Go above -1db with Audition/Special/Mastering plugin

Chris.Ludacris wrote: Hi Paul, Thanks for answering. I totally understand your point of view. But why do most artists opt for going beyond -1db anyway?

Hi Chris, and you're welcome.

It's sort of old-school thinking. In fact (and with respect) - many people still don't realize that Peak level or maximum amplitude has nothing to do with Loudness or perception.

A few other things to keep in mind:

- If the master is going to be converted to a lossy codec (AAC, MP3), you must supply a certain amount of headroom to prevent clipping  and/or distortion that may occur due to data compression. For instance a WAV file with a -0.1 dBFS ceiling will most certainly contain Intersample Peaks when converted to MP3. In essence the distribution copy will be clipped. The headroom in the lossless master will prevent this. I would say you need at the very least 1 dB of headroom before converting. In the broadcast world 3 dB of headroom is widely recognized and recommended, especially when converting to MP3. It's all about the short comings and results of lossy data compression. In my world of intelligible spoken word/dialog processing, I generally target -2.0 dBTP at the source,

- Loudness no longer matters. Sure, this is a subjective statement. However, the point is loud mastered tracks are most certainly going to be turned down by the delivery platform (Loudness Normalization, SoundCheck, etc.) In essence, it no longer makes sense to push for a loud master, which in most cases will be squashed and crappy sounding due to heavy processing. Producers should be focusing on well balanced mixes and high quality masters with full use of dynamic range. And don't forget the headroom issue.

Earlier I mentioned delivery targets which include an average Loudness descriptor (aka Integrated/Program Loudness) as well as a True Peak max. Compliance is important, and it's all based on what is specified in the spec. for a particular delivery platform. Music production best practices may be different when compared to something like spoken word or dialog/broadcast processing. OTOH the general concepts are consistent with regards to overshoot prevention and perceptual loudness optimization.

-paul.

@produceNewMedia

Chris.Ludacris wrote: Hi guys, Thank you for answering I found a solution, the WAVES L3 Ultramaximizer. Basically you just set the Out Ceiling to -0.1db, and then play with the threshold to increase the volume. The reason for which I'm trying to get to -0.1db is that most artists tend to do that (loudness war thingie ) and it feels weird having a track that is not as loud. And most artists go above the -1db limit that I was used to as well. Thanks, Chris

The maximum Peak Amplitude has no bearing on the perceptual Loudness of the audio. It's simply an indication of signal level, or a representation of voltage. In fact, it's possible for a piece of audio with a maximum Peak of something like -2.0 dBFS to be perceptually louder than something with full scale Peaks at 0dBFS or slightly lower (-0.1dBFS, etc.).

Think about it -  let's say your audio has an inherent wide dynamic range. In other words, wide variations between low level and high level passages. The Peak ceiling may very well be -0.1 dBFS. However, the overall average, perceptual loudness of the entire clip may be quite low. Conversely, if the audio has been heavily processed (compressed, limited, etc.) reflecting a reduced dynamic range with a maximum Peak ceiling of -2.0 dBFS, - in most cases it's going to be  much louder than what I previously described.

In today's world the maximum Peak or True Peak is specified in ubiquitous Broadcast and Distribution specs as a clipping failsafe. The average (Integrated/Program) Loudness descriptor reflects how we perceive audio with regards to loudness. It is equally specified in the described specs.

Bottom line is the Mastering plugin is leaving headroom, and that's totally fine. This is especially important if and when the master is converted to a lossy codec such as AAC or MP3.

-paul.

@produceNewMedia.

I've had a look and, like Steve, all I can come up with is using the overall output gain control to boost things--but, since this specifies the gain to be added rather than the maximum output, it would indeed be hit and miss. 

I suspect the problem is that altering any of the many settings can affect the output level and it would be very easy to get above 0dB(FS).

If -0.1 is really important to you I fear normalising remains the best way.  However, I'd add the warning that, if your material will eventually be converted to MP3, there's a strong possibility that some samples will be clipped in the conversion process.  Even when "mastering" totally manually, I limit (pun intended) myself to around the -1dB level anyway.