Please help me tune my control room….please.

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

Moderators: Aaronw, sharward

Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Please help me tune my control room….please.

Post by Soundman2020 »

I've read where lots of people will mount general store "peg board" (the perforated fiber board you'd use for a tool rack) to the front of their 703 panels. A porous membrane absorber, right?
Actually, it's a "perforated panel tuned resonator". The term "porous absorber" applies to things like mineral wool, fiberglass, acoustic foam, etc. It is very broadband, not tuned. And a "membrane absorber" or more accurately a "membrane trap" is a tuned device where the surface density and dimensions of the panel and the depth of the sealed air gap behind set the frequency. There's also a "limp membrane" absorber which is vaguely similar, can also be tuned, but the front "panel" really is a membrane, in the sense that it is not rigid at all. Things like tar paper (roofing felt), and MLV, and some types of plastic can be use to make limp membrane traps.

But perf panel ("perforated panel") is different: it is a Helmholtz resonator, and is tuned. Mathematically, it is pretty much the same as a slot wall, and works on exactly the same principle. The "slugs" of air in the holes (slots) vibrate in sympathy with sound waves of the same frequency, and they are tuned by the size of the hole (width of slat), thickness of the panel (thickness of slat),, spacing between the holes (height of the slat), and the depth of the sealed air cavity behind (ditto). If all of the holes are the same size, and the same spacing, over the same cavity depth, then the whole things is tightly tuned to one very specific frequency. If you put porous absorber inside the cavity, directly behind the front panel and in contact with it, then it is a tightly tuned absorber for that specific frequency, with very high Q. That would be the case with pegboard over a panel of 703, for example.

On the other hand, if the hole size or spacing varies, or the depth of the cavity varies, then it acts more like a broad band absorber that covers a whole range of frequencies evenly, with a low Q... provided that the open area of the panel is greater than about 10%. If the open area is less than about 5%, then each hole acts on it's own, tuned to it's own frequency, so the device acts more like an array of small Helmholtz resonators, each tuned to it's own specific frequency, once again with a reasonably high Q.

So you can do that in many different ways, for different purposes, by making the holes, spacing and depth in different ways. You can even get micro-perforated membranes now, where the holes are really, really tiny, and the membrane itself is very thin, but when installed correctly they can be very effective. One advantage is that micro-perf panel can be transparent plastic, so it can be placed in front of windows: you can have a see-through tuned absorber!

But anyway, to answer the question: ".... Mmmmmm... maybe." Yes, you can put pegboard over a sealed cavity with insulation inside and get acoustic absorption. But once again, it ain't that simple! If you did that, would it be tuned to a useful frequency? Or might it end up sucking out frequencies that were already weak, while not touching the ones that actually needed it? And since perf-panel is mostly solid wood with only a small percentage of holes, in addition to absorbing the tuned frequencies, it reflects most others right back at you, except at the low end, which it allows through... If the holes / slots are large enough, then it can also provide some diffusion (slot walls do, for example).

So that's the situation. Pegboard night or might not be useful in your specific room. I'd suggest finding out the characteristics of the pegboard that you can get in your area, in terms of hole size, hole spacing (center to center) and panel thickness, then we can look at the REW data and see if there might be some way to use that.

Here's a perf-panel module under construction that I designed for one of my customers last year. It's part of a series of slightly different panels that went on the walls of the control room, each tuned slightly differently to treat certain small problems in the overall room response:
Perf-Panel-Broadband-RDMOUS-01.jpg
You can see the different hole sizes and spacings that I used to cover the range I wanted.
I had thought of using this method of cutting certain sized holes into hard faced panels to help tune while providing some broadband absorption but I didn't get very far because I couldn't find reliable info. Care to weigh in?
Well, yep, it can be done! But as usual with acoustics, it ain't as easy as it looks on paper! For example, Helmholtz resonators are pressure-based devices, not velocity based (like porous absorbers are) so you have to make sure that your device is placed at a pressure peak for the frequency you are wanting to treat. So if you are wanting to treat standing waves of some sort, the you can't just put it any old place that looks nice: it has to be where the pressure component of the wave is at maximum, or close to maximum. You also have to ensure that your device has enough air volume inside to be effective, etc. And you have to use the correct formula! That last comments sounds dumb, but there's an issue here: there are two equations that have been published in books for calculating Helmholtz resonators in slot walls and other types of perf panel: one of them is wrong; It has a "plus" sign where there should be a "multiplication" sign. It seems that at some point, many years ago, some careless proof reader missed that error, and the book went to print. Then nobody noticed! Other writers of other books then copied the wrong equation into their own books, and the error was propagated. For years, nobody noticed ... :shock: Until a few years back, a very smart, sharp-eyed acoustician by the name of Eric Desart saw the error, started investigating, notified the publishers of the books, as well as the Internet, and slowly the error is being corrected. But if you happen to have an older book, or even a newer one where the publisher didn't get the memo or didn't care, then you'll be using the wrong equation. Eric was a highly respected and very valued member of this forum, and posted on many threads here too over the years. He's also the guy who designed the very best isolated studio on planet Earth (Galaxy Studios, in Belgium). Unfortunately, he passed away earlier this year, and he is greatly missed. But his legacy lives on, and one part of that is his correction of the Helmholtz equations.

