Why is there a Difference between RTA and SPL Meter Readings between Sine and Pink PN Signal Types?

[AudiocRaver]

John, anyone...

Probably already asked and answered, and it seems like there was even a recent discussion about it, but I could not find it anywhere. Why is there a difference between these SPL readings?

1. Using an analog mic (with audio interface, not a USB mic), I calibrate the SPL meter with a 200 Hz sine wave from the Signal Generator set at -20 dBFS, so a -20 dBFS reading on the SPL Meter corresponds with 85.0 dB SPL. The meter on the RTA panel will also read 85.0 dB SPL. All good so far.
2. Changing the signal type to PPN, with the Generator still at the -20 dBFS setting, the SPL meter now reads -20.8 dBFS and 86.4 dB SPL, while the RTA gives a 84.3 dB SPL reading.

Shouldn't the RTA and SPL Meters, given the same meter settings, yield the same SPL?

And, if RMS is RMS is RMS (is it?), then shouldn't a 20.0 -dBFS reading on the SPL Meter always correspond to an 85.0 dB SPL, regardless of signal type?

Thsnks, John.

 

[John Mulcahy]

By PPN I guess you mean the full range Pink Periodic Noise signal, and that the SPL meter was on Z weighting (i.e. no weighting applied) and the High Pass filter was not active? That being the case the factors that would have the strongest influence are:

• Any cal files that were applied (soundcard and/or mic) - that's the most likely way for a lower dBFS input reading on the meter to end up in a higher SPL reading. The dBFS figure on the SPL meter is the rms value of the input data, but the SPL reading is generated from the FFT of the input data so that the cal file frequency responses can be applied. As the full range Pink PN is genuinely full range the cal files need to be accurate over the whole frequency range. A cal file generated from a soundcard loopback measurement will show the same figures for a sine wave input and full range Pink PN on SPL meter and RTA.
• The window on the RTA, the RTA needs to use a rectangular window when being fed a periodic noise source otherwise the window will alter the structure of the signal
  
• The RTA FFT length, that should match the Pink PN sequence length - REW generally takes care of that if the RTA is already running when the Pink PN signal starts playing

 

[AudiocRaver]

Super, thanks for the explanations.