Is there any way to objectively measure headphone resolution?

[Feelas]

The FR/IR will predict the system's complete response to a stimulus only if the system in question is linear and time-invariant ('LTI' system). If there are non-linearities (which exist in most real systems to some extent), those are unfortunately not described by the IR/FR.

Oh, yes, I forgot the LTI assumption, thank you! I've seem THD+N measured headphones for years, yet is IMD taken into THD+N (I bet, the "+N" part if yes)?

 

[chi2]

These can be described by a system's transfer function / curve (here's a nice visual example of that).

Only problem with the visualization is phase inversion .

 

[dominikz]

Oh, yes, I forgot the LTI assumption, thank you! I've seem THD+N measured headphones for years, yet is IMD taken into THD+N (I bet, the "+N" part if yes)?

THD(+N) you typically measure with only one tone as stimulus. IMD(+N) you measure normally with two tones as stimulus. So not really possible to read one from the other directly - though both are caused by the same non-linearity of the transfer function, so they are indeed linked. It is just that this non-linearity causes non-harmonic components (IMD) as well as harmonic ones (THD) when there are multiple tones on the input.

Only problem with the visualization is phase inversion .

Hmm... seems to me phase is OK - positive half-wave on the input corresponds to positive half-wave on the output. Though I'll admit it's been a very long time since I last looked at it so I may be looking at it wrong

 

[100rounddrum]

I guess same can be said about mostly any target curve if you account for what Dreyfus compiled for us. It's closer than both FF and DF targets, I think. I'd just not agree about too big bass boost, since getting the character in-line with room curves is actually a nice treat. The headphone crowd is just too accustomed to thin sound.

IMO I haven't yet found albums which translate badly onto K371s, and that means (I guess) that I can be hardly taken as a non-average person

The B&K target, is closer to perceptual neutral than the Harman curve, or any other curve.

The FF and DF target curves are indeed too thin and bright, but even they are closer to perceptual neutral than the Harman curve, since they peak at the upper-treble, instead of the sensitive upper-mids area.

The Harman curve is a rollercoaster of peaks and dips.

 

[dresler]

The B&K target, is closer to perceptual neutral than the Harman curve, or any other curve.

The FF and DF target curves are indeed too thin and bright, but even they are closer to perceptual neutral than the Harman curve, since they peak at the upper-treble, instead of the sensitive upper-mids area.

The Harman curve is a rollercoaster of peaks and dips.

I'd love to see your sources on this.

 

[Feelas]

The B&K target, is closer to perceptual neutral than the Harman curve, or any other curve.

Yet, because the Harman target is based on a "natural, speaker-like FR in a good room", you could possibly argue that it's B&K plus the averaged HRTF with a bit of a bump here and there, yet it's also averaged over the preferences of a group of 70+ people. I believe if you'd be to substract the HRTF part from Harman, you'd get B&K back. By the way, B&K is actually hyped too and doesn't hightlight anything else than a common room response, I don't think it was ever regarded as a "perceptual neutral" - this is not quite right. FF & DF don't ever happen in natural rooms.

 

[100rounddrum]

Yet, because the Harman target is based on a "natural, speaker-like FR in a good room", you could possibly argue that it's B&K plus the averaged HRTF with a bit of a bump here and there, yet it's also averaged over the preferences of a group of 70+ people. I believe if you'd be to substract the HRTF part from Harman, you'd get B&K back. By the way, B&K is actually hyped too and doesn't hightlight anything else than a common room response, I don't think it was ever regarded as a "perceptual neutral" - this is not quite right. FF & DF don't ever happen in natural rooms.

If you think the Harman curve is close to neutral or enjoyable, you must have hearing damage in the upper-mids. Honestly, I can’t think of another reason, since we humans are already sensitive to the 3 - 5khz range. Even a measured neutral response there, will equate to a slighly boosted perception of those frequencies, let alone if that area is boosted several DB’s.

Coupled with a dip in the lower-mids, it makes for a distant, harsh sound.

Maybe those 70 people had hearing damage, since it’s very common in the American youth to have hearing loss? So if you boost the upper-mids, you compensate for that hearing loss, thus it would sound neutral.

 

[Feelas]

If you think the Harman curve is close to neutral or enjoyable, you must have hearing damage in the upper-mids. Honestly, I can’t think of another reason, since we humans are already sensitive to the 3 - 5khz range. Even a measured neutral response there, will equate to a slighly boosted perception of those frequencies, let alone if that area is boosted several DB’s.