So that's a long way of saying: make sure you use the right equation when you are calculating the tuning for your perf panel and slot walls!
From the point of view of someone who knows the program, whats the best way to label the data so it makes it easier to understand?
What I do is to use the name block of each test to give me the most basic data, such as which speakers where on and what point in the series of tests I am looking. Like this:
REW-tabs-sample-2.jpg
I always start the name with a series of three characters to show which speakers were on, and filling in the other characters with a dash "-". So "L--" means only the left, "-R-" means only the right, "--S" means only the sub(s), "LRS" means all three, "LR-" means left and right but not the sub, etc. Then I add a very brief description of the series, such as "baseline" or "superchunks in" or "cloud in", to indicate what was the major change that happened just before this test was done, then I use the comment block at the bottom of each test to note down additional information that might be useful in the future.

That way I can see at a glance what test I am looking at, and I can easily flip back and forth between related tests to see what changed. So for example, if I want to see what effects the cloud had on the overall modal response, I'd probably look at the waterfall plots window, and flip back and forth between the "LRS - Baseline" and "LRS - Cloud in" tests, while concentrating on how the peaks changed, and how the decay patterns changed.

Etc.

(For 5.1 rooms, I use the notation "LCRlrS" to note the speakers that were on, where "C" of course is the center channel, "lower case "l" is the left surround channel and lower case "r" is the right surround channel.)
Thanks again for all the advice and criticism. I'm gonna gather some more data and post tomorrow and we can start sorting things out.
:thu:


- Stuart -
JasonC
Posts: 31
Joined: Thu Aug 15, 2013 1:24 am
Location: Alpharetta, GA
Contact:

Re: Please help me tune my control room….please.

Post by JasonC »

Stuart,
Sorry for the delay. This weekend turned into a marathon as usual.

I did some more tests with speaker placement and with and without racks on the desk. I've linked the mdat here....
http://we.tl/L8aLSWKRbR
Hopefully I was descriptive enough in the titles and descriptions for you to know what is what.
Currently the speakers are only about 4" away from the front wall. I did some tests while moving them further away from the front wall in 4" increments and it just seemed to get worse as I went on so that tells me that they need to be as close to that wall as possible, or in it. As I was doing the tests I saw that moving the speakers 6" in on both sides to be about 60" apart and aiming them more down the room as opposed to angled in at the mix position helped to smooth some response mainly in the upper mids between 500 and 1.5k. When doing the tests with the 4space racks on and off of the desk there is definitely more problems in the same 500-1k range with them on the desk. Makes me want to experiment with getting the speakers higher off the ground and angled down slightly to the mix position. I want to be able to have gear on my desk, of course, so I need to find the compromise. The table that I'm currently using is fairly large and I don't need that much space. I was planning to build my own desk that'll be more efficient in less space. Any tips in that department? Take a look at the data and let me know what you think. Im sure you'll see more than I can.
-- question on REW data... when looking at the SPL&Phase window after taking a measurement. In the lower phase graph where it shows what looks somewhat like a curve opposite to that of the SPL graph (I'm assuming thats the phase graph) where it shows the vertical dotted lines.. Is that where frequencies are phasing out? Standing waves? Since the SPL lines either peak or null at those frequencies that's what I assume?

