John Sayers' Recording Studio Design Forum Archive

Hi everybody,

this is my first post, so I'm hoping to do it right.
I'm a professional musician and have worked as a recording engineer for a while now.
I built my own studio a couple of years ago - and that's where the problem starts: I didn't do it right.
Lacking knowledge, I did not make any measurements in the room but instead started building right away. I read some stuff about studio acoustics but I guess I was to young and not patient enough.
Right now after having worked at other places for a while, I want to start over and improve the (bad) acoustics in the room.

Let me talk you through:
The room is 6m long and 3,20m wide. What I did first, was to construct a second (mainly wooden) wall to soffit mount my speakers (Adam Audio P33 Midfields). You can see that in the drawing below and on the pictures.
What is not shown in the drawing (but on the pictures) is that the top part of the speaker wall is angled and - maybe more importantly - that this wooden wall doesn't go all the way up to the ceiling but only up to about 2,50m. The black stuff above it which you can see on the pictures is just cloth. I hope I explained that right...
The speakers are mounted in the angled part of the wall at about 2m.

The ceiling:
The ceiling is at a variable height (sorry, I don't know how to say that in english). At the entrance of the room it's 3,85m and on the other end it's 3,55m high. I made a large absorber with lighting in it, as you can see on the pictures.

More absorbers...:
I constructed two bass traps which are located in the back corners of the room. They are about 110 x 70 x 125 cm.
Above them there are two more broadband absorbers in each corner.
There is two more pictures attached to the post.

Now for the bad part: The sound is horrible. I have found out by listening and playing with the eq that where I'm sitting (the circle in the drawing) I have a major problem at about 80 Hz - it's just completly not there! And after that - the ringing mids take over. I found out again by listening that my main problems are about 100, 135 and 170 Hz.
After reading a lot in this great forum, I decided to have a look at RWE. I did a first measurement today (I could not attach it - how can I supply it to you?).
As I didn't have any experience with the program nor an SPL meter, the measurement might not by to accurate, but I could see already, that the frequency problems are exactly where I found them by listening.

Now you know my situation. Now for the simple question: Should I get rid of everything and start over, or do you see a way of treating the room?
What I'm mainly interested in is: What did I do wrong (or better - did I do anything right)?

Any help or suggestion would be much appreciated! Thanks so much in advance!

johannes

Hi
I'm new here myself and must admit my knowledge is pretty limited, However i have learn from my own mistakes in the past. One of your main problems you stated was a low-end issue.
Maybe getting a couple of bass traps behind the wooden 'wall' where your monitors are could help. I had a similar problem but with a rehearsal space. The bass was booming out of control and was making practicing in there horrible. To solve the problem i made 2 rock wool boards the same height as the room and around 350mm thick. These were then angled behind the speakers (with space behind the panels as well). This worked pretty well, I'd originally only put two bass traps at the far end, the same as you have in the room plan.
Also some diffusion panels may help. There is plenty of stuff on diy diffusion panels that could help on the internet. this link is a calculator that calculates what size and pattern would help at certain frequencies.
http://www.mh-audio.nl/user/acoustic%20calculator.asp
Maybe try some affordable methods such as these, see if it helps and then look into a complete over haul or further modifications.
Hope this is useful
Jord.

Hi there Johannaes, and Welcome!

The room is 6m long and 3,20m wide. ... ceiling ... 3,85m and on the other end it's 3,55m high

Your room ratio is 1 : 1.1 : 1.87. That's very long and thin. I would consider trying to improve that.

What I did first, was to construct a second (mainly wooden) wall to soffit mount my speakers

How are those built? In terms of internal structure and materials used.

The speakers are mounted in the angled part of the wall at about 2m.

That is far too high! you don't say how much that part of the wall is angled, but anything more than about 7° is a major problem. Personally, I have never designed a studio that tilts the speakers more than 5°. High tilt angles produce a whole series or detrimental acoustic issues.

