• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). 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!

My first attempt at REW measurement. Am I doing this correctly?

mushjoon

Member
Joined
Nov 29, 2021
Messages
18
Likes
9
Hi everyone, I am trying to set up Neumann KH80 for mixing practice and as you can see, my room is super tiny.

The dimensions are 280cm x 250 x 240 (L x W x H) and that would be 9.18ft x 8.2 x 7.8 for my American friends.

I used Behringer ECM8000 and played sine sweep playing through both speakers simultaneously so I can simulate how I would normally listen to music.

The first three sets of photos include speakers placed at the edge of the desk to minimize reflections bouncing off my desk surface.

The speakers are 57cm away from the front wall and roughly 110cm from side walls.
The speakers are 68cm apart from each other.
The mic is placed 87cm away from the front wall.

KakaoTalk_20211219_210949930.jpg

KakaoTalk_20211219_210926559.jpg

KakaoTalk_20211219_210950178.jpg



Now, I placed the speakers as close as possible to the front wall. And as you will see in the graphs below the dips in the low end are worse than "speakers on the edge" position.

KakaoTalk_20211219_212620154.jpg
KakaoTalk_20211219_212643247.jpg


Here are the graphs for the desk-edge position:

measure1_dec_19.png


decay_dec_19.png


spectogram_dec_19.png


waterfall_dec_19.png


measure2_dec_19.png


I apologize if I forgot to provide a crucial graph so I'll include REW file as an attachment.

As you can see, even with the on-the-edge position (red) there are some horrible dips at 75hz, 160hz, 390hz and 700hz.

What can I do to fix those dips and make my response as linear as possible?
 

Attachments

  • dec_19_ver2_measure.zip
    3.5 MB · Views: 48

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
I think you have the potential for really, really good sound. I played with your measurements and had some thoughts about them and what I might do:

1. You took good starting measurements
2. Many of the mostly narrow (and some broader) peaks and dips are likely SBIR. I would enter my dimensions into this to look at it further/confirm: http://tripp.com.au/sbir.htm
3. To get the potential sound quality you are going to need EQ (via parametric EQ and perhaps shelving filters). Not knowing what you have, the software you use to play your music, whether you want to EQ other sources, budget, etc. I can just think of what I would use (Mac-based):
- the boundary settings, EQ on the monitors (those highs have to come down, it must sound very bright now)
- VST (PC) or AU (Mac) plugins. I use Audirvana as my music player, would maybe use "Audio Hijack" to capture full-system sound/host AUs
4. It is much easier to bring down peaks that fill in gaps from room modes or SBIR. That perhaps suggests using your closer to the wall positions (look at approx. 600hz- to 1.2khz below)
5. A moving mic measurement around your ears/head might add nice information re. what FR aberrations are significant.
6. I suspect(?) a nearfield setup makes it less important, but you might try adding frequency dependent windows for consideration
7. I would make sure the mic is at the height of my ears (maybe you are tall? maybe you set the mic up at tweeter height?)
8. I have never set-up a nearfield system, but if I did I would wonder about potential coupling to the desk and having it resonate, become another sound source. I would probably couple the speaker to the stand with "Blutak" and maybe spikes to the desk (or look into the products designed for studios for this). I would even consider some constrained layer damping sheets (CLD) to the bottom of the desk :)
9. Adding even a single sub where you could vary XO freq., phase, and further EQ would be very likely to smooth out the low frequencies (and extend them of course). The KH 750 DSP would be the natural choice, but it could likely be done less expensively.

Bill
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
I think a room curve is very important. This shows the one I look and the settings I used to develop one for your setup.
Screen Shot 2021-12-19 at 11.54.00 AM.png
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
I aligned your frequency responses at 1 Khz, chose a baseline level around 1 khz for the house curve/EQ generation (also trying to find a middle range where the peaks and dips are relatively balanced) and came up with this. The first is with variable smoothing. Less smoothing allows finding the exact frequencies for EQ and the Qs of the filters. Sometimes looking with psychoacoustic smoothing makes one feel a little better :) and see the broader trends. That is the first:
psychoacoustic smoothing.jpg


FR with target curve.jpg
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
Here is the EQ section with the settings I used. They are tight so generated a huge number of filters:

Screen Shot 2021-12-19 at 12.13.48 PM.png
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
As I said, it generated a million filters. I probably wouldn't approach it exactly that way:

1. I suspect the highs may simply need a shelving filter at 2khz or so.
2. I would balance using multiple high Q filters in the lows v a broader (lower Q) filter that encompassed several of the peaks. As you go higher in frequency (perhaps especially > 600hz) narrow peaks and dips are less audible so you should perhaps be less fastidious there. Sometimes, though, when you do that you will see nice improvement in the decay/waterfall.

