Is it possible to display total RMS of a measurement?

[neodog]

The RTA window presents an RMS computation in a box in the upper hand corner of the graph window. When saving the aggregates as a measurement, the RMS gets added in the form of notes on that measurement:

Input RMS -30.60 dBFS
-31.5 dBFS C, -47.8 dBFS A
-30.6 dBFS 22 - 22k UNW
-56.2 dBFS >22k

The RMS unit/scale listed in the note (dBFS here) is determined ahead of time, by the choice of scale/unit in the RTA graph. If I go to the measurement window to look at past captures, potentially in different units there, all I have right now is the note. I don't know how to get the RMS for SPL or V, for example. Can I get REW to present these RMS calculations for the current measurement somehow?

 

[John Mulcahy]

In the Info dialog for a measurement you can see the "Data offset" that is added to convert dBFS to dB SPL and the input voltage that corresponds to a full scale rms input, with those you could manually calculate the values.

 

[neodog]

Okay, I'm seeing a "data offset" of 120db, and Input FS Sine of 182.5 mVrms. So for this data:

Input RMS -79.84 dBFS
-88.7 dBFS C, -89.9 dBFS A
-83.8 dBFS 22 - 22k UNW
-82.1 dBFS >22k

To get RMS SPL:
-79.84 dBFS + 120db = 40.16 RMS dB SPL

To get mVrms:
182.5 * 10 ^ ( -79.84 / 20 ) = 0.01858 mVrms (18.58 uVrms)

Does this look right? The 40 SPL seems about 3dB high, but the 18.58 uVrms It seems to correlate closely with realtime figures I'm seeing.

This isn't exactly trivial, so it'd be nice to have a feature for it. Maybe a tooltip when hovering over a dataset in the legend.

 

[John Mulcahy]

You will see a 3 dB difference in the dBFS values depending on the setting of the View preference "Full scale sine rms is 0 dBFS". REW takes that into account when applying the data offset.

 

[neodog]

Are the RMS values in RTA computed off of the sample stream instead of the FFT? I notice that the different windowing settings can push the spectral plot around quite substantially, but the RMS numbers never budge in response. That would help explain why there aren't any methods of recomputing RMS on existing measurements, since it depends on the realtime sample stream. In tracking down noise contributors, I find myself wanting to compute RMS on a particular frequency range, but there doesn't seem to be a way to do it.

 

[John Mulcahy]

They are generated from the FFT data, taking into account window effects and the distortion high pass and low pass settings, if used.