• 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 daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

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

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,184
Likes
2,642
Location
Salvador-Bahia-Brasil
based on you meassured response the "toole target" (I think it was extracted from one of his books) looks more suited. The Harman one does boost you bass (realativly) so you will feel loss of mids

attached
 

Attachments

  • Dr Olive Toole Target.txt
    1.4 MB · Views: 59
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
I've started reading some of the in depth info on the REW website, and found this nugget of info on regarding the Impulse Graph.

The default settings for the windows will usually be suitable. In smaller rooms it may be necessary to use a shorter right-side window duration, around 300-500ms - if the frequency response plot appears noisy and jagged try reducing the right window period and hit "Apply Windows" to recalculate the frequency response. In very large rooms the window can be increased to improve the frequency resolution.

I've still got some reading to do, but that looks like an avenue worth exploring. Also I take it that if I wanted to try out the alternative method I have talked about (less late reflections in the final measurement), then that would be achieved using IR Windows. I wonder if it needs a combination of IR Windows and accompanying frequency range filtering? I also had an idea for a "manual" method of doing such a measurement, by attaching a little cardboard shield to the mic. I will see what changes this makes.
 
Last edited:

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,184
Likes
2,642
Location
Salvador-Bahia-Brasil
I've started reading some of the in depth info on the REW website, and found this nugget of info on regarding the Impulse Graph.

The default settings for the windows will usually be suitable. In smaller rooms it may be necessary to use a shorter right-side window duration, around 300-500ms - if the frequency response plot appears noisy and jagged try reducing the right window period and hit "Apply Windows" to recalculate the frequency response. In very large rooms the window can be increased to improve the frequency resolution.

I've still got some reading to do, but that looks like an avenue worth exploring. Also I take it that if I wanted to try out the alternative method I have talked about (less late reflections in the final measurement), then that would be achieved using IR Windows. I wonder if it needs a combination of IR Windows and accompanying frequency range filtering? I also had an idea to a "manual" method of doing such a measurement, by attacking a little cardboard shield to the mid. I will see what changes this makes.

zoom out fully of the IR window and it will become obvious how big the window has to be.
in my case I use 250ms:

gg.jpg
 
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
Why are you not just using the moving mic method?

You already have the anechoic data for your speakers above the schroeder frequency.
There's all sorts of methods people suggest in articles/posts/videos/etc, but I can only do 1 at a time. But what exactly are you suggesting to do here? How should I incorporate the anechoic data? As the room target?
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
1,992
Likes
2,598
Location
NL
What I would say is just take a MMM around the MLP and compare it with a single point measurement at the MLP. Then also compare it with the anechoic data that’s available.

My conclusions from an MMM are pretty much always that my bass issues are less severe than the SPM suggests. And above say 1kHz, anechoic anomalies often shine through in the MMM, giving me another starting point as to what and what not to correct there and to what extent.
 
Last edited:

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,379
Likes
5,455
Location
Melbourne, Australia
There's all sorts of methods people suggest in articles/posts/videos/etc, but I can only do 1 at a time. But what exactly are you suggesting to do here? How should I incorporate the anechoic data? As the room target?

Which city are you in? I am in Melbourne. There are a couple of us around in Melbourne who could drop by and give you a hand?
 

moonlight rainbow dream

Active Member
Joined
Nov 24, 2020
Messages
154
Likes
219
There's all sorts of methods people suggest in articles/posts/videos/etc, but I can only do 1 at a time. But what exactly are you suggesting to do here? How should I incorporate the anechoic data? As the room target?

I think Amir laid it out very succinctly. Take the MMM measurement... spatial average is very important to get a realistic appraisal of the bass performance. Use parametric EQs to pull down peaks below the room transition frequency. Do not EQ above the transition frequency according to your own measurements; it's very easy to screw up the sound this way. Only EQ according to anechoic data to flatten... and in your case with the IN-8, I wouldn't even do that because the on-axis anomalies in the frequency response are diffraction effects common in concentric designs and resolve themselves once you toe them in a small amount.