Raw, target, and predicted:

Raw and target.jpg
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
The above would sound very, very good to me. Here is a busy image of the filters (again, not necessarily what I would do exactly):

FR, predicted, filters.jpg
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
Lastly, the filter details. Too many, but interesting. I'm done :). Have fun. You have a lot of potential with those great monitors.

Also note I don't consider myself an REW expert, having only recently switched over from Fuzzmeasure. I am sure there are many, many here with much more expertise. In that regard I am open to correction and learning.

Bill

Screen Shot 2021-12-19 at 12.35.17 PM.png
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
2,262
Likes
1,630
Location
Canada
I presume there isn't a correct calibration for the microphone. I would not touch anything above 1kHz based on these measurements. Download the klippel data files from Amir's review and use the curves there for your reference -- primarily the listening window (LW) and on-axis.

I also would be wary of any correction in the transition (middle) zone from any single point measurement, esp. without any kind of pre-filtering applied beforehand.
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
I wondered about the high frequency measurements as well and had looked up his mic, which they claim to be flat and doesn't have a calibration file.


But yes, one has to wonder, though he is measuring at < 1m, it seems. Making sure he is on the correct axis, many more measurements where his head will be and more, a moving mic measurement, using FDWs might shed some light. He still will need a house curve, I think, for nearfield listening, perhaps basing it on Amir's measurements (of the direct sound?) and tailoring to suit.

Bill
 
OP
M

mushjoon

Member
Joined
Nov 29, 2021
Messages
18
Likes
9
I think you have the potential for really, really good sound. I played with your measurements and had some thoughts about them and what I might do:

1. You took good starting measurements
2. Many of the mostly narrow (and some broader) peaks and dips are likely SBIR. I would enter my dimensions into this to look at it further/confirm: http://tripp.com.au/sbir.htm
3. To get the potential sound quality you are going to need EQ (via parametric EQ and perhaps shelving filters). Not knowing what you have, the software you use to play your music, whether you want to EQ other sources, budget, etc. I can just think of what I would use (Mac-based):
- the boundary settings, EQ on the monitors (those highs have to come down, it must sound very bright now)
- VST (PC) or AU (Mac) plugins. I use Audirvana as my music player, would maybe use "Audio Hijack" to capture full-system sound/host AUs
4. It is much easier to bring down peaks that fill in gaps from room modes or SBIR. That perhaps suggests using your closer to the wall positions (look at approx. 600hz- to 1.2khz below)
5. A moving mic measurement around your ears/head might add nice information re. what FR aberrations are significant.
6. I suspect(?) a nearfield setup makes it less important, but you might try adding frequency dependent windows for consideration
7. I would make sure the mic is at the height of my ears (maybe you are tall? maybe you set the mic up at tweeter height?)
8. I have never set-up a nearfield system, but if I did I would wonder about potential coupling to the desk and having it resonate, become another sound source. I would probably couple the speaker to the stand with "Blutak" and maybe spikes to the desk (or look into the products designed for studios for this). I would even consider some constrained layer damping sheets (CLD) to the bottom of the desk :)
9. Adding even a single sub where you could vary XO freq., phase, and further EQ would be very likely to smooth out the low frequencies (and extend them of course). The KH 750 DSP would be the natural choice, but it could likely be done less expensively.

Bill
First of all, thank you very much for such detail and informative advice!!!

I entered all the numbers from my on-the-edge position into the SBIR calculator mentioned in your reply and by god, I didn't expect that the calculator would correctly predict that I'll have a dip in 78hz caused by the side wall!

Screenshot 2021-12-20 121915.png


Just a few more questions:

1) Why did you use Harman curve for the target line instead of a flat, linear line? I thought EQ-ing for music production/mixing requires matching the speakers' response to a flat line.

2) Why did you choose the closer-to-wall position over on-the-edge position? I know that you chose it because you said that it's easier to bring down peaks than to increase the dips caused by room modes and SBIR. But isn't it a better option to choose the red line because it looks more linear than either the green or purple line?

3) I forgot to mention this. The purple line in the REW data zip file is from the same closer-to-wall position as the green line EXCEPT the mic was positioned much closer to the front wall (60cm instead of 87cm) because I suspected that I would be able to mitigate some desk reflections by doing so.

4) For the basic measurement procedure, I tried to position the mic to my ear height (approx 112cm) and maintain the same distance from both speakers.

