Sonar says 'not clipping'. Soundforge says 'clipping'.

www.elementalaudio.com

download Inspector. it will tell exactly what is happening.

Also, to ensure everything is equal, bounce to track in Sonar.
Make sure the track looks/sounds right.
Solo that track so you know nothing else is being added.
Export

Edit: download the free version of Inspector, I think it is better than there pay version.

I guess it depneds on the algorithm each different app uses to decide if a file is clipping or not e.g. maybe Sonar will allow you 7 (or whatever) consecutive samples at 0dB before it gives a clip indicator, but Sound Forge only allows 3 (or whatever) consecutive samples at 0dB before it gives a clip indicator.

This all makes sense - thank you.

It doesn't sound like it's clipping... and I think wigworld has nailed the issue.

Keith - my favorite answer is absolutely almost - Some cheap playback devices have no tolerance for any samples at full scale - but most consumer devices made after 1995 (as I have been told) can tolerate between 4 and 7 samples at full scale without completely cracking apart. Sound Forge (and it's CD Architect partner) are conservative and CDA adheres to strict Red Book standards - Which I seem to recall may have more to say about number of full scale samples - anybody know the Red Book tolerance for this ?
-D

I think that you should rely on your ears. Loop the transient- If you can't detect any distortion, fuggedabout it.

ORIGINAL: piano39

I think that you should rely on your ears. Loop the transient- If you can't detect any distortion, fuggedabout it.

I agree with you and rely on my ears as much as possible, and it doesn't sound like it's clipping - I hear no distortion.

My concern, however, is the one mentioned above: different devices have different ways of calculating clipping. So if my files play back on an overly sensitive or older device, it could distort. Murphy's Law says it will be an A&R person's device that this happens on, too.

I guess the larger point is that I expected Sonar's meters to be as sensitive as possible and not have to rely on external plug ins or other softwares to get a 'true' read on what's going on.

T.

ORIGINAL: piano39

I think that you should rely on your ears. Loop the transient- If you can't detect any distortion, fuggedabout it.

Absolutely, Almost - some playback devices and or clients demand nothing hitting DBFS - even for one sample. Most of what I do gets broadcast on FM radio - the FCC has requirements for % moduation in AM and what I call FreqPeak in FM. Of course they have limiters and compression/expansion in front of the transmitter but - I don't want the chief engineer to tell me my recording is smacking the red too often on the processors.

Our ears are mostly sensitive to RMS or average loudness not Peak. In fact our ears are a poor judge of peaks as most peak measurement occurs across the full bandwidth of ten octaves. The four outside octaves of our hearing is significantly less sensative at these edges and can be very very deceiving depending on the level you are monitoring at. And depending on your meter ballistics you'll see what you can't hear because clips occur at tens of thousands of a second - we just can't 'hear' that fast.

With respect Piano39 - your ears are a poor judge of Peak / an excellent judge of what sounds good.

-D

[ i don't think that is the point. he is trying to understand an inconsistency between tools.
it is always good to try to understand these things, as it can often point to a bug somewhere.

and if not, then you learn something that might help in the future... ]

From a technical standpint, you are absolutely correct. To use this stuff correctly, you are forced to understand every nuance of the hardware, software, etc. It sounds like everyone on this thread is conscientious.

On the other hand, my point is, that you have to rely on your ears as the final judge.

Don't forget, you could have the exact opposite problem- all of the meters, etc say that you are not distorting, but in reality, you are. The more complex the signal chain- submixers, efx sends, etc- the more likely that this is to happen.

Dreamkeeper:
I use the PSP meters sometimes - I love 'em
-D

ORIGINAL: DonM

Dreamkeeper:
I use the PSP meters sometimes - I love 'em
-D

Yeah! They're so ... Vintage, hehe!

Technically, this is a quality control/ calibration issue. The answer is to use a known standard that is certified to be at zero db with unity gain. I think that there are calibration CDs on the market for just this purpose- it will have audio or WAV files with various frequencies recorded at various levels. ( I also think that I own one, but can't find it).

Import this file directly into Sonar. What does the VU meter read with everything set at unity?
Import into Sound Forge and do the same.

This should at least tell you if Sonar or Sound Forge are accurate.

If you send me the right WAV to do this, I will... but I ain't got time to be hunting down a callibration CD.

I had the same concern and recently went through the calibration for my system. There are some good tone generators as recommended link=http://forum.cakewalk.com/tm.asp?m=620029]here[/link].
I created a reference sine tone and check the meter of each tool including my hardware input/ouptut, and computer output.
I found that the meters in SF and Cakewalk are identical. 0dbfs in both tools also is 0dbfs of my soundcard. I use asio driver btw.
winamp and media player came in about +1db with the window control set to max.
I don't like the PSP vintage meter as the needle bounces too fast for accurate measurement, but it sure brings back memory of the good 'o days.

Micv

I have found the same thing. If you go to the Presonus site at the central station page here and go to the bottom of the page they have a 1K calibration sine wave there for download that is calibrated at -18dbfs.

If I play that wave file in Sonar I get exactly -18dbfs on the Sonar meters. Ditto in Sound Forge meters.

I my mind the two applications meters are the same.

John

SONAR will register clipping only if you go fully over full code, IOW if some float/double sample goes outside -1.0 - +1.0

Ron... I just got confused - do you mean 'any sample' when you same if 'some F/D sample' goes between -1 and +1? Isn't that quite some wide range in DP FP? Even before Full Scale? I apologize if I am really in remediation mode!

-D