What can I use to measure the frequency in my room? (1 Viewer)

[DanielKellmii]

I am interested in measuring the frequencies that I actually hear in my room. Other than lugging my computer in there and buy a decent microphone, what can I use?

 

[Wayne A. Pflughaupt]

An cheap and easy (but time consuming) way is to use a test disc with 1/6-octave sine wave signals, an SPL meter and some graph paper. You take SPL measurements of the individual tones and plot them on the graph paper for a visual of your response.

Regards,
Wayne A. Pflughaupt

 

[ChrisWiggles]

An SPL meter.

 

[DanielKellmii]

Chris, a SPL meter just measures the sound pressure, that is the amplitude, What I am looking for is the frequency.
Think of a sine wave or just a line that rises and lowers to the same point as time goes on. The SPL meter measures how high the line goes. Frequency is how many peaks and valleys are squeezed in a given length. The more peaks and valleys, the higher the frequency (a flute), the fewer, the lower the frequency (a bass). A picture would be so helpful here, but I have no server to host it on.
Waynes suggestion is great, but I need to get a test disk and some time to do this. I don't think I will ever have the time for that.

 

[ChrisWiggles]

I know. You use sine tones and an SPL meter, and graph it. That's the hard, but cheap way without spectral analyzing equipment or computer stuff.

 

[Wayne A. Pflughaupt]

I can’t see it taking more than 30 minutes or so to accomplish – maybe 45 tops.

Regards,
Wayne A. Pflughaupt

 

[Oachalon]

sorry to jump in on this thread but lets say i had a computer and a decent microphone in my room how can i find the frequency response of my room. What program do i use and how do i set it up. Would u count the microphone that came with the yamaha 1400 as a decent microphone. Thanks

 

[DanielKellmii]

Wayne, that is almost a possibility. Does the AVIA disk have those kinds of tones? If not, where can I get them? Maybe I can generate them myself and record them on a CD on my computer.

Now all I have to do is convince my wife to take the kids out shopping while I "play." To her, if I am not cleaning the house, I am playing. Cutting the grass...playing..Fixing the car....playing..Pressure washing the driveway.... playing....:frowning:

 

[Nathan Stohler]

Daniel,

Do a web search for NCH Tone Generator. I used that to create test tones, and then I burned them to a CD.

 

[WayneO]

If you have broadband I can email you 2 3MB files that have 10-100hz 1/12 octave tones and has an excel chart as well that you can enter values and will apply the correction factors for the RS meter and graph it for you. PM me your email if you'd like them.

 

[Bill Waxman]

I downloaded the NCH tone generator, but I am unsure of which tones I should burn and at what levels. I'm also not sure where various octave levels are or what levels they would equate to. Also, are these signals being created by a mono or stereo output in the program?

I guess, basically, that if anyone has a step-by-step resource for creating a test cd and how to use one, I would greatly appreciate a link or the help. I thank you very much in advance for the information.

-Bill

 

[Steve::Weaver]

That's a good thinking, Bill. Why figure out something yourself when you can have someone do it for you?

All kidding aside, as an engineer (not audio) turned management-type, it doesn't make sense for me to spend a couple of hours figuring things out on my own if someone has already done it and can be persuaded to share what they already know along these lines.

I know it'd take a few minutes (at least) to put this down on paper (or in bits) but if someone already had a procedure and test tones handy, it'd make a great addition to the FAQ. (I'll admit that I didn't look for this in the FAQ just now, but I don't recall seeing it in there when I read completely through it a couple of months ago.)

Q: How to I determine the frequency response response of my sub/speakers/'72 Ford Pinto?

A: First, you'll need an SPL meter (Radio Shack catalog #33-4050), and a set of test tones, which you should download from...

 

[Bill Waxman]

Steve,

I use the FAQ and Glossary quite often (still relatively new), and you are correct that there is nothing there yet (that I could find). Frankly, reinventing the wheel has never appealed to me, but I do agree that it is best to learn things on one's own. I've read that calibrating one's system can improve it's sound. My thoughts were that, as a novice, if some one had a proven method, it would be most likely be far more reliable than anything I could piece together.

In my research, the only test tone suggestions I've found are for calibrating sub woofers. It was suggested that one measure the sound levels from 16-120hz by factors of 8 and then try to correct the various levels. However, this fails to explain higher frequencies as there would be 2,485 levels to test between 120hz and 20khz for calibrating the speakers.

I appologize if I offended your belief that knowledge should be earned. I generally agree with this philosphy when others are simply being lazy. But I can assure you that this is not the case. I have learned a great deal since finding this forum, but I fear that my knowledge is not yet where I wish it to be. Perhaps I will eventually find enough information on test tones to add a page to the FAQ! (Probably dreamin' on that one though, eh?)

So, if anyone does have some sort of step-by-step guide, or is willing to create one, I'm sure that many people, perhaps Steve as well, would be greatful for it. Otherwise, stay tuned for mine...

As always, thank you in advance.
-Bill

 

[Steve::Weaver]

Bill, I think you may have mistaken my joking first sentence as me being serious, but I definitely was not.