After doing that round of tests I took all remaining absorbers off the walls ( in the tests linked above the absorption that was up in the room were the clouds, the angled 2" panels that are in the ceiling corners, two 4'x4' 2"703 panels in the rear 1/3rd of the room on the side walls, the corner traps in the rear corners, the two rugs, and the 4" 703 panel that spans almost the whole rear wall, for what its worth) and did another round of tests with the speakers in their new, slimmer width positions starting at the mix position and working away from it in 6" increments to the rear wall. Hopefully this will let you see what my room is doing at a bare-wall level, even though there is still some absorption in the room overhead and in the rear corners. The mdat is linked here...
http://we.tl/jCjsThpdgE
One thing I did notice: Is that carpet on your floor? Or is it something else? Ceramic tile maybe? Please tell me it isn't carpet.... Please...
No, It's not carpet! :wink: The floor is porcelain tile. There is a 6x8 rug at the mix position/desk area, and another rug under the couch at the rear of the room. The rugs are almost exactly the same color as the tile so it looks kinda like carpet. Don't worry. Even though you totally reemed me last year when I told you I was building this room on top of a "kick drum" I took that into account and massed the crap out of this floor. It's about as close to solid concrete as I could get it. Nice and solid.
Before I do the next round of tests while adding treatment back into the room I'd like to hear what you have to say after reviewing some of the data on whether you think this is a good baseline to go from and if you notice anything specific like where I should be hanging my absorbers on the walls based on the data. Right now I just had them fairly evenly spaced on the area of each wall. I tried using the 3D calc that you linked me to to get a rough idea of where absorption is needed for certain frequencies but as we know, with my room shape the basic rectangular calculators don't really apply. And as you said before about diffusion placement not being within 10' of the mix position, I was wondering if using diffusion on the rear wall where I previously had the 4" absorber that spanned the wall might be a better choice. Thoughts?
In all the tests in and near the mix position there's the significant lack of lows between 60-75hz, the significant boost at 106hz, and another dip at about 145hz. Seeing that these are "problem areas" that seem to be consistent is it safe to assume that I need to build traps tuned to these frequencies and have them placed around the mix position such as doing some slat absorbers on both sides of the mix position to try and counteract whats happening there? You talked about absorbing back in standing waves like i have at 60-75hz. Could you maybe elaborate on that or help me nail down a potential solution to those particular frequencies in that particular area of the room?
Definitely seeing light at the end of this tunnel. Thanks again for all your efforts and patience! :thu:

Jason
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Please help me tune my control room….please.

Post by Soundman2020 »

I did some tests while moving them further away from the front wall in 4" increments and it just seemed to get worse as I went on so that tells me that they need to be as close to that wall as possible,
Yup. That's why the recommendation is to get them up against the wall, leaving only a gap that's just big enough to get useful absorption into. As you move them further away, the SBIR issues move further down the spectrum, into the lows, where they are more noticeable and more objectionable.
as close to that wall as possible, or in it.
:lol: :shot: :cop: Yep! There are reasons behind our crazy recommendations! :)
I did some more tests with speaker placement and with and without racks on the desk. I've linked the mdat here....
:shock: I think you need to calibrate your system! Either it needs calibration badly, or the entire studio is now cracked and crumbling from the extreme shockwaves that you caused by doing your tests at an average of over 120 dB!!! Time to get out your sound level meter, and calibrate REW properly, then never change the levels again while you are testing...
As I was doing the tests I saw that moving the speakers 6" in on both sides to be about 60" apart and aiming them more down the room as opposed to angled in at the mix position helped to smooth some response mainly in the upper mids between 500 and 1.5k.
Did you come to that conclusion by doing one speaker at a time individually? Don't forget that when you have both on at once, there are interference patterns that produce their own peaks and nulls. It's possible you were seeing that, if you were doing that with both speakers.

Also, are you looking at all of the data there, or just the frequency response? FR is only a small part of room tuning. If you take a look at the Spectrogram for the two tests "L no racks" and "L no racks 6" in", you'll see that you are exciting a mode at around 78 Hz much more strongly at the 6" in position. Yes, it is smoother overall, and it might be an option, but we'd need to address that additional modal issue. Also comparing the same data for the R speaker, you can see the same result, plus the 1.0.0 mode is a lot stronger are better defined at 6" in...

Maybe you could try a position around 3" in, and see if that is a good compromise.
When doing the tests with the 4space racks on and off of the desk there is definitely more problems in the same 500-1k range with them on the desk.
Yeup! Hence the recommendation to never have speakers over a desk, or on a meter bridge... (despite what you see in many magazines, showing off high-end studios... :) )
Makes me want to experiment with getting the speakers higher off the ground and angled down slightly to the mix position.
You can do that, but only very slightly. The higher you raise and angle the speakers, the WORSE the desk issue becomes, since the bounce angle is now also increased, and from a higher starting position...ñ so you get more reflections into your ears... Murphy... There's also the psych-acoustic effect of having sound coming down at your ears from higher up: that messes up frequency perception, as well as directionality. Don't ever go more than a 10° tilt on your speakers, and that is already major, huge. I personally don't ever go over 5° tilt. Depending on your speakers, and their directivity patterns ("Q"), it is often better to just raise them an inch or two WITHOUT tilting them...
The table that I'm currently using is fairly large and I don't need that much space. I was planning to build my own desk that'll be more efficient in less space. Any tips in that department?
Make it as small as possible, tilt the console towards you, embed it into the surface, keep the surface as low as is comfortable, and even consider tilting some or all of the surface.
when looking at the SPL&Phase window after taking a measurement. In the lower phase graph where it shows what looks somewhat like a curve opposite to that of the SPL graph (I'm assuming thats the phase graph) where it shows the vertical dotted lines.. Is that where frequencies are phasing out? Standing waves? Since the SPL lines either peak or null at those frequencies that's what I assume?
Nope. That's just where the phase "wraps" off the bottom end of the REW scale, and comes in back at the top again. Since phase is measured in angles, relative to a circle, that's the point where the phase came around full-circle and kept on going around. If you look under "controls" tab, there's an option to "unwrap" the phase, so you can see the full curve. Technically, it's not correct to look at it like that, but for us mere mortals it's a lot easier to understand "unwrapped".