Coupled with a dip in the lower-mids, it makes for a distant, harsh sound.

Maybe those 70 people had hearing damage, since it’s very common in the American youth to have hearing loss? So if you boost the upper-mids, you compensate for that hearing loss at normal a slightly b, thus it would sound neutral.

See papers above.

 

[Dreyfus]

The upper mid boost is probably there as a part of the PRTF / HRTF, representing the frontal localization in Harman's listening room.
A lot of Oratory's settings are a bit too hot in the 1-4 kHz region for my ears. Probably due to a mix of personal taste and a non-average pinna and canal.

I like the B&K systems for their better interpretation of the human ear's anatomy. Also the coupler ("ear impedance simulator" to be exact) is technically much better than the old 711 and its enhanced iterations. But then again, the best it can serve is only some kind of average.

Comming back to Harman, have they ever done any tests comparing their averaged target to the in-situ response of participants?
To my knowledge they did only some in-ear measurements to verify that their GRAS rig was on point capturing the in-room speaker response. Correct me if I'm wrong here.

I would also like to see some data proving that B&K is better at recreating tonal balance in practice.

Regards
Dreyfus

 

[Feelas]

If you think the Harman curve is close to neutral or enjoyable, you must have hearing damage in the upper-mids. Honestly, I can’t think of another reason, since we humans are already sensitive to the 3 - 5khz range. Even a measured neutral response there, will equate to a slighly boosted perception of those frequencies, let alone if that area is boosted several DB’s.

And sources on that?

This is FAR from not peaking in the 3kHz - 5kHz range - neither FF, nor DF, in fact it's completely the opposite. Please elaborate on that?

See the old Tyll's article. Tyll's HMSII ear simulator had almost -10dB compensation between 3-5kHz to account for the specifics of the head.

I would also like to see some data proving that B&K is better at recreating tonal balance in practice.

Sonarworks posted in an article that the HF attenuation should nicely even up with room reflectivity.
B&K 1974 target - originated probably here - is mentioned as merely accounting for getting the FR properly into the shape of near-field and far-field response? It makes an assumption, that the microphone catches both reflected and direct sound w/ flat (or near) manner - which is a realistic one to be done. Thus we'd need to consider rolling off the HF to emulate the realistic balance - keep in mind, that it was written in 1974 and regarded a specific scenario, since they mention targeting an "average hall" but with half intensity of rolloff. And yes, the reflected sounds in rooms are an issue, which cannot be accounted for easily. Still, the paper in itself is old and amounts for 1) the materials used to build homes back in the '70s (which means that the curve might not translate to modern rooms) and 2) well, they write explicitly that the point of interest was between 60Hz and 6kHz...

It'd actually be easy to recreate the B&K from Harman (if we consider it perceptibly flat, disregarding the bass boost) - merely treat Harman as flat & overlay that respones to look like B&K. Since the basic curve itself (w/o bass boost) was derived (afair) from speakers EQ'd to measure flat in-room response in a reference room, then overlaying B&K should amount to just overlaying.

Comming back to Harman, have they ever done any tests comparing their averaged target to the in-situ response of participants?
To my knowledge they did only some in-ear measurements to verify that their GRAS rig was on point capturing the in-room speaker response. Correct me if I'm wrong here.

That one is actually a great question! In A Statistical Model that Predicts Listeners’ Preference I see no mention of such comparison. It'd be an interesting one to see whether the applied curve lends itself to being measured as flat in-ear. I don't actually think that they're after that kind of experiement, seeing how far this is into supplying something that just sounds satisfying to mostly anyone.

On ResearchGate I can see Olive et al. paper called "Listener loudspeaker preference ratings obtained in situ match those obtained via a binaural room scanning measurement and playback" - it seems distantly connected.

Seeing that Harman Curve sparks discussion (albeit one really tinged into Reddit-like or even insulting territory, which is saddening, seeing how we'd like to discuss the matters to measure and not whether one is deaf), I'd like to delve into something I've noticed today:

In The Ultimate Paper (TM): Olive et. al. A Statistical Model that Predicts Listeners’ Preference they've used the (by now it seems standard) methodology of EQ-ing the K712 and convoluting other headphones FR's onto the K712s. So, they must've measured K712s on GRAS, then applied a differential EQ between K712s and measured pairs. EQ'd K712s to 30 pairs + one "high anchor" (Harman Curve) and measured preference across listeners.
The idea of virtualizing headphones over each other came from the idea before, based on (their own) papers concerning the blind-testing case on headphones. It's obvious that perception can be skewed by what you feel on head and that might get all the typical bias-train going. To avoid that, you'd have to virtualize what's on head. Also, what was mentioned before: they explicitly state that the match was within +-dB up to 12kHz; so we're actually talking about tonal balance of different pairs, but with K712s 12kHz+. Does that show any red flags? It's interesting, to say the least. Now, that could pretty much tilt the favour into Harman-EQ'd K712s side, because of merely emulating the rest badly. Hmmm....

