Sonar's RMS meter display doesn't follow the AES17 specification. According to the spec, a 997 Hz full-scale sine wave should read 0 dB RMS, but on Sonar's meters it reads -3 dB RMS (technically, it's probably reading -3.01 dB but the meter resolution isn't that high). I've already submitted a feature request to Cakewalk to have Sonar either conform to the AES17 spec, or to let the user decide where to put 0 dB. Anyone else feel like adding their voice to the request?

I've also requested this - it would allow for the meters to be used more easily as K-meters if the difference between peak and RMS were "corrected" in this way.

Exactly! I was trying to calibrate some monitors to the K-Metering system, and I kept having to remind myself that Sonar isn't following the AES spec.

I agree. The effect on my work is that exporting audio seems to follow the AES17 spec in terms of the resulting dB so what happens is I tend to mix too hot because I look at the meters, forgetting they are really -3dB below what I'm hearing. That is, I cannot use the visual to see how hot my mix is. (of course I know some people will argue one should NEVER use the eyes - use the ears!) Thorne

I think it's fair to say that in the absence of a well-calibrated mastering envronment (which I'm sure not many can lay claim to), it is impossible to tell, by ear, how "hot" the final mixdown is.

Tarsier:
Thanks for submitting the feature request - how come not a lot of other folks have brought this up?

-D

Voxengo Span now uses the K-system... It seems to work pretty good.

ORIGINAL: Otis
Voxengo Span now uses the K-system...

And it looks like you can adjust the RMS reference point by 3 dB as well. Very nice.

Now it's 2016 and  several updates to Sonar are we accepting the master meter accuracy?
 

I'm not trying to be difficult, but I've been through AES17 a few times over the last several years, and I went through it again tonight, and I don't see anywhere that  it states that "a 997 Hz full-scale sine wave should read 0 dB RMS". To be honest, I'm not really sure what that means?
 
Any dB measurement requires a reference - we have dBu, dBV, dBm (heaven help us), dBFS, and so on.
 
The other fly in the ointment is that dB is, by definition, an RMS measurement, and it is not well suited to peak measurements.
 
The other fly in the ointment (who left the door open) is crest factor - the ratio of peak to RMS value of a periodic (or non-periodic) waveform. A sine wave has a peak instantaneous level that is 3 dB greater than the RMS value.

A 997 Hz sinewave that peaks at full scale in the positive direction, and does not reach full scale in the negative direction is considered 0 dBFS, or "all ones" or whatever you wish to call it. It has nothing to do with analog levels, and has very little to do with digital metering, except that in theory that is the maximum measurable level (ignoring all the inter-sample stuff of course.)

I've fed test signals into Sonar from calibrated test equipment and the meters have behaved - they are NOT VU meters, but they do provide reasonably accurate RMS and Peak readings. It's been a while since I've tested the metering, and I too would love to be able to set a threshold, e.g. 0 dB on the meters is some specific (arbitrary would be nice) dBFS value. That would be handy for folks using K-Metering, and in fact handy for the rest of us too.

I seldom pine for the days of analog any more - aligning 16 meters on one tape deck two on the others, and 48 meters on the console was NOT fun, but... if you were patient you could get all the meters to read 0 VU at the same level. And you could even choose your reference level (with in reason).

Nah, I'll stick with digital, and just hang a decent set of analog VU  meters off the 2-Mix!