• 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!

EQ strategy advice for this room please

tw99

Senior Member
Joined
Mar 13, 2019
Messages
467
Likes
1,069
Location
West Berkshire, UK
The following measurements are from a 2.0 setup, although I do have a powered sub that could be added if it was going to make a big difference (currently it only gets turned on for films). I'm using a MiniDSP SHD, and I've got one preset set up using Dirac to correct the response up to around 200hz, which I've now got sounding OK, if not perfect.

I was interested to try another preset just with the MiniDSP filters, with Dirac disabled, using REW to measure and generate the filters. I'm finding it a bit difficult to optimise because I seem to have some particularly uneven response below 100hz. I've read that it's usually unhelpful to try to boost dips in the response, but I don't see how I can avoid some level of boost in this room, or there will be a pretty obvious dip remaining.

Graphs are all 1/6 smoothed. Speakers are ATC SCM40s, a 3-way with the crossovers at 380 and 3500hz.

This graph is an average of a few measurements at and around the normal listening position. There are some big resonances that need reducing, and the overall bass response is a bit lower than the mid and treble, but I'd like to maintain a Harman-like slope through the low frequencies.

raw.jpg


Now with the measured response including the REW-generated filters in green.

rewfilters.jpg


This was the target response and predicted filter behaviour, which seems to match what was measured quite well. But to get this, there are filters boosting by 4, 5 and just over 6db in places. The small fall in treble was just done with a manual filter afterwards as I found that much more of a fall ended up sounding dull.

filters.jpg

This is what the filters look like

filters1.jpg

Am I doing about as well as I can here ? Would be interested to hear any suggestions.
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
You did well! I would have done the same as you, except I wouldn't have filled the two small dips. How does it sound? Not too bright now?
 

Pio2001

Senior Member
Joined
May 15, 2018
Messages
317
Likes
507
Location
Neuville-sur-Saône, France
Hi,
Instead of 1/6 oct smoothing, you can use Var smoothing. It will give you more accuracy in low frequencies. Especially around 35 Hz where you can make very accurate corrections in order to remove small peaks that are masked by the 1/6 oct smoothing, while being perfectly audible.

To make faster measurements, you can play a pink periodic noise 64 k and measure it in the RTA window, using 1/48th RTA, 64k FFT, forever averaging.
Then you hold the microphone vertically, and move it around all the measurement zone that you are interested in until the curve stops changing. Then you can export the result in the main window with the "current" button.
This way you avoid making several measurements and averaging them.
 

Mike-48

Active Member
Joined
Feb 2, 2020
Messages
163
Likes
224
Location
Portland, Oregon
To my eyes, the 80 Hz to 200 Hz region is depressed by a few dB. I might like to do something about that, at least try it and see.
 
OP
tw99

tw99

Senior Member
Joined
Mar 13, 2019
Messages
467
Likes
1,069
Location
West Berkshire, UK
Thanks for the comments. I'll experiment with the VAR smoothing. @Mike-48, I agree, that region does have problems.

I'm not sure if it's best to try to address this by cutting more aggressively from 200hz up, or by trying to boost more in the 80-200 area.
 

Mike-48

Active Member
Joined
Feb 2, 2020
Messages
163
Likes
224
Location
Portland, Oregon
I'm not sure if it's best to try to address this by cutting more aggressively from 200hz up, or by trying to boost more in the 80-200 area.
I'm pretty empirical with this stuff. I'd try a low-q boost of a decibel or two centered around 120 Hz. In general, I don't think there's anything wrong with boosting broad areas a little. It's big boosts, or trying to fill sharp nulls, that are exercises in futility.

Oh, I almost forgot to say: if the woofers are the same distance from two or more surfaces (e.g., side and front wall), try changing that. If there is some boundary-related cancellation going on, maybe you can change that.
 
OP
tw99

tw99

Senior Member
Joined
Mar 13, 2019
Messages
467
Likes
1,069
Location
West Berkshire, UK
Revisiting this after a while, I've found that I somewhat prefer the results using Dirac compared to the REW-generated filters. Haven't done any blind testing of REW vs Dirac, so can't prove anything there. However, since Dirac can be switched in and out instantaneously with the SHD it's certainly very easy to hear that it provides a big improvement over no correction at all.

In the end I've settled on a Dirac curve as below. The extra dip at 73hz is to remove a resonance that didn't fully go away if there was just a straight line through that point. Subjectively this is better. So overall there's a mild to moderate bass boost and a small cut to midrange/treble to give an overall more pleasing response.

Screenshot 2021-03-19 123826.jpg
 
Top Bottom