The virtual headphones only simulated the magnitude and minimum phase part of the headphones and excluded any nonlinear or excess phase distortions that were present in the actual headphones. However, the validation studies together with other headphone investigations suggest that the magnitude response is the dominant factor in how good or bad a headphone sounds.

... and as commented in conclusions: Finally, we wish to address the limitations of this study so that the results are not generalized to conditions outside those tested. This study did not address or simulate non-linear or excess phase distortions. Ourexperiences and others suggest these are not dominant factors in how a headphone sounds, but more research may change our views.
As such, they're aware that they're not accounting for this.

As such, it would be good to know whether EQ-ing into a foreign characteristic doesn't cause new non-linear issues to occur and distortion to happen, thus lowering the preference rating perceived by the listening group -> in the end, the preference rating could be different when faced with a "real" pair, accounting for the non-linear phenomena that the specific FR in a specific enclosure causes. It is not mentioned, however, whether they've checked for distortion happening after EQ-ing.

There's an interesting paper dated 2017/2018: doi: 10.1121/1.5031030 (Perception of nonlinear distortion on emulation of frequency responses of headphones). It seems - please correct me if I'm wrong - that what they've done is pretty convoluted: forcing 11 pairs to emulate HD800's response and afterwards measuring their FR & non-linear distortion. Then, bending HD800's FR to match what was measured on pairs & emulating the measured 2nd & 3rd order distortion. Oof! I don't know whether it seems trustable, but interesting nonetheless. Despite that, it seems that what is lacking is the comparison of starting vs resultant non-linear performance, but that's not what the study was after, it seems. And it's all about detecting 2nd/3rd order distortion by ear, not exactly whether it causes anything bad to happen.

The previous work of Gutierrez also seems interesting: doi:10.3390/app6040117 (Influence of the Quality of Consumer Headphones in the Perception of Spatial Audio). Conclusions are regarding the proper (1dB or more) channel balance, FR uniformity & quality, uneven response between 100Hz & 1.6kHz causing front-back confusion (widely mentioned in opinions of some pairs!) and 4-7kHz causing bad lateral positioning.

It would thus be of interest to get the the oratory1990 EQ god to measure more than merely the FR... There might be a gigantic bag of conclusions to be found there.

 

[Dreyfus]

It'd actually be easy to recreate the B&K from Harman (if we consider it perceptibly flat, disregarding the bass boost) - merely treat Harman as flat & overlay that respones to look like B&K. Since the basic curve itself (w/o bass boost) was derived (afair) from speakers EQ'd to measure flat in-room response in a reference room, then overlaying B&K should amount to just overlaying.

Well, my hope was that there was any comparable reference curve by B&K that addresses the HRTF of human listeners in a good listening room. There are a few curves for certain rooms and azimuths for their HATS. Harman's approach is still different, though, because they include the actual perception of humans.

The old B&K in-room speaker curve is probably out of date if you consider the technical limitations of the time versus the more recent research done by Harman.

I don't actually think that they're after that kind of experiement, seeing how far this is into supplying something that just sounds satisfying to mostly anyone.

That's what I thought, as well. After all, their strongest interest is to find a curve that qualifies as a good average target for product designing.
I still think that diving deeper into the head- and ear-related differences on headphones might reveal some interesting findings that could lay the foundation of tomorrows personal audio market (which definately will be a thing in the near future).

Now, that could pretty much tilt the favour into Harman-EQ'd K712s side, because of merely emulating the rest badly. Hmmm....

