• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Harman target

KehaDNb

Active Member
Joined
May 5, 2022
Messages
223
Likes
48
is there a study or a video that shows proof to your study where actually people come to the same conclusion as Sean olive (Harman target) that you get that target curve in well treated studio room? I can’t find anything online.
 

Doodski

Grand Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
20,745
Likes
20,757
Location
Canada
Ch 20 of Sound Reproduction The Acoustics and Psychoacoustics of Loudspeakers and Rooms, Floyd Toole touches on this a little bit.
 

Doodski

Grand Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
20,745
Likes
20,757
Location
Canada
I'm not so sure you are going to find literature as you describe, "in well treated studio room?"
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
I think the "well treated studio room" is essentially saying that there won't be excessive room modes or other misbehavior that can adversely color the tonality. I would say the Harman curve is there to make headphones sound perceptually neutral or "flat" for the bulk of listeners. Given it was derived imperially, and is based on measurements on a specific type of measurement fixture, it may not be what some will prefer either due to preferences, age, or anatomical differences. But in a nutshell headphones that do comply with the curve sound the same as neutral speakers to me if I ignore the differences in imaging. The speakers will sound more open and natural (due to a larger image and being able to interact with my head and torso), and the headphones will be more granular and detailed. But with respect to tonality there is little difference outside of some minor room modes, which is what one would expect given listeners have been shown to prefer a flat, neutral response to begin with. Due to its empirical nature, I probably wouldn't waste too much brainpower on trying to look for proof it sounds like studio monitors or speakers in an acoustically treated room (which it should to most people). Rather, that is the measured FR needed to get them to sound neutral to most listeners. Just as a flat response anechoically with controlled directivity is what's needed to get speakers to sound neutral to listeners. For most typical speakers and domestic environments, this translates to the roughly 1 dB/octave downward sloping in-room response. Different systems with different measurements and target responses, but with the same goal: Perceptually flat response from ~20 Hz to 20 kHz.
 

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,074
Likes
1,353
Read Toole's book. Don't work with partial or incorrect assumptions.
 

DavidMcRoy

Addicted to Fun and Learning
Joined
Jul 10, 2018
Messages
574
Likes
989
I decided to just "shop" for a curve that I liked. I settled on EBU 3276. It's the least aggressive one at the top. I especially don't the bass head approach. Sue me.
 

Attachments

  • 0DADE854-FC14-40EE-8F76-BF49BD411BFC.jpeg
    0DADE854-FC14-40EE-8F76-BF49BD411BFC.jpeg
    101.4 KB · Views: 236
Last edited:

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
I decided to just "shop" for a curve that I liked. I settled on EBU 3276. It's the least aggressive one at the top. I especially don't the bass head approach. Sue me.
The tilt seen in the Harman target for speakers isn’t sloped entirely for more bass. Much of that comes from the room acoustics and the fact that box speakers become more directional with increasing frequency (up until you hit the listening window). My experience is if I take and use semi-anechoic measurements to EQ a speaker to be flat, and take measurements where one would normally listen, it ends up being a line with a slight downward slope that is roughly 1 dB/octave, give or take.

Regarding the in-room response target for speakers, it’s a topic that’s worthy of its own discussion. I understand intuitively, but there are a number of factors that go into it, and not every system or listening configuration will conform to it. Easy example is when you are using speakers in the near field. The room measurements at the seating position will look substantially different since the direct sound is much more dominant. I would definitely say it’s worth at least having some idea of where the target response comes from (whether it be headphones or speakers) so you have some idea of where it will be valid and when it might not. But the bottom line is the target is not to get X bass or Y treble, but to ensure the system has neutral tonality with the caveat being that the more one deviates from the conditions under which the target was developed, the less valid it will be.
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
I decided to just "shop" for a curve that I liked. I settled on EBU 3276. It's the least aggressive one at the top. I especially don't the bass head approach. Sue me.
Now in the context of the Harman targets for headphones and speakers the deviation seen in the green line you chose would give a roll-off to the treble in both cases. Now if that’s an actual raw measurement from a microphone or some form of DIY headphone fixture than that’s a different story. I wouldn’t want to arbitrarily apply something without some understanding of the underlying principals. That’s defeating the whole purpose of having a target to begin with.
 
Last edited:
OP
K

KehaDNb

Active Member
Joined
May 5, 2022
Messages
223
Likes
48
The tilt seen in the Harman target for speakers isn’t sloped entirely for more bass. Much of that comes from the room acoustics and the fact that box speakers become more directional with increasing frequency (up until you hit the listening window). My experience is if I take and use semi-anechoic measurements to EQ a speaker to be flat, and take measurements where one would normally listen, it ends up being a line with a slight downward slope that is roughly 1 dB/octave, give or take.