The correct height for speakers is 1.2m above the floor, and that refers to the height of the acoustic axis of the speakers, which is generally close to the tweeter (the manufacturer can tell you exactly where the acoustic axis is on your speakers). So the bottom of the speaker cabinet will be a lot lower. 1.2 m is the correct height, because that is the height of your ears when seated, for most people. The acoustic axes of the speakers should be at the same height as your ears, or maybe just a little higher. But 2m is way too high.

I constructed two bass traps which are located in the back corners of the room. They are about 110 x 70 x 125 cm

Those are very small. They should be the complete height of the room, from floor to ceiling, and built using the "Superchunk" style, for maximum effectiveness.

You show your listening position as 3.41m from the rear wall, but that is too far back. It's hard to calculate where it should be, since we don't know the details of the "soffit" wall construction, but I would estimate about 3.5m to 3.6m from the rear wall.

You also show your speakers in the corners, angled at 31° inwards, and apparently aiming directly at your head, which is also wrong. In that room, the speakers should 90cm away from the side walls (which makes them 140cm apart). Once again, I'm talking about the acoustic center od the speaker, not the side of the cabinet.

You also show some type of perforated panel trap mounted on the left soffit but nothing on the right soffit, so the room is asymmetrical, and you'll never be able to mix well in there. What is the purpose of that device, and what frequency is it tuned to?

The sound is horrible. I have found out by listening and playing with the eq that where I'm sitting

Download REW (Room EQ Wizard) from the Home Theater Shack website, and use that to do a test on your room. You do need to calibrate it using a good quality hand-held sound level meter, and you do need to do the tests at 85 dB (not 75, as stated in the manual). Do three tests, all of them with your measurement mic exactly at the currently listenign position, and exactly on the room center line. THe first test is with just the left speaker on, the second with just the right speaker on, and the third is with both speakers on. Make sure that there is no equalization or compression at all in the signal path. Do not adjust anything at all between the three tests! They must all be done with identical paramaters. Also,l you must not be in the room when the test runs: use the "start delay" option to give you enough time to get out, close the door, and another few seconds so the room can calm down after you are gone.

When you have done that, post the resulting MDAT file here on the forum, so I can download it and analyze it for you.

I did a first measurement today (I could not attach it - how can I supply it to you?).

Assuming you did it correctly (as above), upload it to a file-sharing service (such as DropBox), and post the link here.

but I could see already, that the frequency problems are exactly where I found them by listening.

Frequency problems are only a small part of the the acoustics of that room. Decay times, spectral issues, reflections, comb filtering, and several other issues are just as important, and will be revealed by REW.


You mention major issues at around 80, 100, 135 and 170 Hz. Theory predicts that your 3.0.0 axial mode is at 86 Hz, your 0.0.2 axial mode is at 97 Hz, your 0.2.0 mode is at 107 Hz, your 5.0.0 axial is at 145 Hz, and your 6.0.0 axial is at 172 Hz. I'm betting that you also have issues at 20Hz and 57Hz, but your speakers probably don't go that low, or you didn't notice them. That tells me one thing: you have a major problem with the x.0.0 set of modes, which are your length modes, related to the distance between the front and back walls. You are going to need to do some serious bass trapping to deal with those. You also have issues wit the side-to-side modes (0.x.0) and the up-down modes (0.0.x), but they are not as severe.

Now you know my situation. Now for the simple question: Should I get rid of everything and start over, or do you see a way of treating the room?

To be honest, I suspect that you would be better off starting over, from an empty room. You might be able to treat some of the aspects, but with your speakers mounted so high up and tilted down so steeply, I doubt that you'll ever be able to get good acoustics in there.


- Stuart -

Hello Stuart and Jord,

thank you so much for your answers.
I never read or heard anything about speakers mounted too high and being tilted too much could affect the sound, but it maks sense. Thanks for that!

anything more than about 7° is a major problem.

Damn... I would estimate they are angled at about 20°...
I'm just curious: What is usually the result of speakers being angeld that much?

How are those built? In terms of internal structure and materials used

The soffit wall is made from plywood. There is a wooden frame under it. The lower part (basically the parts on the left and right below the angled part) contain quite large boxes filled with sand on their back. I thought that would eliminate any unwanted vibrations.
As I said, the soffit wall doesn't go all the way up to the ceiling but ends at about 2,50m. Is that a problem?
At least it makes it hard to calculate frequency problems, I guess...?