Let me quote Sean's blog:

An important questions is: "How accurate are the virtual headphones compared to the actual headphones"? In terms of their linear acoustic performance they are quite similar. Fig. 2 compares the measured frequency response of the actual versus virtualized headphones. The agreement is quite good up to 8-10 kHz above which we didn't aggressively equalize the headphones because of measurement errors and large variations related to headphone positioning both on the coupler and the listeners' head.
[...]
More importantly, "Do the actual and virtual headphones sound similar"? To answer this question we performed a validation experiment where listeners evaluated 6 different headphone using both standard and virtual listening methods Listeners gave both preference and spectral balance ratings in both standard and virtual tests. For headphone preference ratings the correlation between standard and virtual test results was r = 0.85. A correlation of 1 would be perfect but 85% agreement is not bad, and hopefully more accurate than headphone ratings based on sighted evaluations.

Virtual headphone simulation appears to be a controversial subject. There are definately a lot of technical uncertainties. More particularly the higher you go up in frequency ... until you finally hit the limits of the measurement system.
That said, a correlation of 0.85 between the virtualized and the original model is quite good already.

Do you have access to the "A Virtual Headphone Listening Test Methodology" (9658) AES paper? Maybe we can find some more details on that topic.

As for the non-linear distortions, that is really something that should be addressed in future studies, indeed. Sean agrees, as he once mentioned during a presentation.

 

[wasnotwasnotwas]

I'd just not agree about too big bass boost, since getting the character in-line with room curves is actually a nice treat. The headphone crowd is just too accustomed to thin sound.

Definitely agree with being used to less bass- I have started to increase my HP bass more in line with Harman but it takes some getting used to . On some HP the Harman adjustments sound "right"- less so on others. So both Oratory measurements to Harman- HE560 works well, sounds "right"- Focal Clear sounds "wrong".

 

[Feelas]

Definitely agree with being used to less bass- I have started to increase my HP bass more in line with Harman but it takes some getting used to . On some HP the Harman adjustments sound "right"- less so on others. So both Oratory measurements to Harman- HE560 works well, sounds "right"- Focal Clear sounds "wrong".

See that oratory's measurements "cut off" after 10kHz - bringing the sub-10k treble down (leaving 10k+ as it is) instead of shelving all the 10k down to get the same downward pointing response will sound strange - probably a tad of detached high-treble sparkle & shimmer. Looking at diagrams, Clears can suffer from exactly that (innatural) character after EQ.

It'd be a great experiment if you apply 10kHz high-shelf (paired with o1990 EQ) onto Clears and just turn the gain down until you get the correct balance, and then tell us if it was possible to get the tonality right.

Anyways, it should be mentioned that there's no point in forcing the self to use Harman "baseline", especially because they've highlighted high variance between subjects. The bass boost will be lower the bigger room you take into account and the low shelf corner frequency is also a moving target, thus it MUST be modified to preference, else it will result in unnecessary hyping. Pretty much the same about high-shelf, since room reflectivity is different when untreated, that might contribute to the mentioned variance.

Anyways, I see that audio communites in general are far and between in understanding, that bass is the source of fundamentals, which are pretty much the most important component of all music... And instead fuzzling around with "upper treble", "upper mids" - if bass is broken, then nothing sounds well.

 

[Feelas]

Do you have access to the "A Virtual Headphone Listening Test Methodology" (9658) AES paper? Maybe we can find some more details on that topic.

Since I've kinda got it from somewhere far, I might consider sharing it via PM if you'd like and I'd not get into jail that way How I see their process from the paper:
0. All done on GRAS 43A w/ IEC711 coupler
1. Measure the HD518's FR, prepare an inverse FR IIR filter to EQ it flat
2. Measure the 6 target pairs (701/550/LCD2R2/BeatsStudio/QC15/V-Moda Crossfade), prepare FR IIR filter to simulate a pair
3. Measure resultant FR of HD518's on GRAS to check the match
<0.5dB match was reported between FRs of headphone vs virtual headphone up to 10kHz, no filtering above.
Pairs were volume-matched to 78dB B-weighted SPL (according to ITU-R 1770.2 - don't know the standard) when used with the "non-virtualized" method. I believe the same was done for the virtual testing, but it's not mentioned directly. 12 listeners were between 24 and 52 calibrated to Harman testing procedure (it is interesting how that might affect the results). Actually, they haven't used minimum-phase filters here, which is interesting. I think that for HD518's 78dB SPL is not enough to cause perceptible distortion, yet I've not used nor measured them. Perceptual evaluation of individual headphone compensation in binaural synthesis based on non-individual recordings paper is mentioned (thankfully, available for free on ResearchGate) as a further reference to whether minimum-phase matters or not. Currently trying to read it, and it's actually a detailed study on using personalized HpTF's and whether they actually work.

