.
It's also quite pointless if the resulting FR curve deviates too much anyway from what your own anatomy would require for you to be able to accurately locate sound cues (which simply won't happen with stereo recordings with headphones anyway but we'll gloss over that). What matters to you is the end result, ie FR curve at the eardrum, not how it got there. If we accept Rtings's assumption underlying their PRTF measurements then to me the HD800 is akin to listening to speakers which FR curve at the listening position is completely out of whack to start with.

 

[Evshrug]

Anybody hear good songs lately? I discovered this, so many varieties of texture!



https://tidal.com/track/160996679

 

[fablestruck]

Any impressions from anyone about pairing the HD 560S with Chord Mojo?

 

[Evshrug]

It certainly has the driving power for the HD 560S, but I’ll let someone who isn’t an industry insider comment on the sound quality/character.

 

[Brave Heart]

I got these yesterday, because I really wanted a set of open-back headphones. I am pleased with them so far. Using them to listen to music via Tidal on my PC.

 

[PhazeCrive]

When using the Fiio K5 Pro connected to my PC, what gain settings are best and what volume should my PC's own volume be?

 

[Ichos]

When using the Fiio K5 Pro connected to my PC, what gain settings are best and what volume should my PC's own volume be?

I use max volume on the PC and adjust from the pot.
Gain depends on the headphones you use.
Best setting is to get very loud at about 3 o'clock.