You also show some type of perforated panel trap mounted on the left soffit but nothing on the right soffit, so the room is asymmetrical, and you'll never be able to mix well in there. What is the purpose of that device, and what frequency is it tuned to?

That was my first try building some kind of mid frequency absorber, but it didn't work quite well. Actually I have two of those. On the right side of the room is the other one usually. When I made the first measurments with REW, I moved one ofthe absorbers to see the difference in the frequency analysis. Actually it didn't seem to change anything at all so forget about those.

Download REW (Room EQ Wizard) from the Home Theater Shack website, and use that to do a test on your room.

I'll get a decent SPL meter next week and will do those measurements.

Thanks again big time!

johannes

Damn... I would estimate they are angled at about 20°...

Not good at all!

I'm just curious: What is usually the result of speakers being angeld that much?

Two major issues. First is the practical geometric issue: with your speakers up high pointed down steeply, you get major big-time reflections off the desk, console, meter bridge, etc. right into your ears, and there's nothing you can do about it. First reflection points in studios are usually treated with deep absorption, but it's har to operate the console or DAW with 6" of fiberglass blanket over it!

Second is psycho-acoustic issues: Your pinnae (the external part of your ear) works with your ear canal and all the rest of the very delicate parts in there to help you detect directionality. It is designed to do that really well with sounds coming from directly in front of you. If the sound comes from much above the horizon, then you can't determine directionality any more! Or rather, you lose accuracy. Our auditory system is very good at figuring the direction that sounds are coming from in the horizontal plane, and pretty bad at figuring it out in the vertical plane.

There are also other lesser issues with psycho-acoustic effects, such as comb filtering, phase shifts, timing differences, etc. that are also not so nice, but those are the two biggies.

You sometimes do see mains mounted high in large studios, but those are for the CLIENT, sitting at the back, not the engineer, sitting up front, when he is making critical decisions. For the client, the angle of those speakers is much lower, because he is sitting much further back...

The lower part (basically the parts on the left and right below the angled part) contain quite large boxes filled with sand on their back. I thought that would eliminate any unwanted vibrations.

It is probably doing some damping of the front baffle, yes, but at the expense of not using that very valuable space for bass trapping. Soffits are normally built with very thick, massive, rigid front baffles, and they are braced firmly with rigid framing, then damped on the inside with isnulation. That leaves the space below the speaker for bass trapping.

As I said, the soffit wall doesn't go all the way up to the ceiling but ends at about 2,50m. Is that a problem?

Not really. That's OK. You could use the space above it for bass trapping, but extending it higher would be beneficial.

At least it makes it hard to calculate frequency problems, I guess...?

Actually, don't rely too much on calculations. In a room that is already built, it's a lot easier (and far more accurate) to just run an acoustic test on the room, and see exactly where the problems are! Use REW for that...

When I made the first measurments with REW,

Upload your MDAT file someplace like Dropbox, and post the link here, so we can download it and take a look.

I moved one ofthe absorbers to see the difference in the frequency analysis. Actually it didn't seem to change anything at all so forget about those.

Tuned mid-range traps like that probably won't change the FR curves much, but they aren't supposed to: You should be looking for effects in the time domain, not the frequency domain.

That said, the trap is rather small and in the wrong place to be effective.

I'll get a decent SPL meter next week and will do those measurements.

When you get that, use it to calibrate REW, and run a full test, very carefully, then post your results as above.


- Stuart -

Upload your MDAT file someplace like Dropbox, and post the link here, so we can download it and take a look.

Ok, here's a link to the first measurement, which I did without proper calibration (lacking a SPL meter):
http://www.joe-h.de/nicktight_noncal.zip (my own server - I'm not on dropbox)

I'll get a SPL meter next week and will do another measurement with calibration then.

They (the bass traps) should be the complete height of the room, from floor to ceiling, and built using the "Superchunk" style, for maximum effectiveness.

