EQ'ing to Harman curve doesn't give me pleasing results - why not?

[Gringoaudio1]

Ringing, more specifically audible pre-echo will be introduced with sharp phase shifts below 500Hz. I can quantify sharp as larger than 45 degrees with a Q higher than 1. REW auto-eq generated filters are minimum phase and are very unlikely to generate phase shifts of that magnitude.

I know these limits because I create all possible filters and listen with these filters to find where ringing starts. In fact, ringing can be checked quite efficiently from step response.

I find it hard to understand that many audiophiles are so desperately seeking for post calibration results to even try a method. The most complicated filter takes less than an hour to generate. You cannot damage a speaker with a filter and if you don't like the results you switch your previous filter back on. Besides, flat frequency response is almost the least useful evidence of a well calibrated setup followed by phase response. I can get a FR flat with a PEQ by ear in minutes and it would most likely sound throttled/boxy.

I suggest people with passion for better sound out there to go and test the methods that sounds logical to them. Testing is learning.

What do you think of Equalizer APO software and the filters it creates? I’ve been measuring in REW and exporting the filters it creates into Equalizer APO. Then tons of tweaks and additional parametric filters. Great looking FR curves at my listening position.

 

[OCA]

What do you think of Equalizer APO software and the filters it creates? I’ve been measuring in REW and exporting the filters it creates into Equalizer APO. Then tons of tweaks and additional parametric filters. Great looking FR curves at my listening position.

It's a powerful and free convolution engine for Windows. I used it a lot for testing my convolution filters instantly in the past as it can be activated during REW measurements. The only "minor" problem I had with it was varying system delays it generated at every reset which could be a problem for accurate driver time alignment for DIY purposes.

 

[goat76]

Harman curve was created for headphones not speakers but it works fine in rooms, too and REW auto EQ will yield descent results given:

1. No smoothing (practically 1/48 smoothing needs to be applied to remove notches from measurments) up to the room transition frequency around 200Hz where you listen to the room not the speaker. Up to 5dB boost for dips can also safely be used in that region in my experience.
2. No smoothing (1/48) between 200 and 500Hz (roughly where room effects end) but no dip boost allowed
3. Reflection free response above 500Hz (up to 20kHz) because in that region you correct for the speaker not the room, no dip boost as most dips here will be related to speaker crossover design. Anechoic chamber measurement is not practical for most users so a suitable windowing needs to be applied to the response to remove reflections i.e. if the distance of LP to the speaker is 3m, apply a right window size of 3 / 343 x 1000 = 8.7ms to the response which will only contain direct sound as no reflection can reach the LP faster than the direct sound. In practice, such tight windowing will not show any peaks to be corrected above 1-2kHz although there're exceptions in some cases.

The only thing I want to add to your information is that a 3-meter measuring distance is probably too long for most normal-sized rooms, the thing we want to avoid with a windowed measure is early reflections reaching the microphone earlier than the direct sound from the speaker. What we want is to maximize the measuring window to get the most possible resolution to the gated measuring window.

• A 2 meters distance should in most cases be enough, to at least make sure the sound from the drivers is summed to one "unified sound".
• Raise up the speaker on a stool or something so that the speaker's acoustic center (in most cases the tweeter but can sometimes be the mid-driver or somewhere between the two) is at an equal distance to the floor and ceiling to maximize the distance from those two reflection points.
• Place the speaker as far out in the room as possible to maximize the distance from the rest of the boundaries, and the same for the microphone's position 2 meters away from the speaker and raised to the same height as the speaker's acoustic center.
• Make a normal measuring sweep in REW.
• Choose the "Impulse" window in REW.
• Click on "IR Windows" and in the popup window you should apply a value of something like 0.1-0.2 ms in the "Left Window" column, and after that, you can drag the small "L" in the window to choose the starting point right before the impulse start (and make sure the small "waves" before the big impulse will be part of the measuring window). For the "Right Window", you can apply a value of something between 2-5 ms, and you can move the small "R" right before the first reflection (which you should easily see as the first small peak) and leave that out of the window.
• When looking at the "SPL & Phase" window you now see the windowed measurement, it's of a low resolution but that should be good enough to see if any equalizing is needed for the speaker's direct sound, which is the dominating sound you hear at the listening position according to Toole.

Here is the gated measurement of my speaker:

 

[neRok]

But there was a bit too much volume somewhere about the midrange, I'm thinking the 300-500 range.

Just thought I would update this thread as I have been listening quite a bit over the last ~2 weeks using the EQ derived from MMM, and it has been rather good. I did apply an additional PEQ at 400Hz of -2dB and Q 3.5, which helped tamed that midrange problem (it was "blaring" a bit at loud volumes). I'm not 100% happy with the bass <150Hz, because whilst there is lots of bass all over, some parts aren't as strong as others. So for example I was listening to an album and the "fundamental" bass line in 1 song sounded and felt great, but then another song with a slightly different note did not give the same feeling. Similarily not all drum beats are as impactful as others (only some have the proper good feeling I know is possible), so that's probably related. I wonder if it's because the speakers are too far away, compared to how I've had them in the past (used to have 6.5" drivers on desk, and those did give the feeling). I haven't integrated the sub properly, so that might be part of the problem. But also I'm dabbling around with "virtual bass arrays" at the moment (a current thread here: https://www.audiosciencereview.com/...d-attenuated-bass-signals.46426/#post-1656525), to see if there is something that can be done there.

