How do you measure the acoustics of your control room ?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ubxf

Well-known member
Joined
Sep 3, 2004
Messages
811
Location
los angeles
I realize that it is a task better left for a specialist, but before i call in an expert i'd like to check the response so i can learn and also see if i have more issues than i'm aware of. The room is L 24xW16xH10  .
If you don't mind sharing your thoughts it would be greatly appreciated.
 
Simple explanation. You look a lot of this up on the net.
You can start with mathematics.
http://www.sengpielaudio.com/calculator-roommodes.htm
Audio testing  is done with running pink noise through your speakers with a calibration mic set in the mixing position.
The mic is fed into a spectrum analyzer  and it will show what your room is doing with the various frequencies.
 
Thanks Winetree
i have done a lot of reading and the preliminary math calculations, the pink noise through the speakers is the part i'm trying to learn about. I thought that you need a special speaker that sends signal equally  all around ( omni ? ) and you place it in different positions in the room while the mic is steady . Then you do the reverse the speaker is in the center and the mic changes position.
Hi Ruairi, i'm downtown near USC.
 
ubxf said:
I realize that it is a task better left for a specialist, but before i call in an expert i'd like to check the response so i can learn and also see if i have more issues than i'm aware of. The room is L 24xW16xH10  .
If you don't mind sharing your thoughts it would be greatly appreciated.

Ideally, you'll use a transfer-function measurement program such as Smaart (yes, I'm a Smaart partisan). You put a measurement mic in your listening area, and run program material through the monitors, generally pink noise, through the system. You want a transfer function rather than the simpler RTA because TF gives you phase information.

Interpreting the measurement results is the hard part. It takes some experience to determine the causes of suck-outs and boosts in your frequency response. The idea is that you can see whether the problems can be corrected by electronic equalization or whether you need to deal with the architecture.

-a
 
There is so much to say on this.  One of my hats involves installing high end PMC speakers for clients in rooms of every description, from refitting all 3 rooms at Capitol Studios to every kind of home studio you can imagine (from bare wall bedrooms to $2M build outs).  This involves positioning, listening, measuring and quite often working with acoustics or a chosen acoustician or designer.

I will say that unless you are prepared to go quite deep in your learning you are better off leaving it to an expert.  There is unbelievable amounts of misinformation on the net re acoustics.    Finding the right expert is the trick, many of them are not quite what they seem.

Unfortunately there is very little that EQ can really do to correct a typical speaker/room combo. Understanding what it can be used for is tricky.
 
Ethan (Winer) may have written a chapter on just that subject in his book, while he has posted a lot about studio acoustic on other forums.

I suspect studio acoustic design is changing with the evolution away from a large console control surfaces smack in the middle of the listening position. .

There is surely a great deal written  on this subject... If you try to make measurement use more than one sample location to minimize errors due to room modes or combing from reflections off nearby surfaces.

Good luck

JR
 
Thank you guys for all the advice, i'm going to have a pro work on it . I managed to get the frequencies fairly well distributed but i have some weird phase issues.
 
I've used many different softwares, fuzz measure is quite handy for fast measurements which gives a lot of info but not useful to work in real time which you may want to tweak eq and acoustic traps for placement or tuning. I just use one speaker at a time, facing away from the mic, in a corner of a room, the mic placed at the center of the room, to capture a general idea of what's going on. For the mic I have an ECM8000 than is really handy for all this measurements than don't need laboratory grade equipment, any smallish diaphragm omnidirectional condenser will work just fine as long as it doesn't have a really bad freq response but even quite cheap electret capsules are good enough for this application.

Usually the first thing is to bring RT times to a confortable amount, depends on taste and room size, by the use of absorption, I wouldn't care too much about traps on the first step.
Then you need some way of getting a waterfall which will show you the problematic frequencies on your room, at the low end, the exact frequency to call it low end will depend on the room size, this is like a more detailed RT analysis. You should attack each long ringing shown on the waterfall independently with traps of various types, if you have a few close to each other a wider band trap will work. Once you have solved those problems you could go for the next cryptical step which is placing the monitors and the listening position.

The one thing to have in mind is called SBIR (speaker boundary interference response) which will cause peaks and dips pretty much independent of the room size but related with the positioning of the speakers themselves, something similar with the placement of the listening position. There was a good software to work with all this (acoustic X, late 90's) and I haven't seen anything quite as handy since then, you would need a W98 or something like that to run it... I've written a little function in mathematica to play around with it, the idea is to get the smallest dip possible and don't care too much about peaks, since they can be attenuated by filters without much of a problem (other than phase shift but that's not a problem any more). Once I solved one really big problem just by shifting the x-over frequency by 1/3rd octave because as the placement is different you will have different problematic frequencies on the sub and the HF monitors, then usually some attenuation in the low end is almost always needed. At this step I usually align the phase as good as I can for the main listening position, cut the peaks and call it a day, as long as I don't end with big dips.

