I've been using Cakewalk since the DOS version, I have over 25 years experience in music technology and over 25 years in PC technology (day job). But I can't figure this out ...
 
I've been working with a well known pro recording studio in Nashville.  I imported WAV files that they exported from Pro Tools. (note, I have noticed this same thing happening with commercially recorded .mp3s downloaded from Amazon for example ... I just never quite believed it or looked into it before since those files were mp3s and at a lower resolution)
 
The waveform image in Sonar clearly shows their wave files peak at -.2 Db, yet when I play them back Sonar shows the peak at +2.8 Db, but I do not hear any distortion, the sound is crystal clear and strong.  Their waves are not "brick walled".  I have verified that all my tracks, busses and output sliders are set to 0 Db, there are no FX, inserts, sends or returns.  This is being output direct to the hardware buss, but also holds true if output to the Sonar Master buss, there is no difference.  My Sonar meters are set to show Peak values (not RMS).
 
My soundcard's hardware meters (Tascam FW1884) are peaking in the red, and seem to agree with Sonar meters.  My soundcard is set up to show red peak only when exceeding 0 Db.  Again, there is no audible distortion.
 
HOWEVER, when I record my own audio in Sonar, and when I make my own mixes, any peak that hits 0 Db or above DOES cause audible distortion (also not "brick walled"). 
 
Bottom line, I think I'm losing the loudness war.  I think maybe I'm having to mix about -3Db lower than commercially available material, or else I have distortion.
 
Does anyone have ideas what is happening?  How I might resolve this?
 
PS. Let's not get into a loudness war debate here, this is really a technical question.
 
Thanks!

Check your Pan Laws.

A 3dB difference is a bit of a give away, at least that's where my thinking takes me.

Interesting ...
 
My pan law is set to the default ...

signal that’s panned hard left or right, and no dip in output level in either channel when the

signal is center panned."
 
 
 
and the wav file from the studio is indeed on 2 tracks, panned hard L & R, so that might explain the 3Db boost on those tracks (yes?) ... but if those tracks are being boosted into the red, shouldn't I be hearing distortion there?
 
Also I don't think this explains (or does it?) why an imported mp3 (single stereo track) from Amazon sounds about 3Db louder than my comparable self-produced track (when A/B-ing between them, I normally have to set the mp3 about 3Db lower than mine for comparable listening, even if my track is brick-walled nearly as much as the mp3)
 
Thoughts?
 

Check your entire signal chain and gain stage. If your audio file peaks at -0.2dB and there is nothing on the track and master bus effects bin and there are no sends (pre or post fader) enabled and if the main outs are set to 0dB, it will play back at the same exact level.

Also check your offset mode.

CJ

yoursong


I've been using Cakewalk since the DOS version, I have over 25 years experience in music technology and over 25 years in PC technology (day job). But I can't figure this out ...
 
I've been working with a well known pro recording studio in Nashville.  I imported WAV files that they exported from Pro Tools. (note, I have noticed this same thing happening with commercially recorded .mp3s downloaded from Amazon for example ... I just never quite believed it or looked into it before since those files were mp3s and at a lower resolution)
 
The waveform image in Sonar clearly shows their wave files peak at -.2 Db, yet when I play them back Sonar shows the peak at +2.8 Db, but I do not hear any distortion, the sound is crystal clear and strong.  Their waves are not "brick walled".  I have verified that all my tracks, busses and output sliders are set to 0 Db, there are no FX, inserts, sends or returns.  This is being output direct to the hardware buss, but also holds true if output to the Sonar Master buss, there is no difference.  My Sonar meters are set to show Peak values (not RMS).
 
My soundcard's hardware meters (Tascam FW1884) are peaking in the red, and seem to agree with Sonar meters.  My soundcard is set up to show red peak only when exceeding 0 Db.  Again, there is no audible distortion.
 
HOWEVER, when I record my own audio in Sonar, and when I make my own mixes, any peak that hits 0 Db or above DOES cause audible distortion (also not "brick walled"). 
 
Bottom line, I think I'm losing the loudness war.  I think maybe I'm having to mix about -3Db lower than commercially available material, or else I have distortion.
 
Does anyone have ideas what is happening?  How I might resolve this?
 
PS. Let's not get into a loudness war debate here, this is really a technical question.
 
Thanks!

rified that all my tracks, busses and output sliders are set to 0 Db,

What happens if you import this same Wav file into a different audio application?
A 32Bit Float file can hit peaks above 0dB with no distortion.  That doesn't explain the mismatched waveform (showing -0.2 peak).

Ed, he's already said that the 3db increase happens even if he outputs directly to hardware. I assume that he's referring to the little red flag that indicates the file is playing back at +2.8db, so the amplitude is established at the track level, even if there were subsequent adjustments or effects at the master bus.

I would have expected the original files to be mono and meant to be panned hard L/R. But if they were stereo files, importing them into a mono track could account for the 3db increase.

Missed that part Bit.

Repeating a bit of the OP ... Yes, I have zeroed ALL faders, checked the signal chains, removed all fx, inserts, sends, etc.

Also I have no offsets.

It seems the panning law is the most likely cause of the extra 3Db boost, since these are 2 tracks panned hard L & R.  However, with such a large boost over peak, why does this not cause audible distortion.  Any ideas?

Thanks

I've done some testing on this, using the same 2 mono tracks that comprise the final stereo mix.
 
1. Panning Test:  As described in the OP, when the tracks are panned hard L & R as intended, the meters show a peak in the red of +2.8Db, there is no audible distortion, and the waveform display shows peaks reaching -.2Db
 
When either track is panned to the Center, the meters show a peak of -.2Db, consistent with the visible waveform display.  When the tracks are panned somewhere between, then the peak is likewise somewhere in between -.2Db and 2.8Db
 
Conclusion - Certainly seems to be the pan law at work here, giving a boost of 3Db to a track panned hard L or R.
 
2. Bouncing Test:  I bounced the 2 panned mono tracks to a single stereo track.  The new stereo track also has a peak in the red of +2.8Db, and no audible distortion, it sounds great. 
 
The waveform display for the new track does show peaks hitting +2.8Db [As stated before, the mix is not a brick wall mix. And so the peaks are nearly all transients from kick and snare hits].
 
3. Export / Import Test:  I exported the 2 tracks to a 192K mp3 (format specified by the Taxi website where I post some music).  Then I re-imported the stereo mp3 back into Sonar and played it back.  The peaks are exactly at 0 Db, and once again, it sounds great, no audible distortion.
 
The waveform display shows peaks hitting exactly 0Db.
 
______________________________________
 
Bottom line ... This is not creating a problem per se, so no reason to change my Sonar pan law setting. But I would like to understand how these tracks can go nearly 3Db into the red without distorting, mainly so I might find a way to increase the levels of my own mixes to be more commercially competitive in the genres I produce (note ... I didn't start the loudness war, I don't like it, but it's a fact of life IMO, that isn't going away any time soon). 
 
If no one has an answer, then at least this posting might help other users who run into similar conditions with their projects.
 
Thanks