I'm out of time right now, and it's going to take a while to analyze the REW data better and get back to you on the rest, but at least that answers some of your questions for now.


- Stuart -
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Please help me tune my control room….please.

Post by Soundman2020 »

I'm looking at your MDAT data, Jason, but some things don't make sense: Why is there a 5 dB jump in the levels when you moved the mic from 80" to 86"? That is inexplicable, unless you adjusted the calibration between those two readings. All the ones before that (62", 68", 74" and 80") are all at around 77 dB, but from there on (86", 90", etc) are all done at a level of about 82 dB. Why the discrepancy? If you did change the calibration, I could adjust the curves to compensate, but I'd need to know exactly how much you changed it by, otherwise the results would be invalid...
I did notice that you take a break of nearly an hour between those two sets of measurements, so it is likely that something happened in that time frame...

- Stuart -
JasonC
Posts: 31
Joined: Thu Aug 15, 2013 1:24 am
Location: Alpharetta, GA
Contact:

Re: Please help me tune my control room….please.

Post by JasonC »

Hey Stuart,
My apologies with the calibration of REW. There must be something that I'm doing incorrectly because it asks me to calibrate every time I open the program. I use my SPL meter and the program's db meter to match the levels like it says to and I do a "Check Levels" before doing my tests and it gives me the OK.
Yes, when I was doing those tests I got pulled away for a little bit and had to come back to it. I must've bumped the mic pre or my speaker volume by accident when I returned to the tests. If you say it went from 77 to 82 db I must have. I use a Dangerous Monitor ST for my speaker management and it's stepped volume control works in 3db increments so I'd guess I may have bumped it up by accident. If you think doing so may have compromised the tests, I'll shoot another round. Really sorry.

Thanks again,

Jason
JasonC
Posts: 31
Joined: Thu Aug 15, 2013 1:24 am
Location: Alpharetta, GA
Contact:

Re: Please help me tune my control room….please.

Post by JasonC »

Stuart,

Hope all is well!

Toss out all my previous mdat's. You were right from the get-go. My calibration was off. When I was calibrating before previous tests I misread a step and corrected my error. I've got a solid cal file and have documented all my system's I/O settings to be exactly the same every time. I can't apologize enough for wasting your time on my previous tests. :oops: But hopefully this will be a much clearer analysis of whats going on in my room.
You can download it here....
http://we.tl/F7KPSITNFC

After properly calibrating I did a test with the speakers in their averaged proper place about 60" apart and about 7" from the front walls with the mic centered starting at current mix position about 66" away from the speakers. I did L+R, L, and R tests starting at 66" back and moving back in 6" increments to 114". There's no absorbers on the left, right, front, or back walls. Just super chunks in the rear corners and the clouds overhead.
Tomorrow I will be bringing down the cloud over the mix position, adding a layer of plastic sheeting behind the face fabric, adding more 703, putting as much hard backing on it as possible, and increasing its downward angle to 20 degrees. Then doing another round of tests.

Sorry again for the confusion. I tried to compare the new test to the previous one as best I could but I know you'll see a lot more than I. Please take a look at it when you get a chance and let me know the new findings!
Thanks!

Jason
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Please help me tune my control room….please.

Post by Soundman2020 »

I will take a look at it again, Jason, but it will have to wait a couple of days, I'm afraid. I'm really tied up right now with several projects for paying customers, and very little time for the forum. But I will get around to it! I did download the MDAT file, and I have it lined up to look at in detail...

- Stuart -
JasonC
Posts: 31
Joined: Thu Aug 15, 2013 1:24 am
Location: Alpharetta, GA
Contact:

Re: Please help me tune my control room….please.

Post by JasonC »

No problem Stuart. As soon as you get a chance.
Thanks for responding and all your help!

Jc
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: Please help me tune my control room….please.

Post by Soundman2020 »

Here I am, once again, at nearly 4 AM going over your thread one more time, to figure out what needs to be done to get this room working.... :)

- Stuart -
Post Reply