If one speaker is appearing louder or the phantom image has shifted then you probably have to lower or raise the gain on one or the other. That has to do with the way the reflections are playing off the two sidewalls. Probably best to lower gain.

I think you were on to something here. I have been moving and rotating the speakers by very minor amounts and have found that it does cause a difference in certain frequency bands. I'm guessing it is because of my right hand speaker and how it interacts with the window recess it is next to, as in the slightly different reflections are causing different gains (+ and -). Because during my tests I have always found the vocals are quite central, and that it's just certain symbols or guitar strums or effects like piano notes that are different "stereo widths".

Using EQ_APO I set up a toggle so I could switch the stereo tracks (and I did set up the channel switch before my individual channel EQ's, presuming it is smart enough). This way I could test during playback if it was a stereo sound that sounded off-center, or an actual off-center (panned) sound. I did find on occasion that when I channel swapped a song with an actual panned sound, the "stereo location" of the sound didn't mirror perfectly, so I guess I still have work to do.

 

[ExPerfectionist]

Regarding all of the differences you notice in bass, drums, etc. from song to song, that's probably in the mixing and the goal of artistic taste of the artist and sound engineer / mixing. One song may have a really pronounced bass line or clear and dynamic drums, but the next song even from the same artist may have a different mix and aural landscape. That's totally normal as every song is going to sound different.

Yes it's possibly you also have some peaks and nulls in your frequency response, which could create some of those differences as well. The easiest way to check is to measure the in-room response at your LP (or 2/several locations where your ears would be or around where you sit). You can EQ manually (like with a DSP device like MiniDSP) or use an auto-EQ like Audyssey or DIRAC built into AVRs/receivers.

With your other observations, yes room treatment can help make the acoustics more symmetrical for the left and right sides. For example if one speaker has a wall to the outside side of it, and the opposite side of the room has no wall and is open to another room or area, you're going to have reflections off the wall on the one side and not the other. Treating that wall's reflection point with absorption, heavy drapes over the window, etc. can help even things out so it's not so lopsided, and is more consistent.

Room acoustics, distance from front and side walls, and toe-angle towards/away from listening position definitely all have a big impact on bass, soundstage, imaging, and overall sound quality. As you're noticing as you move the speakers a few inches here and there.

The Four Secrets of Speaker Placement - The Absolute Sound

The Four Secrets of Speaker Placement -

www.theabsolutesound.com

How to position your speakers for great sound

Location, location, location. No matter what audio equipment you have, proper speaker placement can significantly improve your listening experience. Here are 7 quick tips to help you get the most out of your stereo speakers: Initial placement. Decide roughly where you will be positioned when...

uturnaudio.com

Speaker Placement Guide: Get The Best Sound From Your Stereo - ELAC

Regardless of what audio equipment you have, optimal Speaker Placement will enhance your listening experience.

www.elac.com

 

[dasdoing]

So I think Harman is basically realy just a loudness compensation

check out this comparison:

ignore the stuff beyond 10kHz. it looks extreme but this isn't really the frequency range people perceive as "bright". this sits between 5kHz and 8Khz or so.
in the bass range we perceive that either the 75vs90 or 70vs90 should give better results as it wont mask the 100-300 range as Harman does.
believe it or not, I am now room EQuing to FLAT var/psychoacoustic in REW. with loud listening I use 80vs90 compensation. no compensation would be club loudness...very loud. stuff sounds very balanced

 

[DVDdoug]

So I think Harman is basically realy just a loudness compensation

I don't think so... That would mean that the loudness standard for the Harmon curve is quieter than whatever the mixing & mastering engineers are working at. The engineers & producers are trying to produce something that's enjoyable for the "average listener" and they don't expect people to be listening at unusually-loud levels.

 

[sigbergaudio]

So I think Harman is basically realy just a loudness compensation

Only that it isn't a dialed in curve, it's the natural response of a selection of mostly anechoically flat speakers in typical rooms.

 

[GXAlan]

Only that it isn't a dialed in curve, it's the natural response of a selection of mostly anechoically flat speakers in typical rooms.

I wonder if typical rooms from the era of Harman’s research and typical rooms of new homes today are different. Dimensions and ceiling sizes and even furniture. We also see more hardwood floors than carpet in the US west coast now.

 

[OCA]

Only that it isn't a dialed in curve, it's the natural response of a selection of mostly anechoically flat speakers in typical rooms.

Except for that it was originally developed for headphones!

 

[NTK]

I wonder if typical rooms from the era of Harman’s research and typical rooms of new homes today are different. Dimensions and ceiling sizes and even furniture. We also see more hardwood floors than carpet in the US west coast now.

