• 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!

I`m lost. Totally awfull measurements in my room. Need advice.

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
Hi there,

I bought a UMIK-1 and tried to measure my Speakers in REW and I`m getting a very uneven frequency response. I used the EQ function and it threw a dozen of filters that are all of very high Q value and massive volume change.

Can you maybe have a look if this all makes sense? Not sure, if I`m doing something wrong. The algorithm reduced the volume by more than 12db in some areas. Are my speakers really crap or do you think the room can "destroy" the sound that much?

Here`s the before and after measurement:

rew.PNG


And this is the implemented EQ which the "after" mearuement is done with with a litte manual optimization above 600Hz

crave_eq_after.PNG


Really not sure if all of this makes sense. The sound changes drastically and I think definitely to the better. But it looks sooo radical.

Is there anything I could do to my room to lessen this rollercoaster bewtween 100 and 400Hz?

Sorry if these are dumb questions, but I really don`t know what to think about it.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,409
Likes
12,294
Location
UK/Cheshire
Look pretty normal to me. Very similar to what I am seeing in my small room. See my thread here:
https://www.audiosciencereview.com/...d-moving-mic-measurements-dont-line-up.26474/

(Skip the first few posts until I post plots after getting the Moving mic measurments working)

I even had to put a manual filter in mine to tame a 49Hz resonance.

To tame the room prior to eq, you could experiment with speaker positions. Especially moving them further from the walls, particularly the wall behind them. I also experimented with partial port bungs.
 

Another Bob

Member
Forum Donor
Joined
Nov 28, 2018
Messages
80
Likes
128
Location
Madison, WI
You're showing a really expanded vertical scale in that first picture, so things are going to look worse than what you see in typical speaker reviews - and in-room measurements will always look worse than anechoic (or gated or Klippel) measurements in any case. A 12 dB room mode is not too unusual. The folks here might be able to provide more insight if you would list the room dimensions, what speakers you're using, and where they're placed.
 

JeffS7444

Major Contributor
Forum Donor
Joined
Jul 21, 2019
Messages
2,347
Likes
3,508
Your room has plenty of low-frequency energy, and that's common for an untreated room, and it's the reason why some auto-eq routines ignore frequencies below about 150 Hz by default. See if you can manually correct using fewer, less-extreme filters, particularly around 800 Hz, 1700 and 6600. Also, if you are using subwoofers which offer a phase control, you might be able to smooth some of the low frequency peaks and valleys somewhat via said control.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Hi there, I bought a UMIK-1 and tried to measure my Speakers in REW and I`m getting a very uneven frequency response. I used the EQ function and it threw a dozen of filters that are all of very high Q value and massive volume change.
...
As others have said, that looks like a normal in-room response and you got excellent results from EQ.
The nulls around 33 and 52 Hz will likely resist EQ; room treatments may be more effective. Bass traps, bass resonators, or tube traps. This will likely also improve the transient response / tightness / spectral decay of the bass.
 
OP
W

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
Thanks, guys!

The speaker sit next to my monitor on a desk and there is a sub too. @JeffS7444 : I already played with the phase and it is set to 180° in both measurements, because with 0° the room modes were even worse/shifted. 60cm behind the loudspeakers is a big window and its a room with a roof pitch, so the room simulation in REW does not really work well. There is litte I can change regarding the positioning of the sattelites, however the sub is connected wirelessly and can go anywhere theoratically.

Additional room treatment would be possible and I will try this if it should help. But what exactly should I try? There are so many different types of treatment... Is there a way to get rid of the 33 and 52Hz dips with proper treatment?

Another question: Is the slope of the curve OK? I found and loaded the harman curve as a house curve and tried to allign to that... Not exactly sure if this was the right thing to do...

curve.PNG



@tonycollinet: Sorry, but what is a partial port bung?

I`ll play with some more EQ in the higher frequencies as Jeff suggests.

Soundwise with the current EQ music sounds transparent, vocals are clear and realistic but the setup is lacking some punch.
 
OP
W

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
You're showing a really expanded vertical scale in that first picture, so things are going to look worse than what you see in typical speaker reviews - and in-room measurements will always look worse than anechoic (or gated or Klippel) measurements in any case. A 12 dB room mode is not too unusual. The folks here might be able to provide more insight if you would list the room dimensions, what speakers you're using, and where they're placed.

Thank you.

Speakers: Nubert Nupro X-4000 RC & NuSub XW-900. Room is 4,5m x 5.2m with a large roof pitch on the right side.

5,20 is the width btw. (distance of the wall with the headphone to the wall next to the sub)

Placing: (sub is in the right corner)
placing.jpg
 
Last edited:

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
Another question: Is the slope of the curve OK? I found and loaded the harman curve as a house curve and tried to allign to that... Not exactly sure if this was the right thing to do...

That looks a lot like the “untrained listeners” curve.
To me it would sound boom-tizzy bassy and bright but what matters is whether it sounds good to you.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,409
Likes
12,294
Location
UK/Cheshire
Then three versions of modified water pipe insulation.