This is going to be your baseline. Now if you still don't like the sound you can tune according to your personal preference and musical genre tastes. You can add in various amounts of bass boost/shelving/rise below ~150hz and you can add in varying downward-sloping tilt above that frequency. A decent rule of thumb is to add more overall tilt to the response if you're listening to modern pop type production (everything recorded separately and then put together by mix engineers)... and less tilt if you're listening to say classical recordings where they're actually sticking mics in a live venue.
 
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
Which city are you in? I am in Melbourne. There are a couple of us around in Melbourne who could drop by and give you a hand?
Perth actually. Hopefully I will get to it today and can report back some actual progress.


A decent rule of thumb is to add more overall tilt to the response if you're listening to modern pop type production (everything recorded separately and then put together by mix engineers)
Interesting. I don't think I have a particular sound preference, as long as its all there and clean. I actually listen to heavy metal pretty much exclusively, so issues in the lower end of the mid range really stand out.
 

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,379
Likes
5,455
Location
Melbourne, Australia
Perth actually. Hopefully I will get to it today and can report back some actual progress.

I know a guy in Perth who can help you. If you are interested, inbox me and I will get him to contact you. He is an enthusiast and loves seeing systems and measuring them and the only thing he wants is more exposure and to learn.
 
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
I'm trying out MMM at the moment (EDIT: With no EQ or HP or anything - raw in room response), testing the affects of different configs of pink vs pink periodic, 50% vs 87% overlap, etc. Here are some examples with psy smoothing.
(The little difference at 82Hz was me mucking around with a PEQ, to see if it would affect much at 160 and 240.)

mm_tests.jpg
  1. The results are quite consistent above 700Hz, but it does slope down more than the estimated in-room response. Why is that?
  2. Is the estimate at a decent relative level? I did not chose that level, it just imported there (so random luck), but it does look okay considering it matches the measurements around 80Hz and 240Hz.
  3. So ~170Hz and ~280Hz peaks are still evident. Is ~850Hz a problem too? If I lower the estimate a little, it would be more apparent.
Actually looking at the screenshot I just uploaded, the MM is flatter above 2kHz than it looked in program. Adjusting the the estimate to match the upper mids gives this;
mm_tests_lower.jpg

I'm going to measure left and right separate now and check for phase issues like this post in MMM thread suggests can happen. Then I might try moving the speakers around, and see what that does.

I know a guy in Perth who can help you. If you are interested, inbox me and I will get him to contact you. He is an enthusiast and loves seeing systems and measuring them and the only thing he wants is more exposure and to learn.
I'll keep slogging away by myself for now, but I'll keep that in mind. Thanks.
 
Last edited:
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
Looking at the Left and Right separately has revealed a problem. Here's the individual MM for the left and right compared to the estimated in room;
mm_l_r.jpg
There's good compliance above 1200Hz, whilst 280Hz and 850Hz peaks are still prevalent. Also 390Hz is looking like it might need some attention too. But the real problem seems to be 80Hz and 160Hz. For whatever reason, the Left channel is responsible for both peaks. I tried different combos of doors and curtains opened and closed, and also tweaked the position even more accurately (it was already within 1cm), but the difference was negligible.

When I compare MM of both speakers (LR=stereo) vs the L+R arithmetic sum, it's not too different, but actually it now shows a phase issue at ~90Hz.
mm_l+r.jpg
But when I look at the L and R compared to the LR measurement, those spots stand out.
mm_lr_comp.jpg

So what should I do here? The left speaker/wall is basically that of a perfect rectangle room, whereas the right speaker has a window recess as well as the janky "far" corner. I'm doubting that speaker position will help, considering I've accurately located them relative to their corners (<1cm). So should I just put a -7dB PEQ on the left channel at 160Hz? Similar at 80Hz? What can I do about the phase issue at 90Hz? Or have I got it backwards and the left is a better response, and perhaps the right is suffering from its own phase issues?

