What kind of equipment is needed to determine a room's low frequency response? (1 Viewer)

[Larry Chanin]

I've got a home theater in which I have a home theater computer (HTPC) with a pretty decent audio card, permanently connected to my A/V receiver. After reading a number of threads on this forum discussing the use of acoustic software to do room frequency graphs, it occurred to me that my HTPC might lend itself perfectly to such an application.
BUT… I'm not a Do-It-Yourselfer, or an audiophile, so I don't want to go overboard (like I usually do), by buying a lot of expensive hardware and software just to calibrate my equipment once in a while.
For my first foray into this area what I'd like to do is see if I can use some of this software to more scientifically integrate a newly purchased subwoofer (which the manufacturer claims will to go down to 18 Hz) with my old, Polk SDA main speakers.
With the help of this forum I've located a number of freeware and shareware programs that provide Audio Spectrum Analysis software that I plan to experiment with, and I've got the following equipment:

• A 1 GHz Pentium III PC with an M-Audio Audiophile 2496 Audio Card with RCA analog inputs.
• A Radio Shack Sound Pressure Level Meter (and I located the thread on meter correction values).
• I also have an old Ekectret microphone (with a low end frequency response of 30 Hz) that came with a Graphic Equalizer. It has a long run of cable terminating in a 1/4" plug.

My specific questions are:

1. Which microphone should I use, the Radio Shack SPL meter's or the Ekectret?
2. If I use the Ekectret do I need some sort of calibration data similar to the Radio Shack correction values?
3. Do I need a microphone preamp between either the Radio Shack or Ekectret and the audio card? My audio card's manual says, "In order to preserve its high dynamic range and minimize distortion, the Audiophile 2496 does not have microphone preamplifiers built into it. Therefore direct connection to a microphone is not recommended. Instead pass the microphone signal through a microphone pre-amp (such as the M Audio DMP2™) and then connect the pre-amp output to the input of the Audiophile 2496."
4. If so, can you recommend a suitable, inexpensive microphone preamp?
5. I've seen a few web sites that have free downloadable test tones, as well as commercially available test CD and DVD's that range in price from $7.95 to $150. For low frequency response analysis what type(s) of test tones would you recommend?[/list=a]
   Thanks in advance for your help.
   Larry

 

[Wayne A. Pflughaupt]

Larry,
A spectrum analyzer (a..k.a. real time analyzer, or RTA) is only as good as the mic feeding it. Stand-alone analyzers, like those made by AudioControl or Goldline, always come with a calibrated mic. Actually “calibrated mic” is a misnomer. Mics with true ruler-flat response cost thousands of dollars; therefore manufacturers of spectrum analyzers actually calibrate the unit for their dedicated mic’s response, rendering its less-than-optimal readings as flat.
Here’s an example of what can happen when using an RTA with the wrong mic. I used to install pro-audio sound systems as part of a crew for a large company. When I moved to a smaller company I became responsible for the final tweaking and EQing of newly-installed systems. Typically we used 1/3 octave EQs and a Goldline RTA for this. Although I was very familiar with RTA theory, I had never actually used one before. After a while I noticed that the RTA was giving similar readings (before equalization) at most of the sites I used it. Specifically there was a broad dip at about 400Hz, and a gradual roll-off of the higher frequencies above 8kHz. I noticed in one instance that boosting the equalizer's 12kHz slider a full 12dB made a difference of only 6dB on the RTA’s display. However, I chalked it up to the inherent poor high frequency response of sound reinforcement speakers.
I finally saw the light when I took the RTA rig to a friend’s house to help her set up her sub and saw the same response curve, including a roll-out of the high frequencies. Of course, any hi-fi speaker worth its weight in MDF has flat response out to at least 16kHz! (These particular speakers were Klipsch.)
It was only a short time later that I found out the company owner had lost the RTA’s original calibrated mic and had substituted a “shotgun” mic, typically used for live video work.
This “what mic to use?” issue is something that has chronically troubled me with the providers of on-line spectrum analysis software. They are really loose about the mics they recommend: “Any high quality mic will do;” “Mics A, B, or C will all give excellent results,” etc.
Another thing they never consider is the sound card. It shouldn’t take a genius to figure out that the sound card also plays a critical part in the signal chain.
Bottom line, Larry, I wouldn’t take much stock in any spectrum analysis program that doesn’t specify a specific sound card and mic that they have designed their software to work with. You will be better off renting an RTA from a pro-audio store, or using frequency-specific test tones with an SPL meter.
Regards,
Wayne A. Pflughaupt
[Edited last by Wayne A. Pflughaupt on July 28, 2001 at 01:03 PM]

 

[BruceD]

I use a PC software system called ETF5. This specific software program has a couple of benefits for use with various sound cards and mics. It uses the Full-duplex capability of a sound card and compares the right and left channels to correct for this specific sound card. It also provides a file for specifying the exact frequency and phase characteristics of the microphone you are using (I bought a mic and pre-amp that were measured with a response correction file supplied on a diskette).
www.etfacoustic.com
The software generates MLS sequences for computer analysis which is more accurate than many RTA units and frequency test CDs.
If you go to the website, you will find an demo room that they use as an example to show how to tune your speaker-room interface. This demo is a very good tool for just understanding the various ways to tune your room.
BruceD
[Edited last by BruceD on July 28, 2001 at 05:16 PM]

 

[Wayne A. Pflughaupt]

It also provides a file for specifying the exact frequency and phase characteristics of the microphone you are using…

Does it do the same for the sound card? If not, there’s your weak link in getting accurate readings.
Regards,
Wayne A. Pflughaupt

 

[BruceD]

Wayne,
ETF5 uses the full duplex capability of the soundcard to test and correct itself automatically for any soundcard. No need to enter a correction file manually, one is generated and implemented by the program itself each time the program is executed. This means you can even change sound cards and not have to worry about re-calibrating the program.
BruceD