 

[Helicopter]

@solderdude do you think Audeze LCD2 Classic would be OK with EQ? I do not like the idea of wood in the standard version, but was not sure if the missing fazors in the classic matter much.

 

[solderdude]

I have heard a few LCD-2 and 1was 'grainy' in the treble and another one 'smooth'.
also heard one of the original LCD-2 and found it too dark.
The LCD2 Fazor had a bit more treble but still an upper mids suckout, just not as bad.
The LCD2 classic is more efficient and is also a bit dark.
Haven't played around with that one with EQ.
I reckon with EQ they can sound tonally the same. No idea if the 'grainy' sound in one of them is something that is more common and whether or not that will/can improve with EQ.

I found the seal of LCD-2 and weight making it no comfortable enough for me so never owned one for that reason.
Loved the bass and mids, didn't like the lack of upper mids.

 

[100rounddrum]

And sources on that?

This is FAR from not peaking in the 3kHz - 5kHz range - neither FF, nor DF, in fact it's completely the opposite. Please elaborate on that?

See the old Tyll's article. Tyll's HMSII ear simulator had almost -10dB compensation between 3-5kHz to account for the specifics of the head.


Sonarworks posted in an article that the HF attenuation should nicely even up with room reflectivity.
B&K 1974 target - originated probably here - is mentioned as merely accounting for getting the FR properly into the shape of near-field and far-field response? It makes an assumption, that the microphone catches both reflected and direct sound w/ flat (or near) manner - which is a realistic one to be done. Thus we'd need to consider rolling off the HF to emulate the realistic balance - keep in mind, that it was written in 1974 and regarded a specific scenario, since they mention targeting an "average hall" but with half intensity of rolloff. And yes, the reflected sounds in rooms are an issue, which cannot be accounted for easily. Still, the paper in itself is old and amounts for 1) the materials used to build homes back in the '70s (which means that the curve might not translate to modern rooms) and 2) well, they write explicitly that the point of interest was between 60Hz and 6kHz...

It'd actually be easy to recreate the B&K from Harman (if we consider it perceptibly flat, disregarding the bass boost) - merely treat Harman as flat & overlay that respones to look like B&K. Since the basic curve itself (w/o bass boost) was derived (afair) from speakers EQ'd to measure flat in-room response in a reference room, then overlaying B&K should amount to just overlaying.



That one is actually a great question! In A Statistical Model that Predicts Listeners’ Preference I see no mention of such comparison. It'd be an interesting one to see whether the applied curve lends itself to being measured as flat in-ear. I don't actually think that they're after that kind of experiement, seeing how far this is into supplying something that just sounds satisfying to mostly anyone.

On ResearchGate I can see Olive et al. paper called "Listener loudspeaker preference ratings obtained in situ match those obtained via a binaural room scanning measurement and playback" - it seems distantly connected.

Seeing that Harman Curve sparks discussion (albeit one really tinged into Reddit-like or even insulting territory, which is saddening, seeing how we'd like to discuss the matters to measure and not whether one is deaf), I'd like to delve into something I've noticed today:

In The Ultimate Paper (TM): Olive et. al. A Statistical Model that Predicts Listeners’ Preference they've used the (by now it seems standard) methodology of EQ-ing the K712 and convoluting other headphones FR's onto the K712s. So, they must've measured K712s on GRAS, then applied a differential EQ between K712s and measured pairs. EQ'd K712s to 30 pairs + one "high anchor" (Harman Curve) and measured preference across listeners.
The idea of virtualizing headphones over each other came from the idea before, based on (their own) papers concerning the blind-testing case on headphones. It's obvious that perception can be skewed by what you feel on head and that might get all the typical bias-train going. To avoid that, you'd have to virtualize what's on head. Also, what was mentioned before: they explicitly state that the match was within +-dB up to 12kHz; so we're actually talking about tonal balance of different pairs, but with K712s 12kHz+. Does that show any red flags? It's interesting, to say the least. Now, that could pretty much tilt the favour into Harman-EQ'd K712s side, because of merely emulating the rest badly. Hmmm....

The virtual headphones only simulated the magnitude and minimum phase part of the headphones and excluded any nonlinear or excess phase distortions that were present in the actual headphones. However, the validation studies together with other headphone investigations suggest that the magnitude response is the dominant factor in how good or bad a headphone sounds.