Regarding the in-room response target for speakers, it’s a topic that’s worthy of its own discussion. I understand intuitively, but there are a number of factors that go into it, and not every system or listening configuration will conform to it. Easy example is when you are using speakers in the near field. The room measurements at the seating position will look substantially different since the direct sound is much more dominant. I would definitely say it’s worth at least having some idea of where the target response comes from (whether it be headphones or speakers) so you have some idea of where it will be valid and when it might not. But the bottom line is the target is not to get X bass or Y treble, but to ensure the system has neutral tonality with the caveat being that the more one deviates from the conditions under which the target was developed, the less valid it will be.
So u end up with no bass shelf?? I’m near field speaker experiences??
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
So u end up with no bass shelf?? I’m near field speaker experiences??
There is indeed a bass shelf at low frequencies where the main speakers and sub are omnidirectional, then the response tilts down once the baffle dimensions become significant.
 
OP
K

KehaDNb

Active Member
Joined
May 5, 2022
Messages
223
Likes
48
There is indeed a bass shelf at low frequencies where the main speakers and sub are omnidirectional, then the response tilts down once the baffle dimensions become significant.
How does ur curve looked if u took measurements urself
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
The bass boost from the Harman target is also based on preference.
I would suggest to simply use a similar shelf as the one used by Harman and play with the amount of bass lift till it suits your system/taste and all will be fine.
You can do the same thing with the tilt of the treble by the way. It is no shame nor stupid to not follow that target.
Several targets exist and they all can't be the only correct one .
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
How does ur curve looked if u took measurements urself
In the near-field (setup is 2x Polk S55's + HTS10 sub) the measurement looks like what's shown below. The sub and part of the speaker response is essentially flat and shelved, with it then having a rather extreme downward tilt above about 150 Hz. Note the sub has to roll off a bit early since sitting this close the sub is easily located if its allowed to play by itself, so its response rolls off with the main speakers. Its principally there to help keep the distortion down in the main speakers at low frequencies and provide a modest extension to the in-room response. If I recall, the tilt is pretty close to the actual responses of the mid-bass drivers themselves, and quite extreme at only a few feet from the speakers. However if I do A/B comparisons with a set of headphones on the Harman target, the tonality is the same, but with perhaps slightly less bass and warmer overall sound with the speakers (I'm not a fan of the "high-end" sparkly elevated treble, even though some people may like that. It just puts me on-edge).
Nearfield.png

If I compare this to the horizontal dispersion taken from semi-anechoic measurements I did, one can see that the speakers radiation pattern begins to narrow with increasing frequency, hence the tilt. The start around 150 Hz is likely where the speaker starts to deviate from having an omnidirectional pattern. Given the Signature Series is a modular speaker design, Amir's measurements of the ES20 show this more clearly with the solid dark red region beginning to break above around 150-200 Hz. Its bass response will obviously be different, but it has (nearly) the same woofer, the same tweeter, and width and depth of the cabinet and distance between the woofer and tweeter are the same as the S/ES55 speakers with the main difference being the height and the dual woofers with a tapered cross-over.

S55 Horizontal Directivity.png

Polk Signature Elite ES20 Measurements Horizontal Directivity Klippel NFS Bookshelf Speaker.png


Now, if measurements are taken in the far field about 10 feet away, the response has a more conventional downward tilt of around 1.5 dB/octave. Note that this would be different if I actually set it up for the far-field since its explicitly set up for near-field use. For starters the sub's response is all wrong and there is a huge suck-out and room mode messiness above 100 Hz and in the sub-bass region. But, it does show the difference in tilt between the near-field and far-field responses and the fact that it would probably end up being more or less in-line with the Harman in-room target for speakers once I got done setting it up for far-field use instead.
Farfield Measurement.png
But, as other posters have said, a target is just a target. I developed mine by starting with listening tests, and then dovetailing that in with the measurements to refine it. The best way to look at a target is it objectively quantifies what subjectively sounds good, and not the other way around.
 
OP
K

KehaDNb

Active Member
Joined
May 5, 2022
Messages
223
Likes
48
In the near-field (setup is 2x Polk S55's + HTS10 sub) the measurement looks like what's shown below. The sub and part of the speaker response is essentially flat and shelved, with it then having a rather extreme downward tilt above about 150 Hz. Note the sub has to roll off a bit early since sitting this close the sub is easily located if its allowed to play by itself, so its response rolls off with the main speakers. Its principally there to help keep the distortion down in the main speakers at low frequencies and provide a modest extension to the in-room response. If I recall, the tilt is pretty close to the actual responses of the mid-bass drivers themselves, and quite extreme at only a few feet from the speakers. However if I do A/B comparisons with a set of headphones on the Harman target, the tonality is the same, but with perhaps slightly less bass and warmer overall sound with the speakers (I'm not a fan of the "high-end" sparkly elevated treble, even though some people may like that. It just puts me on-edge).
View attachment 241550
If I compare this to the horizontal dispersion taken from semi-anechoic measurements I did, one can see that the speakers radiation pattern begins to narrow with increasing frequency, hence the tilt. The start around 150 Hz is likely where the speaker starts to deviate from having an omnidirectional pattern. Given the Signature Series is a modular speaker design, Amir's measurements of the ES20 show this more clearly with the solid dark red region beginning to break above around 150-200 Hz. Its bass response will obviously be different, but it has (nearly) the same woofer, the same tweeter, and width and depth of the cabinet and distance between the woofer and tweeter are the same as the S/ES55 speakers with the main difference being the height and the dual woofers with a tapered cross-over.

