The SPL meter shows an RMS value when in SPL mode (root mean square, the square root of the average of the squared values). It is the correct reading to use when balancing levels.

If you want a long term average select the Leq mode on the SPL meter, that will generate a value averaged over the period since the Reset All button was last pressed - the time period over which the average is calculated is shown on the SPL meter Elapsed Time display. To get accurate readings using this mode start the test signal, press Reset All to reset the averaging, then wait for whatever period you are happy with and take the reading.

 

Hi JohnM,

Thank you very much for the reply and advice.
I tried calibrating each speaker using the SPL-meter and a pink PN-signal played through jriver media center.
Oddly, when I go to the spectrogram, using a pink PN-signal, the subwoofer is approx 5db higher than the rest of the channels.

So the bass is not neutral but attentuated by 5db. No house curve is applied. I'm using an ECM8000 mic (using calibration file from this forum).

I've attached a picture below showing this in effect (subwoofer + center; XO 120hz).

Another observation is that the SPL-meter is around 2.8 db lower than the spl-reading in the RTA. I suppose there is a good explanation - only I, who am not a techie don't know it.


Best regards,
Mikkel

 

I tried calibrating each speaker using the SPL-meter and a pink PN-signal played through jriver media center.
Oddly, when I go to the spectrogram, using a pink PN-signal, the subwoofer is approx 5db higher than the rest of the channels.

You cannot set the relative levels of the speakers and sub using a full range test signal. If you take an SPL reading from the subwoofer alone (other speakers off or disconnected) using a full range signal it will produce a lower SPL reading than a main speaker playing the same signal (if level matched) because the frequency range the sub is reproducing is much less, so there will be less energy. The sub reading would be further reduced by the SPL meter weighting curve (see end comment).

AV receivers use test signals that cover a limited frequency range, for the main speakers the signal rolls off below 500Hz and above 2kHz, for the sub it rolls off below 30Hz and above 80Hz. The levels of the test signals are adjusted so that they produce the same SPL readings when the speakers are matched. If you want to use a full range test signal you need to look at the RTA plot and adjust the speaker/sub gains so that the flat parts of the respective responses are at the same levels. This can be tricky as the levels will vary with frequency, but overlaying the plots will let you find a good compromise.

Whilst it would be nice if a single number could encapsulate the overall level of a speaker this is not the case in practice, since the level varies with frequency differently for each speaker depending on what surfaces and materials the signals from the speaker encounter on the way to the listening position as well as the characteristics of the speakers themselves.

Another observation is that the SPL-meter is around 2.8 db lower than the spl-reading in the RTA. I suppose there is a good explanation - only I, who am not a techie don't know it.

SPL meter readings have weighting curves applied, by default C weighting. See the SPL Meter help for more info abut the weighting curves. The RTA reading has no weighting curve so will always read higher than a weighted SPL meter reading, how much higher depends on the frequency range of the speakers.

 

Thank you for the explanation. I get the general picture - at least I hope I understood your reply correctly... otherwise I stand corrected.

So if I want to use the pink PN-noise (-20 RMA) I should do following:

1. Send the signal to the subwoofer only, and then afterwards the other channels
2. Use the RTA-plot to align the SPL of the speakers as much as possible

One question then:

Jriver sends all bass (XO at 120hz, 24db filter) in all channels to the sub. It means the sub-signal is actually only the portion of the pink PN-signal that is below the applied filter. Should I instead send the full-range signal to the subwoofer before applying the cross-over filter?



Best regards,
Mikkel

 

Jriver sends all bass (XO at 120hz, 24db filter) in all channels to the sub. It means the sub-signal is actually only the portion of the pink PN-signal that is below the applied filter. Should I instead send the full-range signal to the subwoofer before applying the cross-over filter?

Either is OK since you will be using the RTA.