Why I stopped trusting my SPL meter at 35Hz
A class-1 meter is class-1 in the band it was specified for. The room below 40Hz is not in that band.
My SPL meter is a class-1 Bruel & Kjaer 2250. It is one of the most expensive single things I own and I trust it with everything except low frequency.
Class-1 is an IEC standard, and the standard’s tolerances are tight in the bands the standard was written for. Below 20Hz the standard does not specify; between 20Hz and 40Hz the tolerance widens by about 1.5dB; the mic capsule’s actual response, which is what determines the floor, has resonance behavior below 40Hz that the meter’s correction curve assumes a free field for.
In a small room, below 40Hz, there is no free field. There are modes. The meter’s correction curve is wrong. The meter does not lie — the meter is being asked the wrong question.
I have stopped using the SPL meter below 40Hz. I use a calibrated microphone into an FFT analyser and I read the magnitude at each modal frequency directly. The number is less satisfying — it is not a single dB SPL — but it is the number that exists.