I'll start with that today. What should be the size of those? By the way - happy new year!

johannes

Ok, here's the new update.
I did construct the superchunk bass traps. They are now in both corners of the back side, from floor to ceiling. I then moved the speakers and changed my listening position as you described.
The speakers are now on stands, 90cm away from the wall and 1,2m high.
My new listening position is now 3,55m from the back side of the room.

I then got my SPL meter, finally, so I could do the REW measurement at 85 dB as you suggested.
You can download it here: http://www.joe-h.de/calib_L-R-LR.mdat

Some things changed. The two major problems I see are 80 and 135 Hz.
I used the signal generator in REW two hear just 80 Hz and as I moved around in the room, the level changed a lot, depending on where I stood. But my listening position is the worst spot by far... You almost cannot hear anything at 80 Hz there.

Is there something I can do about that?
As I have plenty of "invisible" room behind the old soffit wall, I thought I could just do some more bass trapping there. Would that be a good idea?
And what about the 135 Hz?

Thanks for any suggestions!

johannes

I did construct the superchunk bass traps. They are now in both corners of the back side, from floor to ceiling.

... and you can see the improvement in the low frequency response and time-domain response!

Here's the decay plot from before:
And how it is now:
You can clearly see the improvement. The low end is much smoother.


The same for you waterfall plot, showing how the sound decays over time. Before:
And now:
Decay times have improved, and once again they are more even, and smoother

And finally, frequency response for the low end of the spectrum. Before:
And now:
There too the improvement in the low end is very clear.

You should definitely be hearing a big difference: the bass should be cleaner and "tighter" now, with better definition and less ringing. Can you hear that when you play music in there?

However, there's something wrong with the way you are doing your tests:

First, why are you only running them from 30 Hz to 8 kHz???? We need the entire spectrum! In the future, set REW to run the test from 15 Hz all the way up to 22 kHz, even if your speakers can't handle those frequencies very well.

Second, I do not understand why your measurements taken with both speakers on, are not louder than with just one speaker on! Did you change some setting on your system between the tests? The test for both speakers should show a level about 6 dB higher than for just the left speaker or just the right speaker. You must not change anything in between tests! After you have calibrated REW, never, ever change any setting: keep everything the same for all the tests that you do from then on, and keep a note of what settings you have, so you can do all future tests with the exact same levels.

I then moved the speakers and changed my listening position as you described.
The speakers are now on stands, 90cm away from the wall and 1,2m high.
My new listening position is now 3,55m from the back side of the room.

Excellent! Did you notice the improvement in the sound?

Some things changed. The two major problems I see are 80 and 135 Hz.

Those have always been there:
As you can see, then problems at 79 Hz and 149 Hz have been there all the time: that did not change. However, the modal ringing at around 100 Hz, 131 Hz, 171 Hz, and 190 Hz are all greatly reduced, and the dip at 121 Hz is gone.

I used the signal generator in REW two hear just 80 Hz and as I moved around in the room, the level changed a lot, depending on where I stood.

Right. That's because you are moving into different parts of the standing wave. At some places in the room the standing wave is close to its peak value, while at other places it is close to its null value.

OK, here's the problem: You said that your room is 6m long, but a major mode at 79 Hz does not match 6m! It matches a room that is quite a bit longer that, somewhere around 6.5 m long. If that's the case, then we should also be seeing problems at about 52 Hz... and yes, we are. It is clear in the graphs. 52 Hz is the 2.0.0 mode, and 79 Hz it the 3.0.0 mode. There are also certainly problems at 26 Hz,... but we can't see those, because you didn't run the test low enough!!! So I am guessing that there is about 50 cm or so behind the soffits, up the real wall of the room. Is that correct? If you were to measure the room without the soffits, would it be about 6.5 m long? If so, move your listening position forwards, so that it is about 4.1m from the rear wall. That should be a better location: it will be out of the modal null for the axial length modes.

As I have plenty of "invisible" room behind the old soffit wall, I thought I could just do some more bass trapping there. Would that be a good idea?

Yes, but to be very honest, I think you will need to re-build those soffits correctly. Based on the REW data, it looks like they are not build properly, and you are not getting the proper effect.