5) I read somewhere about "fixing it right at the source" mentality regarding the music production and room acoustics. So that would entail that I should invest in some acoustic treatment even trying to fix my speaker response with EQ. If I had to install 100T glasswool insulation panel for low-end frequency absorption, where should I put it first? Front wall, side wall or both?

Once again, thank you for your reply
 
OP
M

mushjoon

Member
Joined
Nov 29, 2021
Messages
18
Likes
9
I presume there isn't a correct calibration for the microphone. I would not touch anything above 1kHz based on these measurements. Download the klippel data files from Amir's review and use the curves there for your reference -- primarily the listening window (LW) and on-axis.

I also would be wary of any correction in the transition (middle) zone from any single point measurement, esp. without any kind of pre-filtering applied beforehand.

I don't really know how I can use Amir's SPL curves for my reference. Should I try to mimic how Amir measured his speakers by positioning the mic pointing directly at a single speaker from very close distance and re-calibrate my ECM8000's response according to his data?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
2,262
Likes
1,630
Location
Canada
To provide some bit of sanity check:

1639973144662.png

From Bill's PEQs in post #8

Based on your measurements and Amir's, I would prefer not to EQ anything above 300 or 500 Hz other than a very simple tailored HF EQ tilt according to your own personal preference -- if, and only if it fits your listening preference. If you aren't sure right now, just leave it flat as-is. Alternatively, you can reduce the HF by pointing the speakers straightforward i.e. no toe-in which is what I currently do with the KH120 for my desk.

It's better if you measure each Lt and Rt channels individually. But with the very limited single point data on hand, this is one way that I'd do it:

1639974477797.png


1639974489085.png


For mid and HF tonal balance EQ, refer to the Klippel data.
 

Attachments

  • CEA2034 KH80.zip
    97.8 KB · Views: 40

Backflash

Member
Joined
Oct 13, 2021
Messages
51
Likes
24
Own same mic, 30cm away from Adam t5v it goes with anechoic data from Amir for me, more or less. For HF room plays quite a bit if it's stuffy I think, I got a huge 5m wide book cabinet with glass and speakers point at niche filled with wooden and metal things in it only 1.5m away from it, I need my setup to stay like this for reasons, so my data is a little weird too. I have a bump in 600 range(among other bumps at lower fq), which is correct as much as I can hear it, and a drop in 3-5k range, that's the desk and Adam HF tuning align to make it worse.
Mic sure got some variance though.
ecm8000_freq_resp.jpg
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
You're welcome. I just played with it while watching the Liverpool match :)

I really like the eq @ernestcarl provided above. His predicted red curve above would sound very, very good in the lows through the midrange. I have been hoping that I pointed out that I wouldn't have used all the eq filters I showed in my post, that it could and probably should be done more simply (and with acoustic approached first), limited above 5-600 Hz at the highest, and after lots more measurements. My goal was to show pictures that could lead to playing around on your own. Firstly, you need to be confident that the graphs represent what you are hearing, this perhaps the most difficult aspect.

It would be interesting to compare the measurements of @Backflash to Amir's in detail to try and get a feeling for how much you can trust your mic.

1) Why did you use Harman curve for the target line instead of a flat, linear line? I thought EQ-ing for music production/mixing requires matching the speakers' response to a flat line.

For reproducing music, a flat line at the LP is incorrect. Trust me. It will be too bright. I am not sure about mixing. It is of course all about translation. If you mix with a flat line will your mixes be too dull in the HFs or sound correct? I am no expert in this regard. I suspect a search on gearspace might lead to the answer.


2) Why did you choose the closer-to-wall position over on-the-edge position? I know that you chose it because you said that it's easier to bring down peaks than to increase the dips caused by room modes and SBIR. But isn't it a better option to choose the red line because it looks more linear than either the green or purple line?

My eye was caught by the dip on the green line, though if you could determine where it is coming from and mitigate it, the forward position may be the way to go. Certainly that position would mitigate reflections from the desk (something I would be nervous about) and might be wonderful. See again the red line above.

3) I forgot to mention this. The purple line in the REW data zip file is from the same closer-to-wall position as the green line EXCEPT the mic was positioned much closer to the front wall (60cm instead of 87cm) because I suspected that I would be able to mitigate some desk reflections by doing so.

You have to figure out what your ears will be hearing. If your ears hear the desk reflections it seems they would have to be included, especially as those reflections will be very early. One idea would be to compare the IRs of the forward and back positions to see if you can find the reflection as a second impulse. you could add the distance from the speaker to the desk and the desk to your ear and will pretty much know how far out in time you might find it. To find the spot where it hits the desk put a mirror on the desk with your head in the listening position and look for the speakers' reflections.