In fact, I have a hard time being serious even when I try to be (please note the the above Ford Pinto comment).

As for actual measurements, it seems to me (and I'm a self-certified audio moron!) that you'd want to measure logarithmically. That is, you take the same number of measurements between 10 Hz and 100 Hz as you do between 100 Hz and 1 kHz, and between 1kHz and 10 kHz. What resolution you choose for this scale would depend on how precise you wanted to be, but you could certainly get a decent response graph with no more than a couple hundred measurements.

Ideally, this would be done with a high-quality microphone and a computer, playing test tones and recording the actual room response, but it could certainly be done by hand with a little patience.

Surely several people on this forum have done exactly that and could shed some light on their methodology. We just have to wait for them to come out of the woodwork...

 

[Chu Gai]

If your intention is to somehow use the SPL meter, whether it be the RS or one costing thousands, and burn test tones, covering a range of frequencies from say 20 Hz to 20 kHz in whatever increments, in order to somehow map your room, then don't. Simply because you obtain data and can plot it does not mean that it will be either accurate or useful. Further, even though thousands may have done it does not give the approach any more credibility.

Using such an approach will ensure that you have data with enormous response variations. They will be dependent upon..

• how the meter is pointed
• where its held in relationship to where you are

This is due to a couple of reasons. In the lower freqency area, much of the data is a result of room effects. In the upper frequencies there will be quite substantial variations as a result of interferences. Moving the microphone of the SPL just a bit will give you an entirely different set of results.

SPL meters were never intended to do such measurements as they are broadband devices to be used for giving a single number for the total SPL. You could though obtain something like 1/3 octave pink noise bands in which case you wouldn't see the anomalous results mentioned above inherent in sine wave mapping. Nonetheless, obtaining bad data can lead you to make bad decisions especially if you try and make some sort of corrections such as equalization in order to flatten out a room's FR.

If your intent is to obtain data that can be used to help you in some way this is not the approach.

 

[Cagri]

If the data is dependant upon where the meter is pointed and where it's held in relationship to the listener, isn't it still broadly accurate? I mean few of us listen to music without moving our head, in the form of a statue... Or there may be a moving body in the room every now and then...

What would be the better approach Chu?

 

[ScottCHI]

he said: "You could though obtain something like 1/3 octave pink noise bands in which case you wouldn't see the anomalous results mentioned above inherent in sine wave mapping."

if you want to see what he's talking about, set your meter up and do the sine wave measurements. then move the meter a foot forward and a foot to the left or right and repeat the measurements. unless you've gone to great lengths to reduce the room's effects, you should see quite a bit of variation, due to the room, from frequency to frequency.

 

[Cagri]

which makes me think he's suggesting a different way..

 

[Bob McElfresh]

There is a product called "Spectra Plus" that can be used to perform Real Time Analysis (or RTA) on your room. It is computer software that sends test tones out your sound card, and samples the results back from a microphone.

Spectra Plus

They have a 30 day trial copy you can download.

If you do a Google search you can find:

Harmony Central

HanDee - a PDA based analyzer

Spectra RTA

True Audio - TrueRTA

 

[Chu Gai]

Well I tried to confine my original reply to Daniel's apparent reluctance to get a decent microphone. That would imply getting a decent software package, that I assume one's going to pay for and not go posting on alt.cracks, possibly a decent soundcard if his isn't supported by the software for some reason, and a decent enough learning curve. That's a lot of decents isn't it?? And a lot of decents are bound to cost you some decent amount of time and a decent amount on your debit or credit card.

My recommendation would be to get a package that supports MLS. You can read more about it at http://www.mlssa.com/ although for a more reasonably priced consumer implementation you might want to consider ETF from http://www.etfacoustic.com/. While ETF will allow you to use the RS meter, I look at it like using an out of focus telescope to look at the moons of Jupiter or the Crab Nebula. The problems with the RS meter is that there is no uniformity among meters rendering any correction software or fudge factors applied to your readings to be largely meaningless. If you want to get decent results you need decent tools. In a pinch you can use a matchbook to gap your lawnmower spark plugs however its useless to set the intake and exhaust clearances on your engine's lifters.

One has to ask oneself, why do you want to determine the FR of your room? If it's just for kicks, then just look at the FR of your speaker system and be done with it. If it's because your profoundly curious about examining the effect of speaker placements (varying their spread and hence their distances from the side walls) then you need better tools. MLS is that better tool. If you want to tame a particularly nasty low frequency room mode, which is a problem in the time domain, then you want to properly identify it. MLS will do that for you and by doing so you can pursue the purchase or design of a bass trap specifically tailored for addressing that particular frequency. If your contemplating DIY speaker building, a package that includes MLS will be enormously useful as it'll allow you to do quasi-anechoic measurements as well as in-room response. The ETF package is but one (you'll note it also does RTA and a host of other techniques) but it is one with a fair user base and hence you can get a lot of feedback. There's a learning curve, strange terms that sound alien at first, but I'm a firm believer that most, if not all of use, given enough exposure and persistence, manage to figure these things out to become sufficiently competent. Osmosis and all that.