John Sayers' Recording Studio Design Forum Archive

Hey, first time posting here!

My parents let me have one room in their house before they demolish the whole thing and build a new house in a year or so. I have already done some basic treatment in the room but I'm not sure how to go on. Anything I spend money on here should be something I can take away with me when they deconstruct the house.

Room dimensions: 4,1 m x 4,3 m x 2,46 m. On one of the shorter walls there are cupboards along the whole room filled with my parents' stuff which kind of makes the 4,1 m dimension 3,57m. They don't need to access the stuff but it's going to be there because it also provides some thermal insulation during the winter, according to my mum... There is a window on the other short wall and a large door with windows on the other. The walls are drywall I think. On the back the cupboard doors are probably particle board.

My near fields (Genelec 8040) are currently placed on the long wall as I think it works a bit better this way. I tried it the other way too but I think the low end improved a bit in this configuration.

I have constructed four 200mm deep broadband absorbers from floor to ceiling in all the four corners. Then there are nine 100mm deep absorbers. All of them are Paroc COS5 which I think is something like the OC 703...

I tried to learn SketchUp quickly to provide some kind of a drawing of the room, I hope it's okay. Then there is a REW plot.

I'm really grateful for any tips how to improve it.

EDIT: I forgot to mention I'm mostly going to use the room to produce and mix electronic music (psytrance to be exact)

Hi "tommiot", and Welcome!

which kind of makes the 4,1 m dimension 3,57m.

That puts the rear surface way too close to your head, and probably places your head too close to the middle of the room. You need to rotate your orientation so that you are facing one of the short walls and have the speakers aiming down the longest axis of the room. That allows you to have your head in the correct location, and also have as much distance as possible between your head and the rear wall.

My near fields (Genelec 8040) are currently placed on the long wall as I think it works a bit better this way.

It might "sound" better like that... before you treat the room! But it won't be better until you have everything set up correctly.

Like this:
First, put a large sheet of thick plywood (or OSB, or MDF) in front of the "stuff" on the long wall, so that you have a hard, solid, rigid,massive surface on that side, the same as on the other side. Symmetry is very important. Use 19mm plywood for that, which should make the room about 3.55m wide.
Put your speakers on stands such that the acoustic axis is 1.2m above the floor (check the Genelec web site for the location of the acoustic axis on your speakers). Set them up so that each speaker is 99 cm from the side wall, and the speakers are therefore about 157 apart. Once again, I'm talking about the location of the acoustic axis, not the edges of the cabinet.
Push them tight up against the front wall, then pull them 10cm away. Put 10cm thick panel of insulation in that gap. Something like OC-703.
Set the speaker controls to flat response, except for the bass roll-off, which should be set to -3 dB.
Make sure that you have not other equalization or dynamics in the signal path.
Set up a mic stand, completely vertical, to mark the mix position. Place it about 190 cm from the front wall, on the room center line. That marks the location where the axes of the two speakers need to intersect, which is about 30 cm behind where your head will be (160cm from front wall).
Aim both speakers so that they are pointing exactly at that mic stand. Now set up your chair so that the center of your head will be 30 cm in front of the mic stand.

That's the correct initial setup for your room. After you have treated the room, you will try some small variations on that, but that's the starting point.

I tried it the other way too but I think the low end improved a bit in this configuration.

Probably because your head was in a null for some of the modes... That's not where your head should be!

All of them are Paroc COS5

Do you have a link to the acoustic test data for that product?

Then there is a REW plot.

But it shows only frequency response! That's not much use... we need the entire MDAT file. Upload it to dropbox (or similar), and post the link here.

That said, there's clearly a major symmetry problem with your current setup: the Left and Right traces are very different. The "average" trace is no use at all: instead, do a third test with BOTh speakers turned on, but do not change any adjustments with your system in between. All measurements must be done at the same settings.

Also, you only ran the test from 30 Hz to 2 kHz. That's no good. You need to do the entire spectrum, plus a bit more. Run your tests from 17 Hz to 22 kHz.

Next, you did your tests at a very, very low level: only 55 dB??? Either that, or you did not calibrate REW correctly. You need to calibrate it using a proper hand-held sound level meter.

And finally, when you run the next set of tests, do them all with the mic point upwards at about 60°.

Basic treatment: Your four corner absorbers are fine, assuming they are made using suitable insulation.
You need the panels that I already mentioned between the speakers and the front wall.
You need similar panels on the side walls and the ceiling at the first reflection points.
And you need lots of thick treatment on the rear wall, covering as much of the wall as possible.

