Unsteady SPL Readings

My subwoofer does not produce a steady SPL reading, unlike my other speakers which only fluctuate by 1db. Any ideas what the problem is?
If you are using a very cheap slm, (ie radio shack type), they are not designed to go that low in frequency response to get an accurate, consistent reading. It may help to get out of "a weighted" if you are in that mode as well. With cheap slm's, "c-weighted" comes the closest to being flat.
What are you feeding it with?

The filtered-noise signals most of us use for level matching start as RANDOM noise. Even on the extra-slow setting of my analog meter, which already is easier to read in a dynamic situation than a digital meter, the reading must be averaged.

Only a sinewave would give you a rock-steady reading, but they're highly subject to peaks and dips in response.
That's what they do. You're going to have to average it out, and really........ listen, to ultimately dial it in. Nature of the beast. Be sure you have the meter on 'slow' response. That helps 'some'.


C-weighting is supposed to be essentially flat, whereas A-weighting is supposed to simulate hearing sensitivity whether its a RS, B&K, or GR sound level meter. I've used them all, but now as a retiree I use only the RS, because Breul & Kjear and General Radio are expensive. I use C-scale and slow trajectory with 0 re 80 dB SPL when measuring the white noise generated by my Proceed PAV. As a reminder, white noise is equal SPL across the pass band, whereas pink is equal SPL across octave bands; 0 dB SPL is re 0.0002 dynes per square centermeter, IIRC.

The Radio Shack is just not designed to accurately measure low frequencies whether it be noise or sinewave. It cannot respond near as well as other slms. However, come to think of it, you could be having some cone breakup on the sub, that may give you that problem. Also, the radio shack is better than nothing I agree, but its absolute accuracy can vary by as much as 3dB from low frequency to higher frequencies (ie 1kHz) I've had to calibrate these things and what I really wish is that dealers would spend money for a type 1 SLM with octave analyzer, and really dial in a customer's system. As you nodoubt know Don, even an exact C weighted curve is going to change the reading when you are in a passband mode with a subwoofer.