4) For the basic measurement procedure, I tried to position the mic to my ear height (approx 112cm) and maintain the same distance from both speakers.

Perfect. Though as mentioned above you should do each speaker individually. There is also a good video somewhere on here that might help. It is a demonstration of the moving mic technique in a near field monitoring situation like yours. I don't have experience with it, but it could be an interesting datapoint.

5) I read somewhere about "fixing it right at the source" mentality regarding the music production and room acoustics. So that would entail that I should invest in some acoustic treatment even trying to fix my speaker response with EQ. If I had to install 100T glasswool insulation panel for low-end frequency absorption, where should I put it first? Front wall, side wall or both?

I strongly believe in treating the room as the first measure, but have seen folks struggle with it in a near field setup like yours. I would love to treat the back wall with adequate absorption to mitigate SBIR, but computer monitors make this difficult. Move your monitor and put a mirror on your back wall. Where you see the speakers is where the reflection is occurring to some degree. Because your monitors have nicely controlled directivity you are in better shape. In addition the angle to the baffle is very high, so it might be that the reflections will be only where your monitors lose directivity in the LFs (where the woofer is radiating omnidirectionally, i.e. the woofer is small relative to the wavelengths being produced). Unfortunately, that means the absorption needs to be thick to be effective. My systems have always sounded better with absorption behind he speakers.

If you find an impulse in your IR from a sidewall that is less than 30ms I would also treat that. And I would actually be tempted to treat all first-reflection points in the room, though this is controversial (perhaps less so in mastering) as some prefer the envelopment of early lateral reflections while I prefer the image specificity of eliminating them. Listening to mostly classical and acoustic jazz, I want to hear the ambience of the hall/studio, not the room.

Anything you can do to fix the room rather than using eq is preferred. If you could get it as good as possible acoustically then limit your eq to as low as possible (300 hz or lower) you would really be doing well. +/- a gentle shelving filter in the highs if you determine it is a good approach.


And please don't consider what I say dogmatic, just my thoughts :)

Bill
 

Bill Brown

Member
Joined
Sep 25, 2021
Messages
83
Likes
85
I got curious about desired FR for mixing. Scanning through gearspace didn't find much, though my search was fairly cursory.

I then thought of Bob Katz. Heck, one of the room curves is named after him, so I know his mastering system uses it. Re. the concept in general, I think I may be right:

Mixing/Mastering for “Flat” Loudspeakers: What is really “flat” anyway?

Dear Bob
,

I write because I noticed you apply a slightly tilted room curve to your mixing room. But why not make a flat response as reference and then mix your recordings from that? To a “straw man” like my that would seem more logical because a tilted curve will introduce a spectrum bias which I as a consumer would need to implement in my system as well. Whereas if a mix made from a flat response means that if my speakers (whether equalized or not) are pretty much flat I will get the approximately same sound impression as in your studio. Of course, a complete replica is not likely unless I also aim for the same RT level as in your studio.

Anyway, it was just something that got me curious and finally I just had to ask for the reason for a tilted curve in a mixing studio, which for the uneducated man (like myself) seems odd.
I certainly mean no disrespect, since your knowledge in this field is manifold greater than mine. So it with the utmost respect that I ask :).

Best regards,

Mikkel G. Hangaard

Dear Mikkel:

Thanks for your letter.

What you have brought up is the classic “chicken and the egg” situation. It is true that our recordings have to complement our reproducers. if we equalize our reproduction system to flat (however that is measured), it would cause us to produce mixes and recordings which would be duller sounding than our current mixes! But historically, that has not been the case. We have a large recorded legacy of recordings that sound perfectly good on playback systems that measure some high frequency rolloff. All the major accepted reproduction systems since the beginning of time measure rolled off at the high end to some degree. So basically you are asking us to try to produce new recordings that would not be compatible
with the older reproduction systems.

It’s academic: If today, suddenly, all playback systems were made flat, then all or most recordings which already have been made would sound too bright. It’s a long, historical precedent and consistency is more important than absolute conformance with “flat”.

Lastly, there is a whole question of how “flat” is to be determined. Loudspeakers which have a wide, flat dispersion characteristic sound very different than loudspeakers which are flat on axis but rolled off off axis. How do you determine what is flat in that case? Then getting into questions of reflections from the side walls in the room and how they are treated. Lastly, the question of the measurement method and the measurement window, should it be anechoic or should it include some reflections in the room? There are so many different ways of determining the window that unless you know exactly how a measurement was taken, you cannot effectively judge a published loudspeaker measurement.