Do two sets of tests with REW: One with no treatment in the room, the other will all the treatment in the room. Both tests must be done as a set of 3: one with just the left speaker, one with just the right speaker, and one with both.

Post the MDAT file that has all those tests in it, as above, so we can download and analyze it.

- Stuart -

Thank you for being so kind and helpful!

I did know the recommended orientation is to place the speakers on the short wall, but after trying it with the first reflections treated (on the walls and ceiling and back of the room) and with the corner absorbers, I was having trouble with low frequency response, which I think got a bit better when I turned the room around. But I'm not so sure anymore so maybe I will turn the room around once more and have the cupboards on the right side of me.

Then I will be facing the window and on the back wall there is the door. I have two 100mm thick absorbers standing on feet, so that I can move them in place after entering the room. I can try to fit more absorbers there but it's a bit difficult.

Do you think having the plywood is really important? It's 150 EUR to get three sheets of 2440x1220x18 mm plywood to cover the whole wall, but the cupboard doors are already quite rigid and the whole wall is flat because the doors are from floor to ceiling. I could also get some 4 meter long slabs and screw all the doors shut to make the whole surface a bit more rigid. I will add some photos...

The corner where the right speaker is in this setup is a problem. There was a very bad 110 Hz dip even with the monitors on the window side. I tried absorbers behind speakers but they didn't affect the response that much and I ran out of wool.

My measurements were only up to 2k because this forum wouldn't let me upload pictures with high enough resolution to make any sense from the 20kHz graph But I will make the measurements and post the whole files. Also I was more concerned with low frequency response, as the music I produce is quite focused on low frequencies...

I measured the room with a DBX RTA-M microphone, M-Audio Octane preamp to RME ADI-2 ADC. I don't have an SPL meter but I didn't think it was that relevant. I did the measurements at the same level I listen to when mixing.

I will try your suggested initial setup and post the requested files when I get back to the room, probably on Monday.

All of them are Paroc COS5

Do you have a link to the acoustic test data for that product?

No, sorry. The wool density is 70kg/m^3 and I think COS5 is very commonly used instead of 703 around here.

Do you have a link to the acoustic test data for that product?

No, sorry. The wool density is 70kg/m^3 and I think COS5 is very commonly used instead of 703 around here.

It is not equivalent, and to be honest, isn't even similar. 703 is fiberglass based, this stuff is mineral wool. No direct comparison.

70 kg/m3 is too high for good bass absorption. If you are using mineral wool optimum density is 48 kg/m3. The higher you go, the worse it gets for low frequencies. Lower is better. You could go down to about 40 kg/m3, or even 35 kg/m3.

If you are using fiberglass insulation, then the density should be about 30 kg/m3.

In fact, it isn't really the density that matters, but rather something called "gas flow resistivity" or GFR. that's the parameter you need to know, but on the website for the product you mentioned, in the specification for gas flow resistivity it says "NPD". I assume that means "No Performance Data", which implies that they don't have a clue, since they never measured it!

I would not use a product for which there is no acoustic data.

What other similar products can you get where you live? Take a trip to your local "Home Depot" equivalent, and see what they have in stock. Note down the type, brand, and product code, then see what you can find on-line about each.

will turn the room around once more and have the cupboards on the right side of me.

That would be the smart thing to do!

which I think got a bit better when I turned the room around

Did you have the speakers set up correctly in both cases? With the speakers in the CORRECT locations, and the mix position in the CORRECT location, and the angles and distances all set correctly? If not, then I'm not surprised you had problems. The dimensions, distances and angles I gave you above are acoustically optimal for that room. If you used anything that is very different from that, you'll have problems.

I have two 100mm thick absorbers standing on feet, so that I can move them in place after entering the room. I can try to fit more absorbers there but it's a bit difficult.

100mm is not thick enough. It needs to be 200mm thick, and with a 200mm air gap behind it. Anything less will not work effectively down to low enough frequencies.

Do you think having the plywood is really important? It's 150 EUR to get three sheets of 2440x1220x18 mm plywood to cover the whole wall,

Try it first without any plywood, to see if you are achieving acoustic symmetry. If not, then put one sheet up covering just the front part of that closet. If that still does not get you enough symmetry, then add a second sheet next to that. I don't think you'd need all three.

The corner where the right speaker is in this setup is a problem. There was a very bad 110 Hz dip even with the monitors on the window side

That's not due to the speaker or the location of the speaker in the room. It is due to the room itself, or a reflection in the room, or a resonance, or the lack of treatment. You cannot judge if a particular setup is working correctly unless you do ALL of it. Only doing part of the complete setup wont' tell you if it is working or not. That's like trying out a new recipe for cheeseburgers, then only using the bread and the tomatoes to start with, but no meat, lettuce, cheese or pickles! You wont' know if the recipe is any good until you make the complete burger and try it...