The "string analyzer" was other method which consist in attaching a string from the speaker to a mic stand at the listening position, the string should be variable in length, adding a few centimeters to the initial length will show the comb filter generating surfaces, usually the desk, mixer, etc. Then adding about 2m will show the surface which can cause image shifting, using a small mirror and a flashlight, this usually take two people, one pointing at the mirror with the light, the other placing the mirror in different surfaces which may be problematic pointed by the string analyzer, if the light is shown at the monitor position (or close enough) you should treat that surface, could be changing the position, angle or using some absorption so the reflections are attenuated or diffusion on it so the reflections are not specular (as the light with the mirror) and much smaller.

Then make a last measurement of all you can from where the monitors are to the listening position to see if everything ended up fine, then maybe a few other places just to be aware of the problems out of the listening position.

That should cover as quick as I can how I do this kind of things, it could be pretty quick or not, easy or not, depending on the room which you started off, where you want to get and the time you have to do so. A proper control room could have some diffusers for example but that's quite more involved and time consuming, since usually a custom made solution is what is needed to work at the proper freq you need them to and to work properly in the particular room you are treating. Also the measurement of that is quite hard to do and quite hard to see what you need or how better it is after the job is done. Is not imposible to measure or just better in subjective POV but much harder and a proper measurement of those parameters to find any specular reflections hitting the listening position would require equipment that's not so easily available or cheap.

JS
 
Thanks Joaquins for your detailed description. I will definitely do more analysis based on your infos.
 
ubxf said:
Thanks Winetree
i have done a lot of reading and the preliminary math calculations, the pink noise through the speakers is the part i'm trying to learn about. I thought that you need a special speaker that sends signal equally  all around ( omni ? )
This is not necessary; use your spekers. that's what you'll be listening through anyway. The requirement for omni is only when you do architectural acoustics (noise transmission/isolation).
and you place it in different positions in the room while the mic is steady . Then you do the reverse the speaker is in the center and the mic changes position.
Again, the latter is not necessary for alignment of a control room.
As a starting point, you have to know that room modes are intrinsic to the room, as the name implies. That means the frequencies are governed by the dimensions of the room. However, standing waves create zones of high SPL at certain frequencies (antinodes) and zones of low SPL (nodes). It is statistically impossible to find a locus where there is no node neither antinode. However, the antinodes of lowest frequency are the more problematic. the position of the nodes/antinodes is quite easy to determine with very simple math. IMO the most useful link is there
http://www.sengpielaudio.com/calculator-roommodes.htm
Then there are other artefacts created by reflections on the walls. These occur at a frequency that is determined by the position of the speaker relative to the walls; that's where experimenting with speaker placement is important.
For proper assessment you need a sweepable sinewave generator and a good pair of ears if you don't have an SPL-meter.
Then you'll need an RTA, and learn how to use it. Many beginners go astray trying to EQ to a fraction of dB, only to find that if they move the measurement mic by a half-inch, everything has changed!
The RTA will tell you in broad terms how your speaker+room combo works, but the resonances must be dealt with sinewaves.. Don't worry too much about dips in the response; they are much less bothering than peaks.
Just google for RTA software; the RTA from REW is excellent but I think the learning curve is pretty steep for a beginner. Ideal configuration is 1/3 octave. There are 1/6th and 1/12th octave RTA's but there's no practical advantage in them for DIY acoustics.
 
I have knocked many rooms into shape. I am no acoustician and don't pretend to be one.  I have never used software to do this. Some rooms I've had a hand in have been shot with tone. On a macro level the report is never very much of a surprise.

Unless you have the experience listening and being able to translate that into home brew solutions then the software won't help you.  Like Andy said, the hard part is interpreting the results and knowing what to do with that information.
 
abbey road d enfer said:
IMO, tuning a room without a generator is akin to tuning a guitar amp without a multimeter. It can be done, but I don't think it's the most efficient way.

If you don't know what you are doing a generator won't help you. Doing it without a generator may not be efficient but you end up in at least as good a place if you are careful.
 
Just to elaborate a little.  I do not like speakers to be ruler flat to 20k. A gentle roll off above 12k-15k sounds right to me. Many studio speakers are ruler flat.  Most studio rooms will be over damped at high frequencies and produce this rolloff naturally.  Most people like it this way.

Knowing how and where to increase reflections can produce an "EQ" change. There are a million little things you can do to change perception. Hopefully you don't have to.