All these variables produce varied amounts of “measured” rolloff. In other words, even the definition of how to define “flat” is put into question. So even if we were starting from scratch with new recordings and new reproducers, there would still be a fundamental disagreement as to how to measure and determine what “flat” really means and we would still get nowhere!

So in the end, the egg came first, and the chicken followed, and we just have to keep on keeping on!
Hope this helps,

Bob

 
OP
M

mushjoon

Member
Joined
Nov 29, 2021
Messages
18
Likes
9
To provide some bit of sanity check:

View attachment 173501
From Bill's PEQs in post #8

Based on your measurements and Amir's, I would prefer not to EQ anything above 300 or 500 Hz other than a very simple tailored HF EQ tilt according to your own personal preference -- if, and only if it fits your listening preference. If you aren't sure right now, just leave it flat as-is. Alternatively, you can reduce the HF by pointing the speakers straightforward i.e. no toe-in which is what I currently do with the KH120 for my desk.

It's better if you measure each Lt and Rt channels individually. But with the very limited single point data on hand, this is one way that I'd do it:

View attachment 173522

View attachment 173523

For mid and HF tonal balance EQ, refer to the Klippel data.

I just entered your filter setting and applied it to Equalizer APO and the low-end rumbling sounded like more in control!!! I still have to make DIY bass traps and measure the response again but thank you for the suggestion!
 
OP
M

mushjoon

Member
Joined
Nov 29, 2021
Messages
18
Likes
9
I got curious about desired FR for mixing. Scanning through gearspace didn't find much, though my search was fairly cursory.

I then thought of Bob Katz. Heck, one of the room curves is named after him, so I know his mastering system uses it. Re. the concept in general, I think I may be right:

Mixing/Mastering for “Flat” Loudspeakers: What is really “flat” anyway?

Dear Bob
,

I write because I noticed you apply a slightly tilted room curve to your mixing room. But why not make a flat response as reference and then mix your recordings from that? To a “straw man” like my that would seem more logical because a tilted curve will introduce a spectrum bias which I as a consumer would need to implement in my system as well. Whereas if a mix made from a flat response means that if my speakers (whether equalized or not) are pretty much flat I will get the approximately same sound impression as in your studio. Of course, a complete replica is not likely unless I also aim for the same RT level as in your studio.

Anyway, it was just something that got me curious and finally I just had to ask for the reason for a tilted curve in a mixing studio, which for the uneducated man (like myself) seems odd.
I certainly mean no disrespect, since your knowledge in this field is manifold greater than mine. So it with the utmost respect that I ask :).

Best regards,

Mikkel G. Hangaard

Dear Mikkel:

Thanks for your letter.

What you have brought up is the classic “chicken and the egg” situation. It is true that our recordings have to complement our reproducers. if we equalize our reproduction system to flat (however that is measured), it would cause us to produce mixes and recordings which would be duller sounding than our current mixes! But historically, that has not been the case. We have a large recorded legacy of recordings that sound perfectly good on playback systems that measure some high frequency rolloff. All the major accepted reproduction systems since the beginning of time measure rolled off at the high end to some degree. So basically you are asking us to try to produce new recordings that would not be compatible
with the older reproduction systems.

It’s academic: If today, suddenly, all playback systems were made flat, then all or most recordings which already have been made would sound too bright. It’s a long, historical precedent and consistency is more important than absolute conformance with “flat”.

Lastly, there is a whole question of how “flat” is to be determined. Loudspeakers which have a wide, flat dispersion characteristic sound very different than loudspeakers which are flat on axis but rolled off off axis. How do you determine what is flat in that case? Then getting into questions of reflections from the side walls in the room and how they are treated. Lastly, the question of the measurement method and the measurement window, should it be anechoic or should it include some reflections in the room? There are so many different ways of determining the window that unless you know exactly how a measurement was taken, you cannot effectively judge a published loudspeaker measurement.

All these variables produce varied amounts of “measured” rolloff. In other words, even the definition of how to define “flat” is put into question. So even if we were starting from scratch with new recordings and new reproducers, there would still be a fundamental disagreement as to how to measure and determine what “flat” really means and we would still get nowhere!

So in the end, the egg came first, and the chicken followed, and we just have to keep on keeping on!
Hope this helps,

Bob

I can't thank you enough for going through all of my questions and answering each one of them. I'll definitely keep your suggestions in mind and use them later as I try to improve my room in the future. Thanks!
 
Top Bottom