I tried absorbers behind speakers but they didn't affect the response

How do you know that? Do you have the REW graphs to demonstrate that it didn't work? What parameters were you looking for? Did you have the speakers tight up against the insulation, and positioned at the correct locations (as I defined them above), and at the correct angles? The reason for that treatment is to deal with the SBIR issues caused by having the speaker in that specific location, as well as to reduce comb filtering from the edge diffraction where the mids and low wrap around behind the speaker itself. Since you did not yet have room symmetry, you would not have been able to hear much of how that treatment was working, or not working. It is related to image stability, sound stage, and the sweet spot.

Once again, you can't test the hamburger recipe unless you cook it using all the ingredients. Saying that you tried insulation behind the speakers8 without doing the rest, but didn't hear any improvement is like saying that you tried the tomatoes on the bread all alone but it didn't taste like a cheeseburger...

My measurements were only up to 2k because this forum wouldn't let me upload pictures with high enough resolution to make any sense from the 20kHz graph

That doesn't make much sense! With REW, the way you do the test and the way you show the results are two entirely different things. You can run the test from 17 Hz to 22kHz but then only show the graph from 20 Hz to 500 Hz if you want. There's no relationship between the testing resolution and the display resolution.

Also I was more concerned with low frequency response, as the music I produce is quite focused on low frequencies...

Cheeseburgers again... "I only ate the lettuce and cheese, because the burgers I want to make are known for the taste of lettuce and cheese"...

I measured the room with a DBX RTA-M microphone, M-Audio Octane preamp to RME ADI-2 ADC.

Excellent.

I don't have an SPL meter but I didn't think it was that relevant. I did the measurements at the same level I listen to when mixing.

Cheeseburgers: "I didn't actually cook the meat, because I didn't think it was relevant to the cheese and lettuce flavor I want in my cheeseburgers!"...

Your ears do not have the necessary precision to judge what 86 dB is. The human ear/brain is not able to determine objective levels of sound, only subjective levels. You NEED to do the tests at 86 dB because that's the standard test level! Not the level that you like to listen at. The reason you need that level is because anything less is probably not going to trigger all the resonant modes in your room, and it is also the recognized level for calibrating and testing audio systems, all around the world. If you go to a movie house, you can bet it was calibrated at 86 dB. If you wanted your mixes to sound the same in that movie house as they do in your room, then your room needs to be calibrated at 86 dB as well. It doesn't matter what level you normally mix at, or listen at: what matters is that the room was calibrated at the correct level.

If you want people on the internet to help you, then you have to be sure you are providing data that allows that to happen. If you post REW data that says it was tested at 55 dB but the actual level in the room was 75 dB, then that isn't valid: what I see on the graph is not what happened in your room. So any advice I give you based on that would be totally wrong, and would not work. That would be like me telling you to grill your burger at 350 degrees for four minutes, but then you not having any way to measure the temperature of your grill, and no way of measuring time, so you just "guess" at both! Your burger will either be raw and uncooked, or it will be overcooked and dry.

You need a sound level meter, and you need to calibrate REW correctly. If you don't measure the actual level in your room and then tell REW how loud it is, REW has no way of knowing.

I will try your suggested initial setup and post the requested files when I get back to the room, probably on Monday.

Great! I'm looking forward to seeing that. Don't forget to do two sets of measurements: one with the room empty (no treatment inside) and one with it treated. You could also do other tests along the way if you want, at each stage of the treatment. That way, you can see what each device does to the room, and you can see the incremental improvements.


- Stuart -

I'll reply to your post later, but I just happened to find the GFR at some ancient Gearslutz thread, referring to some Paroc document which is unavailable now. It's stated that the GFR for Paroc COS5 is 22000 Pa*s/m^2 and OC 703 somewhere between 16 000 and 24 000 depending on the document.

It's stated that the GFR for Paroc COS5 is 22000 Pa*s/m^2 and OC 703 somewhere between 16 000 and 24 000 depending on the document.

Great! OC themselves say that 703 is about 16,000 rayls. Their 705 is a bit higher around 24,000, so that might be the reason for the incorrect higher number.

But 22,000 rayls is on the high side for good bass trapping. I'd go with somthing lower than that. That would be fine for mids and highs,but not optimal for lows.

- Stuart -

Soundman2020 wrote:

It's stated that the GFR for Paroc COS5 is 22000 Pa*s/m^2 and OC 703 somewhere between 16 000 and 24 000 depending on the document.

