Is the cakewalk meter peak reliable? Seems not

[Marcello]

Hi there, I always use the cakewalk meter on each track to see if peaking and track is running too hot.

I usually try to maintain each track around -6 db or maximum -4db.

I noticed that if I use the youlean loudness meter 2 plugin on the same track and I compare the highest peak with the cakewalk embedded meter, there is a difference of 1db or so.

I believe I should maybe rely on the Youlean meter for more precise measurement?

What do you reckon?

[John Vere]

As far as True Peak reading yes use the You Lean meter if you want a more accurate reading. 
If you read the info on the web site it explains why it is more accurate. I believe the term is sample accurate. 

I actually have ran tests and funny enough I’m just reviewing and testing Sonars Concrete Limiter. 

To test take a Brick wall limiter and set it at -1.0 db. Feed the track or bus a signal that goes over -1.0 db . Now put a bunch of analyzer plug ins after the limiter including the You Lean meter.
Observe that all the other analyzers are reading exactly -1.0 db as well as the Cakewalk meter. 
But the You Lean might show-0.9 db or even as low as -0.6 db if the limiter is garbage. 
Cakewalk/Sonar meters are actually as accurate as all those other Analyzers. Just not sensitive as the You Lean. 

It is actually one of the features we overlook.You’d be surprised that the majority of DAWS don’t have very good meters. 

Edited April 11 by John Vere

[Marcello]

I see, 1 DB difference is a lot though.

[David Baay]

I was incidentally comparing Youlean and Sonar's numeric peaks recently and was actually surprised to see them reading identically because I'm pretty sure CW meters are not "true peak" (i.e. processing the samples as though converting to analog to find intersample peaks rather than just reporting the raw sample level). It's more likely the material I was metering just didn't have significant intersample peaks. I suspect calculating true peak on every track in a big project might add significant CPU load and would not be worth the overhead - maybe useful if it could just be selectively enabled on the Master bus which would usually be the only place it matters. 

In any case, I agree 1dB is a lot, and would make me suspect there might be some additional gain somewhere. Youlean is in the FX bin so any make-up gain, volume automation or post-FX Prochannel plugins could alter the track peak after Youlean sees it.

[chuckebaby]

1 hour ago, David Baay said:

I'm pretty sure CW meters are not "true peak"

This would be my guess as well. Youlean uses True peak (Measures ISP- Inter sample peaks) Cakewalk meters are peak meters.

The extra 1dB your seeing is probably the ISP.

[Mr No Name]

cakewalk would be measuring DBFS levels, youlean would be measuring measuring LUFS, as mentioned the intersample peaks are obviously measured differently aswell, cakewalk may be more accurate infact. imo.

Interesting fact, some mastering engineers say not to worry too much about intersample peaks, some will go as far as +1 over, lets the audio sound more natural after limiting.

 

only spotify streaming people worry about these things, some people just worry about playing the music on PA's or in nightclubs.