Understanding the REW measurement of my small room

How to use REW, What is a Bass Trap, a diffuser, the speed of sound, etc.

Moderators: Aaronw, sharward

wrath_child
Posts: 2
Joined: Tue Mar 18, 2014 7:48 pm
Location: Cologne, Germany

Understanding the REW measurement of my small room

Post by wrath_child »

Hi guys!
long time reader, first time poster. :)
Now I am dealing with room acoustics and understanding the measurements of my room after treating it a bit.
Room info: (metric)
- length: 3,70m
- width: 2,53m
- height: 2,33m
- specific characteristic is the slant wall in the back. Behind it is our staircase… Kills the symmetry of the room but I have to live with that.
Materials:
- front wall: concrete
- right wall: concrete
- back wall: drywall (gypsum)
- left wall: drywall (wood)
All walls are covered with thin fabric, just for the look ;)
- floor: wood
- ceiling: suspended wood
Treating done: first reflexion points (walls, ceiling) with 10cm basotect absorbers. Basstraps behind wooden slabs in front corners.
IMG_5845klein.jpg
sketch: (all numbers are metric)
Skizze2.jpg
When measuring, I felt that the 83db recommended are very quiet. But maybe my iPhone db measuring app wasn’t right… ;)
Now the measurements:
both speakers
Beide.jpg
left speaker
Links.jpg
right speaker
Rechts.jpg
Hyperlink to .mdat:
Hyperlink
Now a few questions:
I think RT is quite good (besides 50hz and 100hz). Correct?
Any possibillty to correct that, or is the room size responsible?
How about the frequency curve. Good or bad? And what can be done to improve?
Hopefully I respected all rules for a first post.
Thanks, best wishes and please excuse my bad English :D
Alex
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Understanding the REW measurement of my small room

Post by Soundman2020 »

Hi there Alex, and Welcome! :)
- specific characteristic is the slant wall in the back. Behind it is our staircase… Kills the symmetry of the room but I have to live with that.
Fortunately, it is at the back of the room, so it's not a huge issue. If that were at the front, you'd have a problem...
Materials:
- front wall: concrete
- right wall: concrete
- back wall: drywall (gypsum)
- left wall: drywall (wood)
That's probably the reason for the very, very large differences between the left and right speakers.... I'm assuming that the three data sets in your MDAT file are L, R and LR, but I might be wrong, since they are not labeled...

Also, your speakers are not shown in the photo or on the diagram, so it is impossible to say if those differences (and the other artifacts) might be due to incorrect speaker placement...
Basstraps behind wooden slabs in front corners.
Those wooden slats are likely causing some of the artifacts visible in your REW data, but once again it's impossible to say, since the location and orientation of the speakers is unknown.

Also: How are those bass traps built? What is behind the wood slats?
When measuring, I felt that the 83db recommended are very quiet. But maybe my iPhone db measuring app wasn’t right…
Definitely something very major wrong there! According to REW, you did your tests at about 75 dB. Are you SURE you calibrated REW correctly? Are you SURE you had your iPhone app set correctly, and calibrated? In fact, forget about the app, and just get a proper calibrated hand-held sound level meter, such as an Extech or Galaxy. That will cost you about US$ 100 on Amazon or ebay.
I think RT is quite good (besides 50hz and 100hz). Correct?
It's reasonable for that size room, yes. But RT60 is not really a major measure of control room response. It is one of several measurements, that, when taken together, define if the performance of the room is good or not. Your RT60 is OK, but the frequency response is not, there are major modal and SBIR issues, there are large reflections within the Haas time, the phase response is way off, and there is clear comb filtering starting way down low.
Any possibility to correct that, or is the room size responsible?
It is room size to a certain extent, but also treatment. There's a LOT that can be done with treatment to change the RT curves...
How about the frequency curve. Good or bad?
Bad. You have dips of over 20 dB and peaks of over 10, with respect to the average, implying swings of over 30 dB. EBU specs call for the response to be flat within +/-3 dB for a high quality control room...
And what can be done to improve?
Bass traps, to start with. Large ones, deep ones, "superchunk" style, in as many of the remaining corners as you can, but also with plastic on the front to prevent losing too much of the high end. You didn't mention what treatment you have on the back wall, which is critical, but from the REW graphs I'm guessing that it isn't much at all...

But before starting with that, we first need to make sure that your speaker / listening position geometry are correct. However, as I mentioned above, you didn't show any of that in the photo or diagram, so it is impossible to say...

:)

- Stuart -
wrath_child
Posts: 2
Joined: Tue Mar 18, 2014 7:48 pm
Location: Cologne, Germany

Re: Understanding the REW measurement of my small room

Post by wrath_child »

Hey Stuart,
thanks for your fast reply :)
Let me try to answer your questions!
I'm assuming that the three data sets in your MDAT file are L, R and LR, but I might be wrong, since they are not labeled...
Also, your speakers are not shown in the photo or on the diagram, so it is impossible to say if those differences (and the other artifacts) might be due to incorrect speaker placement...
You are right, the three sets are L,R,LR (i thought i labeled them with links, rechts, beide. In german, sorry for that!).
My speakers are placed like this:
Speakerplacement.JPG
Those wooden slats are likely causing some of the artifacts visible in your REW data, but once again it's impossible to say, since the location and orientation of the speakers is unknown.
Also: How are those bass traps built? What is behind the wood slats?
Behind is 30cm of Rockwool.
Definitely something very major wrong there! According to REW, you did your tests at about 75 dB. Are you SURE you calibrated REW correctly? Are you SURE you had your iPhone app set correctly, and calibrated? In fact, forget about the app, and just get a proper calibrated hand-held sound level meter, such as an Extech or Galaxy. That will cost you about US$ 100 on Amazon or ebay.