Th estimated in-room response curve (EIR) is a completely synthetic curve -- a weighted sum of each of the frequency response curves at the 70 measurement points. No room will have frequency independent reflections (wall absorptions are frequency dependent), which the EIR does not consider.

 

[sigbergaudio]

Except for that it was originally developed for headphones!

I think you are confusing two different things.

 

[GXAlan]

Th estimated in-room response curve (EIR) is a completely synthetic curve -- a weighted sum of each of the frequency response curves at the 70 measurement points. No room will have frequency independent reflections (wall absorptions are frequency dependent), which the EIR does not consider.

But the EIR wasn’t the Harman speaker preference curve though? Not to mention

JBL Synthesis SDP-75 Press Release

www.jblsynthesis.com

“…newly-developed room correction target curve for improved performance.”

 

[Ruspamen]

Harman curve was created for headphones not speakers but it works fine in rooms, too and REW auto EQ will yield descent results given:

1. No smoothing (practically 1/48 smoothing needs to be applied to remove notches from measurments) up to the room transition frequency around 200Hz where you listen to the room not the speaker. Up to 5dB boost for dips can also safely be used in that region in my experience.
2. No smoothing (1/48) between 200 and 500Hz (roughly where room effects end) but no dip boost allowed
3. Reflection free response above 500Hz (up to 20kHz) because in that region you correct for the speaker not the room, no dip boost as most dips here will be related to speaker crossover design. Anechoic chamber measurement is not practical for most users so a suitable windowing needs to be applied to the response to remove reflections i.e. if the distance of LP to the speaker is 3m, apply a right window size of 3 / 343 x 1000 = 8.7ms to the response which will only contain direct sound as no reflection can reach the LP faster than the direct sound. In practice, such tight windowing will not show any peaks to be corrected above 1-2kHz although there're exceptions in some cases.

So for the eq below 200hz, no leveling needs to be performed? or do you have to set it to 1/48 before the eq?

 

[DualTriode]

Harman curve was created for headphones not speakers but it works fine in rooms, too and REW auto EQ will yield descent results given:

1. No smoothing (practically 1/48 smoothing needs to be applied to remove notches from measurments) up to the room transition frequency around 200Hz where you listen to the room not the speaker. Up to 5dB boost for dips can also safely be used in that region in my experience.
2. No smoothing (1/48) between 200 and 500Hz (roughly where room effects end) but no dip boost allowed
3. Reflection free response above 500Hz (up to 20kHz) because in that region you correct for the speaker not the room, no dip boost as most dips here will be related to speaker crossover design. Anechoic chamber measurement is not practical for most users so a suitable windowing needs to be applied to the response to remove reflections i.e. if the distance of LP to the speaker is 3m, apply a right window size of 3 / 343 x 1000 = 8.7ms to the response which will only contain direct sound as no reflection can reach the LP faster than the direct sound. In practice, such tight windowing will not show any peaks to be corrected above 1-2kHz although there're exceptions in some cases.

Hello,

This is evidence of not knowing and or not understanding.

In the beginning Harmon started with a "Standard" listening room with speakers equalized to be flat in an anechoic space.

Then Harmon gave listeners headphones and a device the listeners used to equalized the headphones to sound like the speakers in the "standard" listening room.

Then Harmon drew a line or curve that best represented the statically determined preference of the sample of listeners.

That statically determined curve is the Harmon curve.

The Harmon Curve was not created it was measured.

Thanks DT

 

[GXAlan]

That statically determined curve is the Harmon curve.

There is the Sean Olive developed Harman curve for headphones and the Floyd Toole developed Harman curve for loudspeakers.

Speakers

Headphones

The Perception and Measurement of Headphone Sound Quality- What Do Listeners Prefer? - Sean E. Olive

acousticstoday.org

 

[OCA]

There is the Sean Olive developed Harman curve for headphones and the Floyd Toole developed Harman curve for loudspeakers.

We can actually ask Sean himself, he's an active forum member

Sean Olive

www.audiosciencereview.com

 

[OCA]

So for the eq below 200hz, no leveling needs to be performed? or do you have to set it to 1/48 before the eq?

REW's "Match target level to response" button does a good job matching the response level with the uploaded target curve. The calculated level is usually a tad bit higher with unsmoothed response. I then use that exact SPL level to auto EQ 1/48 smoothed response below Schroeder's.

 

[OCA]

a device the listeners used to equalized the headphones to sound like the speakers

Thanks for the heads up (I really didn't know if that's the case) but that seems like an absurd way of testing hearing preferences to me. Harman should give us all from these magic devices and we can stop fiddling with sophisticated algorithms/measurement mics to equalize audio

 

[GXAlan]

We can actually ask Sean himself, he's an active forum member

Sean Olive

www.audiosciencereview.com

As is Floyd Toole.

But the point is that Harman Curve used colloquially in internet chatter can mean the target curve above, also seen in any of the downloadable Dirac targets from Arcam or JBL Synthesis, and the target curve that Dr Olive has published which has the odd looking bumps (but is what you want for a neutral sounding headphone).