The one that worked best for me was the closest in this picture (with a quarter cut-out)


IMG_1806.png
 
OP
W

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
That looks a lot like the “untrained listeners” curve.
To me it would sound boom-tizzy bassy and bright but what matters is whether it sounds good to you.

No, that`s really not how it sounds to me. Without EQ it was like that maybe. I wasn`t aware that there is something like a untrained listener curve to be honest and do not remember where I found this curve unfortunately. Do you maybe have a reference link where to find alternative curves I could work/test with? Thank you.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
... Additional room treatment would be possible and I will try this if it should help. But what exactly should I try? There are so many different types of treatment... Is there a way to get rid of the 33 and 52Hz dips with proper treatment? ...
The short answer is "maybe". It depends on what is causing it. It is often either SBIR "speaker boundary interference response" or LBIR "listener ...".
For example, if your speaker is 2' from the wall behind it, the portion of the sound wave that travels backward and bounces forward has traveled 4' when it joins the forward wave, which is half a wavelength of 136 Hz, so it's 180* out of phase with 136 Hz and thus cancels or creates a null at that frequency. It is also a full wavelength of 272 Hz, so it would cause a peak at that frequency. The same can happen due to the listener position. If you sit 10' forward of the wall behind you, then the sound that hits that rear wall travels 20' after it hits the wall and bounces forward. That's half a wavelength of 52 Hz and a full wavelength of 104 Hz, so you'd get a null & peak at those frequencies respectively.

One way to test whether it is SBIR or LBIR is to move the speakers and listener/mic position around and measure at those different positions/distances. If the nulls change frequency then you've got SBIR/LBIR. If they stay the same then it's likely caused by the room, not by the positions.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,409
Likes
12,294
Location
UK/Cheshire
As I posted in my thread, I'm pretty certain my 49Hz peak is resonance in my room. The wavelength of 49Hz is almost exactly double the length of my room - so perfect for a single half cycle standing wave.
 

GimeDsp

Senior Member
Joined
Oct 30, 2020
Messages
418
Likes
362
Location
Earth
YO, how did you measure the speakers?

Did you download the cal file for you mic?

Did you load the cal file into REW?

When you do REW you REALLY first want to fix around 400hz and below first.
After that midrange room interaction takes place and causes lots of issues.

When fixing 400hz and below you will almost always be only subtracting with EQ, not boosting
Make sure to set REW to "0" boost and also allow narrow filters below "200hz"

PS--even with the best placement of subwoofers it's normal for small rooms to need around -18DB of cut to smooth things out.
 

FrantzM

Major Contributor
Forum Donor
Joined
Mar 12, 2016
Messages
4,337
Likes
7,731
SBIR and LBIR are insidious, yet poorly understood. A primer and ways to mitigate or correct their effects would be most appreciated.

Peace.
 
OP
W

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
YO, how did you measure the speakers?

Did you download the cal file for you mic?

Did you load the cal file into REW?

When you do REW you REALLY first want to fix around 400hz and below first.
After that midrange room interaction takes place and causes lots of issues.

When fixing 400hz and below you will almost always be only subtracting with EQ, not boosting
Make sure to set REW to "0" boost and also allow narrow filters below "200hz"

PS--even with the best placement of subwoofers it's normal for small rooms to need around -18DB of cut to smooth things out.


Yo!

How did you measure the speakers?
--> Placed the mic (umik-1) on my listening position on a mic stand and moved it a bit left/right and in the height. Then averaged the 5 or 6 measurements.

Did you download the cal file for you mic?

Did you load the cal file into REW?
--> Yes and Yes

Make sure to set REW to "0" boost and also allow narrow filters below "200hz"
--> Narrow filters were on but I have 1 positive gain filter in the calculated results... (+8,7db @116Hz) The others are all negative gains. (The +2db @55Hz were my own ideas as well as the peak filter @780Hz)

-18dB :-o Wow...
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
No, that`s really not how it sounds to me. Without EQ it was like that maybe. I wasn`t aware that there is something like a untrained listener curve to be honest and do not remember where I found this curve unfortunately. Do you maybe have a reference link where to find alternative curves I could work/test with? Thank you.

If you like how it sounds that is what matters.
Now that I see the figure yours looks more like the “All Listeners” which is “halfway” between trained and untrained

index.php
 
OP
W

Wegi76

Active Member
Joined
Apr 8, 2021
Messages
107
Likes
85
@tuga Agree. The low shelf is approx 6db in both curves. It looks like a smoothed version of the all listeners curve.
 

GimeDsp

Senior Member
Joined
Oct 30, 2020
Messages
418
Likes
362
Location
Earth
If you like how it sounds that is what matters.
Now that I see the figure yours looks more like the “All Listeners” which is “halfway” between trained and untrained

index.php
A full range curve and the actual summation between mains and sub are very important.
I was always taught to get the summed response flat then use global tilt to change the target, this results in same curve but allows mains to have more energy in Xover region which is Preferred for for panned LF content like rock music toms/symphonic stuff.
 
Last edited:
Top Bottom