You would have a wacky sounding room if you obsessed about a straight line on the software.
 
For literally for the first time in the history of my time here I'm going to disagree with Abbey.  I strongly recommend against trying to tackle room issues with a sweepable signal generator and an SPL meter.  Perhaps for someone with the technical understanding and restraint necessary (like Abbey) it may work but for most everyone else it will be an infuriating and fruitless exercise.

I see two possible routes 1) dig in very deep and learn what needs to be learned or 2) hire a pro.

EQ can only be used to tackle response issues in the speaker itself, and even then only certain issues.  It will be all but useless when dealing with room issues. 

I've spoken with the O.P. on the phone and I'm going to make time in the coming weeks to take a listen and see what needs doing.  Hopefully UBXF can report back on what we find.

 
ruairioflaherty said:
For literally for the first time in the history of my time here I'm going to disagree with Abbey.  I strongly recommend against trying to tackle room issues with a sweepable signal generator and an SPL meter.  Perhaps for someone with the technical understanding and restraint necessary (like Abbey) it may work but for most everyone else it will be an infuriating and fruitless exercise.
I believe you misunderstood me. I have recommended the use of a sweepable generator for assessing room resonances and the position of nodes and antinodes, not for equalization (have I mentioned equalization?). It doesn't take an EE degree to do that; I agree that the SPL meter is not 100% necessary, but it's part of the education. Just the same, my suggested use of an RTA was not for EQ, but for assessment of the sonic "character" of the listening environment.
I see two possible routes 1) dig in very deep and learn what needs to be learned
That's part of my suggestion. Then I answered the OP's specific question "How do you measure the acoustics of your control room ?"
  or 2) hire a pro.
Isn't this site about DIY? If we answered all questions about "how to do this or that" by "leave it to a pro", we'd lose the essence of our group. I believe we should always A) answer the question  B) expose the caveats
 
EQ can only be used to tackle response issues in the speaker itself, and even then only certain issues.  It will be all but useless when dealing with room issues. 
  That may be another point where we may disagree; this is a common belief in the audio community. Reflection-induced room response cannot be improved by EQ because of the non-minimum-phase nature of these.
Although it is true that a room cannot be perfectly equalized in the mathematical sense, a room's response can be decomposed as a product of non-MP and MP polynomials. The latter can be perfectly eq'd. Thousands of listening environments testify to the fact that a well eq'd one sounds better than non eq'd. This takes a lot of preliminary knowledge, expertise AND equipment and is out of the scope of DIY. That's is definitely NOT what I suggested.
 
Hi Abbey,

I re-read your post and don't really see how I misunderstood but no matter, I very much respect your input and don't mean to criticize or derail.

My mention of EQ was not a direct reference to your post.

I agree that where possible we should "A) answer the question  B) expose the caveats" but in this case my experience suggests that I can't do either in written form in this case in a succinct manner.  It is too complicated and there are too many caveats.  Hence my offer to visit UBXF's space and help out.

That may be another point where we may disagree; this is a common belief in the audio community. Reflection-induced room response cannot be improved by EQ because of the non-minimum-phase nature of these.
Although it is true that a room cannot be perfectly equalized in the mathematical sense, a room's response can be decomposed as a product of non-MP and MP polynomials. The latter can be perfectly eq'd. Thousands of listening environments testify to the fact that a well eq'd one sounds better than non eq'd. This takes a lot of preliminary knowledge, expertise AND equipment and is out of the scope of DIY. That's is definitely NOT what I suggested.

I will disagree with you here in a practical sense based on my experience in literally thousands of rooms.  I absolutely understand the non MP and MP issues you describe but EQ can only ever correct MP issues in one listening position and that is not a satisfactory solution in my eyes.

I agree with you 100% that is there is no need for a special speaker, measuring with what you have is best.  REW is a fantastic tool and there is lots of help available online.










 
ruairioflaherty said:
Although it is true that a room cannot be perfectly equalized in the mathematical sense, a room's response can be decomposed as a product of non-MP and MP polynomials. The latter can be perfectly eq'd. Thousands of listening environments testify to the fact that a well eq'd one sounds better than non eq'd. This takes a lot of preliminary knowledge, expertise AND equipment and is out of the scope of DIY. That's is definitely NOT what I suggested.
I will disagree with you here in a practical sense based on my experience in literally thousands of rooms.  I absolutely understand the non MP and MP issues you describe but EQ can only ever correct MP issues in one listening position and that is not a satisfactory solution in my eyes.
The same limitation can be said of speaker positioning. Optimizing speaker position is what gives the best response in a given hearing position. Does it mean it is not worth being done?
 

Latest posts

Back
Top