Great! OC themselves say that 703 is about 16,000 rayls. Their 705 is a bit higher around 24,000, so that might be the reason for the incorrect higher number.

But 22,000 rayls is on the high side for good bass trapping. I'd go with somthing lower than that. That would be fine for mids and highs,but not optimal for lows.

- Stuart -

I found another wool that is actually quite a lot cheaper but they come in a bit smaller size per slab (565x1170mm vs 600x1200mm I have in my COS5 absorbers).

Paroc Extra
density 27kg/m^3
GFR: 8300 sPa/m^2 (this is calculated from the airflow permeability coeficcient that is given it the datasheet, which is the reciprocal of resistivity)



Now already have quite a lot of the COS5 because I was under the impression that it's good for absorbing bass in the corners. I can buy the Extra if it is a lot better for absorbing bass, but I'd still like to make some use of the 200mm COS5 absorbers I already have in my corners. Would you recommend stacking e.g. superchunks of Parox Extra to the corners and having the 200mm panels as rear wall absorbers? I'd probably have to leave like 500mm room between the back wall (or the door on the back wall) to be able to enter the room.

To shortly reply to your previous post: I did try to correctly place the speakers in my earlier measurements, I'm not sure if it was exactly like your guidelines. I placed the listening position to the 38% position in the front of the room and had the speakers in an equilateral triangle, trying different distances from the speaker to the wall. When I had the absorbers behind the speakers the speakers weren't firmly pushed against the absorbers.

I will try to get my hands on an SPL-meter but I can't promise. Sorry the dB scale was off, I was under the impression that only the relative values in the graph matter as long as the measurement is done at a reasonable level, but of course 50dB looks weird on the graph.

Thank you for your patience.

I did the measurements according to your guide. Monitors on the short wall with the cupboards on the right and the door behind me. Monitors against 100mm absorbers against the wall. Cloud has 130mm gap to ceiling. First reflection absorbers have 70mm gap (I can easily increase this if it is necessary)

Here are the results: https://dl.dropboxusercontent.com/u/515 ... 0room.mdat

Now I am having a horrible ringing/echo when I clap my hands. This has to be because there is not any treatment on the parallel walls other than the first reflection points.

Now I am having a horrible ringing/echo when I clap my hands. This has to be because there is not any treatment on the parallel walls other than the first reflection points.

Exactly. That's the purpose of doing the incremental tests, starting with an empty room. You can see each problem as it arises, and treat it accordingly, one at a time. You get to see what is working, what isn't, what needs more treatment, what is over-treated, and deal with it, one at a time. It takes a lot of the guesswork and unnecessary experimentation out of room.tuning, and leads to better results.

I downloaded the REW file, and I'll take a look at it tomorrow for you. It's 2 AM where I am, right now!


- Stuart -

I just took a quick look, and the improvements are very good!

Did you look at your results yet, in REW? I'll post some graphs tomorrow, but the results so far are very promising. The treatment is working pretty much as it should. It's not done yet, but it's getting there. Very nice progress.

And don't worry too much about the flutter: flutter echo is relatively easy to deal with...


- Stuart -

Soundman2020 wrote:I just took a quick look, and the improvements are very good!

Did you look at your results yet, in REW? I'll post some graphs tomorrow, but the results so far are very promising. The treatment is working pretty much as it should. It's not done yet, but it's getting there. Very nice progress.

And don't worry too much about the flutter: flutter echo is relatively easy to deal with...


- Stuart -

Yes I did look at them and now it's a bit better than before! I guess the absorbers behind monitors did work after all, I just didn't have my monitors placed close enough to the wall before, because everything else was pretty much like they are now (in my first setup before I changed the orientation to face the longer wall).

The flutter was significantly reduced in my earlier setup by just placing two absorbers on the opposite walls, but what kind of treatment would you recommend?

What kind of absorbers would be optimal behind the monitors? I have COS5 there now but I think 600x600mm panels would be large enough so I could build new ones and have the current ones on other walls.

And would you recommend replacing two corner traps with Paroc Extra superchunks and having the 200mm panels as back wall treatment instead like I mentioned earlier?

Looking forward to hearing how you would improve the room! I'll try to make somekind of a sketch how the room is now, but I'm a bit busy today.

EDIT: One thing I forgot to mention. I also have a desk in the room in front of the monitors. Just a flat IKEA Galant table made of MDF. Speakers are standing on cinder blocks (with bitumen damping between the blocks) behind the table.

Here is what the room looks like now. And another file with the superchunks, but I'm not sure how to arrange the back wall treatment because I need to be able to enter the room through the doorway