View Full Version : Determining room and speaker response


mrdavie
11-18-07, 05:32 PM
It's been months since I purchased the Radio Shack SPL meter and a copy of Rives Audio Test CD 2. I hesitated to move forward with testing because frankly I don't know what I am doing. But today I decided to forge ahead.

I did test the battery in the meter to ensure that it is still good. I get stuck early on in setting up the test system. Following the instructions with the CD I set the meter to a decibel level of 80 and set its switches to C weighting and slow sample. Then I played the 1000 Hz track and adjusted the volume of my receiver until the meter read 0.

To make sure the setting was close to 0 I was supposed to check a couple of other frequencies. All other frequencies barely register on the meter. Now, maybe 1000 Hz is the crossover frequency for my speakers. If that is the case then in order to measure 0 on another frequency I would have to turn the volume up REALLY loud. I can't do that. First, I don't want to damage the speakers. Secondly, I would not be able to stay in the room.

What am I doing wrong? OR does the sound have to be really loud in order to make these meter readings? :confused:

JKnPA
11-18-07, 06:36 PM
Article...............
http://www.techlore.com/article/10037/;jsessionid=1EB44AF24FE014D8DE008AB481B8A36F

trekguy
11-18-07, 08:27 PM
There are two different sorts of measurement that you might with your speakers. One is to measure the frequency response of the speakers; that is how much the response varies across the frequency range of the speakers with a constant input power. In theory you can make this measurment with individual test tones and an SPL meter and graph the result.

The other is setting the volume trim on each of the channels of your receiver to deliver equal sound pressure at the listening position and to set the Dolby reference level. The article in the first reply is about doing this. Pink noise or white noise is normally used for level matching.

While you can get a rough measurement of the frequency response curve of your speakers in your room, by using test tones and the RS SPL meter it is not the best method. The RS meter does not have a flat response across the audio spectrum. It is not designed for that purpose and has a marked fall in sensitivity as frequency increases. Correction tables are available for the lower frequencies but not the high end.

If your reciever has a built in level setting feature (with mic) you can try using that mic with the free Room EQ Wizard software (supported on Home Theater Shack and AVS). It may work well enough for your purposes--the mic included with my Onkyo is much better than the one in the RS SPL meter.

mrdavie
11-18-07, 11:37 PM
Ok. These two responses are great and I have more homework to do. First I will set the level with the pink noise and then focus on the second reply. My receiver does have a mic to set speaker size and levels.

Ethan Winer
11-19-07, 01:47 PM
Yes, using pink noise is key for level matching. Sine waves are not suitable for that. I have an article coming out in a few months in AudioXpress magazine about using an SPL meter. I can't post it all here :D but the snippet below is relevant. I use the DVD Essentials DVD which is excellent for this, and for many other calibration tasks too.

--Ethan

Using sine waves is not a good choice for matching loudspeaker levels because they cause standing waves in the room. Standing waves create peaks and deep nulls in the frequency response that are highly positional. If you play a 1 KHz sine wave and note the level on the meter, then move the meter only a few inches, the level will likely be very different.