... and as commented in conclusions: Finally, we wish to address the limitations of this study so that the results are not generalized to conditions outside those tested. This study did not address or simulate non-linear or excess phase distortions. Ourexperiences and others suggest these are not dominant factors in how a headphone sounds, but more research may change our views.
As such, they're aware that they're not accounting for this.

As such, it would be good to know whether EQ-ing into a foreign characteristic doesn't cause new non-linear issues to occur and distortion to happen, thus lowering the preference rating perceived by the listening group -> in the end, the preference rating could be different when faced with a "real" pair, accounting for the non-linear phenomena that the specific FR in a specific enclosure causes. It is not mentioned, however, whether they've checked for distortion happening after EQ-ing.

There's an interesting paper dated 2017/2018: doi: 10.1121/1.5031030 (Perception of nonlinear distortion on emulation of frequency responses of headphones). It seems - please correct me if I'm wrong - that what they've done is pretty convoluted: forcing 11 pairs to emulate HD800's response and afterwards measuring their FR & non-linear distortion. Then, bending HD800's FR to match what was measured on pairs & emulating the measured 2nd & 3rd order distortion. Oof! I don't know whether it seems trustable, but interesting nonetheless. Despite that, it seems that what is lacking is the comparison of starting vs resultant non-linear performance, but that's not what the study was after, it seems. And it's all about detecting 2nd/3rd order distortion by ear, not exactly whether it causes anything bad to happen.

The previous work of Gutierrez also seems interesting: doi:10.3390/app6040117 (Influence of the Quality of Consumer Headphones in the Perception of Spatial Audio). Conclusions are regarding the proper (1dB or more) channel balance, FR uniformity & quality, uneven response between 100Hz & 1.6kHz causing front-back confusion (widely mentioned in opinions of some pairs!) and 4-7kHz causing bad lateral positioning.

It would thus be of interest to get the the oratory1990 EQ god to measure more than merely the FR... There might be a gigantic bag of conclusions to be found there.

https://www.resonancehearingclinic.com/resonance-of-the-human-ear-canal/

 

[Feelas]

https://www.resonancehearingclinic.com/resonance-of-the-human-ear-canal/

Yes, and? You're actually contradicting yourself, since you bothered to write 3-5kHz and yet now linking that typical ear canal resonance is 2-4kHz. Well... And actually, that'd mean that DF/FF targets should be even higher, because the measurements were probably taken with blocked ear-canal, instead of at eardrum. Even so, it'd be unnecessary to apply the ear resonance, when applying personalized HpTF it should take the ear resonance level into account & not force this bump to happen.

Please don't try if you're not going to bother discussing, merely using half-truths. Actually ear resonance isn't merely 3-5kHz, it's a moving target .I believe I read it somewhere on Sean Olive's blog that there was a study noticing VERY high variance between subjects in ear resonance frequency.

So NO, if you are willing to suggest to anyone that they have broken ears, get your facts straight first, asshole. Either start behaving or I'll merely ignore anything you say. Maybe contribute something else than just some unfirm opinions, which don't get us anything further.

 

[wasnotwasnotwas]

See that oratory's measurements "cut off" after 10kHz - bringing the sub-10k treble down (leaving 10k+ as it is) instead of shelving all the 10k down to get the same downward pointing response will sound strange - probably a tad of detached high-treble sparkle & shimmer. Looking at diagrams, Clears can suffer from exactly that (innatural) character after EQ.

It'd be a great experiment if you apply 10kHz high-shelf (paired with o1990 EQ) onto Clears and just turn the gain down until you get the correct balance, and then tell us if it was possible to get the tonality right.

Anyways, it should be mentioned that there's no point in forcing the self to use Harman "baseline", especially because they've highlighted high variance between subjects. The bass boost will be lower the bigger room you take into account and the low shelf corner frequency is also a moving target, thus it MUST be modified to preference, else it will result in unnecessary hyping. Pretty much the same about high-shelf, since room reflectivity is different when untreated, that might contribute to the mentioned variance.

Anyways, I see that audio communites in general are far and between in understanding, that bass is the source of fundamentals, which are pretty much the most important component of all music... And instead fuzzling around with "upper treble", "upper mids" - if bass is broken, then nothing sounds well.

I'll have a play. And no, I don't force it. If it sounds off I either tweak or go back to stock or ham-fisted eq I did.

That's the beauty of software eq rather than physical mods or ludicrous hardware alchemy. Quick to on/off and infinitely variable.

 

[wrigglycheese]

I just love this place.