Just as i assumed. I am going to buy myself a proper hand-held an redo the measurement!
It's reasonable for that size room, yes. But RT60 is not really a major measure of control room response. It is one of several measurements, that, when taken together, define if the performance of the room is good or not. Your RT60 is OK, but the frequency response is not, there are major modal and SBIR issues, there are large reflections within the Haas time, the phase response is way off, and there is clear comb filtering starting way down low.
You have dips of over 20 dB and peaks of over 10, with respect to the average, implying swings of over 30 dB. EBU specs call for the response to be flat within +/-3 dB for a high quality control room...
I understand, so its way worse as I thought.
Bass traps, to start with. Large ones, deep ones, "superchunk" style, in as many of the remaining corners as you can, but also with plastic on the front to prevent losing too much of the high end. You didn't mention what treatment you have on the back wall, which is critical, but from the REW graphs I'm guessing that it isn't much at all...
Right, no treatment in the back yet! :)
So how deep should those superchunks be? is a triangle with 20cm edge lenght enough for the start?
But before starting with that, we first need to make sure that your speaker / listening position geometry are correct. However, as I mentioned above, you didn't show any of that in the photo or diagram, so it is impossible to say...
:)
As addition to my sketch i tried to make a panorama foto of the speakerplacement, looks wierd because of the image deformation :)
IMG_6339klein.jpg
Regards. Alex
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Understanding the REW measurement of my small room

Post by Soundman2020 »

My speakers are placed like this:
OK, there's your first problem. They need to be set up much better than that.

First, you have them the same distance from the front wall and side walls, meaning that all SBIR and comb filtering artifacts will occur at twice the amplitude, at the exact same frequency! :shock: Like that, the frequency response will have a major dip at 106 Hz, a major peak at 53 Hz and 124 Hz, plus other peaks and dips at increasingly lower magnitude all the way up the scale, creating a comb-filtering effect (but probably not audible above about 2kHz). Sound familiar? :) (Take a look at your current REW graphs....)

Next, the speakers are too far away from the front wall. For a small room like that, they need to be right up against the front wall. The room is not big enough for you to be able to get them far enough away from the walls, so they have to be right up against the front wall. That way, the SBIR dip moves up into the lower mid range, where it is a little less objectionable, and can be damped better with insulation behind the speaker.

Also, you have the speakers laying on their sides, which would be OK for a large room where you can have them a long distance away from you, but for such a small room, that's not a good arrangement. You are way to close to them to use that arrangement. As you move your head left-to-right, and front-to-back, your ears move closer-to / further-from the tweeters and woofers: your ears will always be at different distances from them, and therefore different timing. You cannot get a good stereo image like that, and your sound stage will be greatly skewed and distorted.

Also, your listening position seems to be in the wrong place. That needs to be fixed too.

So here's what you need to do get your geometry correct:

Set up your speakers standing vertically, with the tweeter above the woofer, such that the acoustic axis of the spekaer is 1.2m above the floor. The acoustic axis is the point from which the sound seems to emit on the front panel, and is on an imaginary line that joins the center of the woofer to the center of the tweeter, much closer to the tweeter. Your speaker manufacturer can probably tell you where that is. That's the point that must be 120 cm above the floor. You could put it a little higher, maybe 130 cm, to help reduce the artifacts from your desk, but no more than that.

Set the speaker stands so that the acoustic axis is 70cm from the side walls (so the speakers will be 113 cm apart), and so that the rear edge of the speaker cabinet is just touching the front wall (or maybe leave a gap of a few mm). Rotate the speaker inwards to an angle of about 25°.

Set up your chair so that your ears are 135 cm from the front wall when you are mixing normally, seated comfortably. If necessary, move the desk so you can work easily in that position.

Now set up a mic stand so that the tip is at the same height as your ears, and is located about 30cm behind your head, on the center line of the room. Adjust the angles of the speakers a little bit so that they are both pointing exactly at that spot (once again, I'm talking about the acoustic axis of the speakers: imagine a line poking out the front of the speaker from that point, perpendicular to the front panel. That's the line that must point at the tip of the mic stand.

That gives you roughly the right geometry for your room. Do a REW test with your measurement mic located exactly where the middle of your head will be while mixing, and post that. Do a listening test with music you know well, and move your head a bit forwards and backwards to see if there is a better spot. Adjust the speakers a little side to side (small steps!) to see if there is a better spot for them (keep them symmetrical, of course). When you find the best spot, do the trick with the mic stand again, so the speakers are angled correctly to point at that spot about 30cm behind your head. Do another REW test like that.

For both REW tests, calibrate to 80 dB for each individual speaker, so the combined level will automatically be about 86 dB. Do a set of three tests, every time: Left-only, Right-only, and Left+Right. Label them! :)

Let's take a look at those before deciding on treatment.


- Stuart -
Post Reply