And while I've mentioned AES17, it really annoys me that Sonar's meters don't conform to it. With all of Cakewalk's advertising hype about precision engineering tools, it is disappointing that Sonar's meters don't indicate RMS values in accordance with the AES17 standard.

What is it in the AES-17 spec that you believe we don't adhere too? Our meter text indicators display with 0.1 dB accuracy. The AES-17 spec [paragraph 4.3.4] calls for true RMS metering with 0.25 dB accuracy. We do better than that.

ORIGINAL: Ron Kuper [Cakewalk]

And while I've mentioned AES17, it really annoys me that Sonar's meters don't conform to it. With all of Cakewalk's advertising hype about precision engineering tools, it is disappointing that Sonar's meters don't indicate RMS values in accordance with the AES17 standard.

What is it in the AES-17 spec that you believe we don't adhere too? Our meter text indicators display with 0.1 dB accuracy. The AES-17 spec [paragraph 4.3.4] calls for true RMS metering with 0.25 dB accuracy. We do better than that.

The problem is not about being within .1 dB of accuracy, it's about being 3 dB too low on the RMS display. Having said that, let me say that I disagree with the spec, and I think Sonar does it correctly. Also, I think the spec is ambiguous when it comes to defining RMS display. But after discussing it with other engineers, and also according to Bob Katz's K-Metering system which references AES17, a full scale sine wave should read 0dB FS peak and also 0dB FS RMS. Sonar displays a full scale sine as -3dB RMS.

Here are the relevant passages from AES17:

AES17-1998 (r2004)

3.3
full-scale amplitude
amplitude of a 997-Hz sine wave whose positive peak value reaches the
positive digital full scale, leaving the negative maximum code unused.

NOTE In 2's complement representation, the negative peak is 1 LSB away
from the negative maximum code.

3.3.1
decibels, full scale
dB FS
amplitude expressed as a level in decibels relative to full-scale
amplitude (20 times the common logarithm of the amplitude over the
full-scale amplitude)

NOTE The rules of the International System of Units (SI) require that a
space appear after the standard symbol dB.

3.3.2
percent, full scale
% FS
amplitude expressed as a percentage of full-scale amplitude (100 times
the amplitude over the full-scale amplitude)

3.4
full-scale signal level
FS
signal amplitude relative to the full-scale amplitude expressed in
decibels, full scale or percent, full scale

NOTE Because the definition of full scale is based on a sine wave, it
will be possible with square-wave test signals to read as much as +
3,01 dB FS. Square-wave signals at this level are not recommended
because tilt or overshoot introduced by any filtering operations will
cause clipping of the signal.

Unfortunately, it is not explicit as to what the RMS value should read, and since (when measuring voltage for example) the RMS value of a sine wave is 3 dB lower than its peak value it is natural to assume that your audio meters should read the same way. And that's what I thought for a long time, until I read Bob Katz's article on his K-Metering system.

Back to AES17, section 3.4 notes that a square wave can be as much as 3.01 dB more than a full scale sine wave. This could only happen if the RMS reading of a full scale sine wave is 0 dB, Or in other words, you reference the RMS dB scale to a full scale sine wave. Section 3.3 states that full-scale is defined using a sine wave, and section 3.3.1 states that the dB scale is referenced to that sine wave. Therefore according to AES17, 0dB FS RMS = 0dB FS peak for a sine wave at 997 Hz and maximum digital magnitude before clipping.

If you go to Bob Katz's website where he outlines his K-Metering system he states:

Bob Katz
The peak and average scales are calibrated as per AES-17, so that peak and average sections are referenced to the same decibel value with a sine wave signal. In other words, +20 dB RMS with sine wave reads the same as +20 dB peak, and this parity will be true only with a sine wave.

In Sonar, a -23 dB RMS sine wave will read -20 dB peak, whereas according to AES17 (and Bob Katz) it should read -20dB for both peak and RMS scales.

On that website there is also a link to Pink noise at -20 dB RMS FS which Sonar's RMS meters display as -23 dB RMS.

As I mentioned above, I like Sonar's method of RMS calculation, but it doesn't conform to AES17. It would be nice if Sonar would allow the user to specify which reference point would equal 0 dB RMS, as that would not only make Sonar be able to follow the AES spec, but also make using the K-Metering system possible by specifying -20 -14 or -12 as your reference point depending on which scale you want to use. It would also allow Sonar's default behavior to remain unchanged with respect to metering, but allow users to calibrate their meters to how they see fit.