Actually, I've often thought the left side was a little louder, to the point where I've gone through windows and stuff to make sure I didn't accidentally pan it 1% or something. But it never is. That's left me to think it's either a slightly biased panning in particular songs, or that my left ear is a little more sensitive, but maybe it's just this peak at 160Hz?

Update: Ok, that actually seemed to work. -7dB and 12Q on the left channel pulled it down equal to the right channel. The resulting stereo measurement is now 3dB lower at that point. The 80Hz EQ wasn't as successful (-5dB and 15Q). It actually caused the stereo measurement to plunge in to the 90Hz issue even worse. I wonder if the problem here is actually the right speaker dips at 60Hz and 84Hz, and perhaps they are SBIR issues from the window recess?
mm_left_eq_test1.jpg
And what does it mean when you add the left and right together, but then the stereo measurement is much higher? Is that just confirmation of room mode issues / energy building up at that frequency? For example;
70Hz has L=72.9dB R=72.5dB SUM=75.5dB Actual=77.0dB.
160Hz is L=78.0dB R=77.0dB SUM=80.2dB Actual=82.5dB.
 
Last edited:

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
1,992
Likes
2,598
Location
NL
Interesting.

It may have been mentioned in one of the previous posts, but where does this estimate come from exactly?

Just trying to understand what you are comparing with what.
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
1,992
Likes
2,598
Location
NL
And just to be sure - your latest measurements are with or without EQ?
 
OP
N

neRok

Senior Member
Joined
Oct 7, 2022
Messages
303
Likes
175
Location
Australia
Just now I turned off the 80Hz PEQ and put in an Allpass filter of 90Hz and 15Q. I'm not 100% sure what allpass does, but I read once somewhere it does stuff to the phase. Then I measured the left only and it was the same (as expected without the 80Hz PEQ), and then I measured the stereo and 90Hz dip has improved! Does this suggest I'm on the right track?
Edit: This seems a legit fix because I did a quick sweep with the allpass on and off, and in this area <300Hz sweep and MMM are the similar, and the sweep shows the improvement.
mm_lr_allpass.jpg

And just to be sure - your latest measurements are with or without EQ?
The MMM measurements in my 2 previous posts were with no EQ applied (nor a highpass or anything), except for the 2 most recent results where I've EQ'd the left only.

I'm doubting that speaker position will help
Quoting myself because now I wonder if speaker position will help if this is an SBIR issue due to the window recess? Because if I move the speaker away from the window, that will change the relationship of distance vs frequency. Further means longer wavelengths, so lower frequency yer? But also it means more multiples of higher frequencies in the same distance? So maybe worse moving it away? One way to find out I guess, but it's getting cold and late, so maybe tomorrow for this test...
 
Last edited:

Gorgonzola

Major Contributor
Forum Donor
Joined
Jan 27, 2021
Messages
1,003
Likes
1,375
Location
Southern Ontario
This has been an erudite and lengthy discussion; apology but I have only scanned it.-- my aged vision makes it difficult for me to read lengthy texts, "subjectively" my eyes bleed.

However I especially appreciate @alex-z's providing the Toole/Olive subject preference curves:
index.php

My own preference corresponds most closely to the "trained-listeners" preference per the graphs; actually on the high end I might be a little closer to the "all listeners" curve but that would be on account of my ears' decline in HF sensitivity. I've had no "training" other than my own 50+ years of listening.

For EQ I use a crude & easy method, that is, MathAudio's Room EQ plug-in in my Foobar2000 player. I use its "Neutral" setting, slightly modified, rather than its "Bright", (i.e. flat) setting, as measured at my listening position. Works very well.
 

ozzy9832001

Senior Member
Joined
Feb 19, 2023
Messages
402
Likes
255
I would just squash the 2 areas centered around 160hz and 280hz.

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.
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,184
Likes
2,642
Location
Salvador-Bahia-Brasil
based on you meassured response the "toole target" (I think it was extracted from one of his books) looks more suited. The Harman one does boost you bass (realativly) so you will feel loss of mids

attached

seeing the graph on post 58 for the perhaps 10th time I now realize that this target is actually the

1688047614046.png
 
Top Bottom