My recommendation would be to design new soffits, properly, take down the old ones, and re-use the materials to build the new ones.



- Stuart -

Hi Stuart,

You should definitely be hearing a big difference: the bass should be cleaner and "tighter" now, with better definition and less ringing. Can you hear that when you play music in there?

I absolutely do! Bass is much better and tighter now.

Second, I do not understand why your measurements taken with both speakers on, are not louder than with just one speaker on! Did you change some setting on your system between the tests?

I think I calibrated the SPL meter again before the test with two speakers on... my bad.
I'll do another "fullrange" measurement without adjusting anything between the single tests.

You said that your room is 6m long, but a major mode at 79 Hz does not match 6m! It matches a room that is quite a bit longer that, somewhere around 6.5 m long.

Let's get to the "the mystery of 79Hz":
The original room is 6m long - without the soffit wall (which, as I said, doesn't go all the way up to the ceiling).
You can see that in the drawing in my first post. The length from the soffit wall to the back of the room is actually 5,32m!
As I moved around in the room, listening to the 79Hz sine, I noticed big differences depending on where I stood - I said that in my previous post. I noticed today that the levels are also quite different depending on the height! Wouldn't that mean, that we are talking about a tangential or even oblique mode?

Last weekend, I did a lot more measurements and tests. I tried several listening and speaker positions.
I achieved the best sound so far (without any major frequency drops!) as I sat at 3,1m (from the rear wall)! The speakers were about 1,65m from my listening position - which in theory should be wrong...
I guess I just moved away from the null at 79Hz and 135Hz. There still is a little drop at 79HZ, but way not as severe as before.

The original room is 6m long -

OK, but that still does not explain why were are seeing a mode that corresponds to a room that is 6.5m long! For a room that is 6m long, we should be seeing a mode at 86 Hz, not 79 Hz. (It's the 3,0,0 mode, by the way).

For a room that is 6m long, there are no modes at all in the vicinity of 79 Hz. That's the mystery. Why does the room seem to be 6.5m long, acoustically, when it is only 6m long, physically? Is there maybe a closet at the back, that is 50cm deep? Or a very thin wall with a passage on the other side that is 50cm wide? There must be some reason why that mode is there.

Unless it is not a mode.... It is possible that something in the room is resonant at 79 Hz... Is there some type of resonant cavity "thing" in the room, that might have a natural resonance at 70 Hz?

Wouldn't that mean, that we are talking about a tangential or even oblique mode?

Possibly, but its unusual to have such a powerful oblique mode at such a low frequency. It's even overly powerful for a tangential mode. If it is tangential, then it would have to be the 2,1,0 mode, and if it is oblique, then it would have to be the 1,1,1 mode.

Do you have any mineral wool or fiberglass insulation on hand? Or acoustic foam? Lots of it? If so, then spread it over the floor, several inches deep, do one REW test, and we'll take a look at the results. If it is oblique, that would probably damp it out of existence, so we'd see a large change. If there is only a small change that would rule out oblique, and leave us with either axial (room length) or tangential: To rule out tangential, set up that same insulation against the side walls, thick, on both sides of the room. Once again, if there is large change in that mode, then that would suggest it is the tangential, but if there is only a small change, we are stuck with the conclusion that it is axial.

But whatever the source of that mode is, I'd still suggest that the best idea would be to re-build your soffits correctly: you might just be chasing a ghost that is due to some strange issue with the soffit, rather than with the room...


- Stuart -

OK, but that still does not explain why were are seeing a mode that corresponds to a room that is 6.5m long! For a room that is 6m long, we should be seeing a mode at 86 Hz, not 79 Hz.

I know... It's very strange. Behind the room there's just the hallway, but that is about 1,50m wide.

you might just be chasing a ghost that is due to some strange issue with the soffit, rather than with the room...

You're absolutely right about "chasing ghosts".

Unfortunately my new SPL meter broke (two days after purchasing it). I jut got it back yesterday, so I hope next week I'll find the time to do the measurement right (fullrange with same settings over the three tests).

Thanks so much for your help!

johannes