View attachment 241551
View attachment 241552

Now, if measurements are taken in the far field about 10 feet away, the response has a more conventional downward tilt of around 1.5 dB/octave. Note that this would be different if I actually set it up for the far-field since its explicitly set up for near-field use. For starters the sub's response is all wrong and there is a huge suck-out and room mode messiness above 100 Hz and in the sub-bass region. But, it does show the difference in tilt between the near-field and far-field responses and the fact that it would probably end up being more or less in-line with the Harman in-room target for speakers once I got done setting it up for far-field use instead.
View attachment 241553But, as other posters have said, a target is just a target. I developed mine by starting with listening tests, and then dovetailing that in with the measurements to refine it. The best way to look at a target is it objectively quantifies what subjectively sounds good, and not the other way around.
How does ur target look??
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
How does ur target look??
Its basically three piece-wise targets. One for the sub and bass response, then the ~ -4 dB/octave or so tilt for the response from roughly 150 Hz to about 1 kHz, and then above the cross-over point the tweeter then have their own "target". I deliberately omitted that part because it gets confusing since there are other things that have to be taken into account. But basically its three piece-wise "functions" that are joined together as shown schematically below, but with the third one for the tweeters missing. Essentially its manually integrating the drivers together since there is not enough distance for that to happen on its own. I think some of the oddities may also come from the speakers and sub being backed up to the wall. I'm sure there are other details that elude me as well...
Target schematic.png

All of this came from my "idea" to use big home theater speakers as my PC speakers. However the only way to get it to actually sound the way it should was through measurements. I remember when I first set it up my initial response was "Oh man, it sounds pretty bad lol..." Saner people would have gotten some small Genelecs, but I decided to explore measurements instead, which was probably a lot better in the end since I was completely ignorant about how speakers and acoustics actually worked before I got this system and read the usual references like Dr. Floyd Toole's "Sound Reproduction". Here's a pic of it from another post I made here at ASR. The headphones on the side are 2020 HiFiMan Sundaras, FWIW. Edit: As a side-note, it images in ways headphones simply cant. It wouldn't make much sense to have it, otherwise since it takes up so much space. But given what it does, it's well worth the inconvenience.

20220708_190228.jpg
 
Last edited:
OP
K

KehaDNb

Active Member
Joined
May 5, 2022
Messages
223
Likes
48
Its basically three piece-wise targets. One for the sub and bass response, then the ~ -4 dB/octave or so tilt for the response from roughly 150 Hz to about 1 kHz, and then above the cross-over point the tweeter then have their own "target". I deliberately omitted that part because it gets confusing since there are other things that have to be taken into account. But basically its three piece-wise "functions" that are joined together as shown schematically below, but with the third one for the tweeters missing. Essentially its manually integrating the drivers together since there is not enough distance for that to happen on its own. I think some of the oddities may also come from the speakers and sub being backed up to the wall. I'm sure there are other details that elude me as well...
View attachment 241554
All of this came from my "idea" to use big home theater speakers as my PC speakers. However the only way to get it to actually sound the way it should was through measurements. I remember when I first set it up my initial response was "Oh man, it sounds pretty bad lol..." Saner people would have gotten some small Genelecs, but I decided to explore measurements instead, which was probably a lot better in the end since I was completely ignorant about how speakers and acoustics actually worked before I got this system and read the usual references like Dr. Floyd Toole's "Sound Reproduction". Here's a pic of it from another post I made here at ASR. The headphones on the side are 2020 HiFiMan Sundaras, FWIW. Edit: As a side-note, it images in ways headphones simply cant. It wouldn't make much sense to have it, otherwise since it takes up so much space. But given what it does, it's well worth the inconvenience.

View attachment 241556
I mean if u compensate it in a headphone target…
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
I mean if u compensate it in a headphone target…
Oh, ok. It would be the Harman headphone target, but with maybe 1-2 dB less bass, and a hair less treble. It ended up being really close tonality-wise when I did set about doing A/B comparisons afterwards with my headphones and IEMs. Edit: To be a bit more clear, in terms of what's perceived, it would be maybe a slight downward tilt below 100 Hz and above 3-4 kHz, but flat everywhere else.
 

Cars-N-Cans

Addicted to Fun and Learning
Joined
May 19, 2022
Messages
819
Likes
1,009
Location
Dirty Jerzey
Did u measured it with a head?
I don't have a dummy head or a headphone fixture, only set up to do speaker measurements at the moment. I think the closest thing I have headphone/IEM-wise to something that complies sans any EQ is the TRUTHEAR x Crinacle Zero IEM I got for travel. When I got it I compared and the tonality is very, very close to the speakers. I think the most distinctive thing would be some of the messiness in the response between 100-200 Hz in the speakers does throw the tonality off slightly in that region.
 
Top Bottom