ORIGINAL: Ron Kuper [Cakewalk]

As I mentioned above, I like Sonar's method of RMS calculation, but it doesn't conform to AES17.

I'm not disputing that we don't do K-Metering, but I'm still not seeing how AES17 mandates a sine wave should should the same RMS value as peak value. If you measure true RMS a sine wave whose peak value is 0dB FS registers at -3dB RMS. That's simply how the math for RMS works out. Even though Bob Katz quotes this practice as being "per AES-17", I don't see anything in AES-17 that actually states this, do you?

No I don't, which is why I think the standard is ambiguous. When Katz referenced AES17 as being the reason 0 dB RMS = 0 dB peak for a sine wave, I had to look it up. Even after I looked it up, I was still convinced Katz was wrong since as you point out the RMS calculation of a sine wave will be 3 dB lower than its peak. I inquired about this discrepancy on an audio forum and all the engineers were adamant that AES17 clearly stated that a meter should display 0 dB peak as well as 0dB RMS for a sine wave. It certainly wasn't clear to me...

So I sat down with AES17 and really tried to reason it through. Here is how I finally explained it to myself:

Every decibel scale is in reference to some value.
When the signal you are measuring is equal to the reference value, you get 0 dB in whatever scale your reference is in.
now I paraphrase AES17, which I quoted above
3.3.1
dB FS is the level relative to full-scale
so what is full scale? go to section 3.3
3.3
Full-scale is a 997 Hz sine whose peak reaches positive digital full scale.

the kicker is section 3.4 which defines full scale signal level
3.4
full-scale signal level. FS. Signal amplitude relative to full-scale amplitude.

Section 3.4 defines signal level in terms of decibels full-scale which is defined in section 3.3.1 which defines full scale in terms of a sine wave which is defined in 3.3. In a nutshell, 3.4 references 3.3.1 which references 3.3 (which is why I think the spec is confusing and ambiguous).

So in 3.4, it defines the reference level, but it doesn't say anything about RMS or peak. But if we are defining what full scale RMS is (which all other RMS values are then calculated from) it is referenced to a full scale sine wave as defined in 3.3. Also, full scale peak is referenced to a full scale sine wave. Therein is the crux of the matter. dB FS RMS and dB FS peak are different scales with separate reference levels--those reference levels are just both defined as being full scale sine waves. That is why according to section 3.4 a square wave can read as much as +3.01 dB FS. It doesn't say whether peak or RMS, but the only way that could happen is if it were the RMS reading.

Since the spec never specifies RMS, when we go to find out the reference value for our dB FS RMS scale we turn to AES17 section 3.4 (amplitude expressed in decibels, full scale) which leads us to 3.3.1 (amplitude relative to full-scale) which leads us to 3.3 (full-scale amplitude is a sine wave).

Therefore our reference level for RMS is the full scale sine wave itself, and when the digital RMS level of the signal being measured is equal to the digital RMS level of a 997 Hz full scale sine wave we get 0 dB FS RMS.

In other words:
the reference level for 0 dB FS RMS is .707
the reference level for 0 dB FS peak is 1.0

That is what I have come to believe the AES17 spec says, in its ambiguous and roundabout way. The paragraph in blue above is probably where the debate comes in. Do you choose your reference level according to the convoluted AES17 method, or do you just take the root-mean-square values of the samples? Sonar just takes the RMS values, which isn't according to AES17 (as I see it).

Do you happen to know the author(s) of the spec? Perhaps you could ask them to clarify.

I'm sure it has something to do with the speed/sensitivity that the meters operate at also.
But actually I have used SF and SONAR as described in this thread and have seen SF clip when SONAR did not.
When looking at the wave I saw no clipping (that I could see) so I wouldn't worry about.
I do normalize the track in SF -0.1 -> -0.5db and that seems to fix any problem I've had.
I use Audacity now a days for wav file editing and I see no clipping as SONAR states.  I trust Audacity in this respect also.  But I've also started mixing not quite as hot as I used too.  I give myself some headroom.
PC

I get identical response in SF9 and Sonar 8.3 (even at -0.1dB on a loop I tested).

I wonder if people getting different response in SF are using the VU/PPM meters instead of the Peak meters (VU/PPM displays +5.4 on my test loop)? Check to make sure you're using Peak meters in both Sonar and SF.

drewfx

Wow - I was having flashbacks there for a moment....

brundlefly

Ancient thread alert: anything before post 37 is approx. 4 years old.

You are the man Brundlefly. I should have picked up on it when they were talking about Sonar 5. I will now look at dates before I start cramming my head full of useless info. lol
 
Thanks again Brundlefly.