In part one we discussed some basics of audio metering, including the relationship of dBSPL in the world around us to dBFS inside the box.  In this part we’re going to take a closer look at how we measure level in Pro Tools, and how it relates to a good finished product.

In part 1 I mentioned that we calibrate our speakers to 82 dBSPL when playing pink noise at -20dB in Pro Tools from the Signal Generator plugin:

It turns out that -20dB is also our nominal operating level in post production sound in Pro Tools.  We calibrate any classic analog meters that we have such as VU or Dorrough meters to read 0dB on their meter when Pro Tools is playing -20dBFS.  This is also the case on audio meters on Sony DigiBeta decks and Sony HDSR decks; when playing their internal signal generator with the meters in “full” scale the reading is -20dB, but when zooming in to “fine” scale the same level is displayed as 0dB.

This means that in post production sound we operate with 20dB of headroom.  Of course the actual total headroom is usually a fair bit less than that since various broadcast network specifications require that audio not exceed certain levels like -8dBFS or -6dBFS.  This forces us to place a limiter with that setting on our overall mix.  Limiters of course are those tools that allow us to set a maximum level above which no signal can go.  It’s the perfect thing to keep us within network spec.  A good limiter choice is the Waves L2:

This limiter has the advantage of automatic release control (ARC) meaning that you don’t need to worry about any settings other than threshold and output; the attack time and release time are managed for you.  Just set both the threshold and output ceiling to your spec (-8dB, for instance) and you’re done!  This is also a very forgiving limiter that can attenuate your peaks by several decibels without any audible distortion, which is yet another reason why it’s a great choice.

So now you’ve made sure that your room is calibrated correctly, your meters are set up properly, and your mix won’t exceed the limits imposed by the network.  It’s time to get to work and keep your mix “in the pocket.”

In the music business it’s a neverending battle for loudness supremacy.  It’s become the new normal to squash absolutely all dynamic range and life out of music, leaving massive loud walls of unchanging noise.  Fortunately in post production sound there’s been push-back in the opposite direction and standards are able to be enforced by the networks.  Commercials, of course, have suffered the same fate as music.  That’s why they often sound so damn much louder than the show you’ve been watching.  But new network specifications for “dialog normalization” are forcing everyone to put their TV mixes back down to a happy place with healthy dynamic range and a consistent presentation from show to show.  This is a Good Thing™.

So the challenge for TV/film mixers these days is to keep their mixes “in the pocket.”  It’s all about maintaining a consistent and appropriate “perceived loudness” for the overall show.  One of the classic tools for doing this that’s been in use for years and is still #1 in my personal arsenal is the Dorrough Meter.

It is an analog meter that operates on a proprietary ballistics scale called “Dorrough Practical Standard.”  When we use the Dorrough Meter for Pro Tools we calibrate the 0dB point on the meter to match -20dBFS in Pro Tools.  The current reading on the meter shown above is right in the “sweet spot.”  The continuous chain of lit LEDs on the left part of the meter measures persistence of sound and is currently reading just below 0dB.  The peak indicator LED on the right is reading 11dB.  Notice how the meter has red LEDs in the middle and on the far right.  When our persistence reading is “in the red” in the middle and our peaks are “in the red” on the right, we’re probably pushing things a bit too much.  Of course we’re welcome to go into the red from time to time for big moments, but the point is to keep the average mix level lower than that.  If you were to use this one meter alone for your entire mix and make sure you stayed in the sweet spot, you’d probably have a great mix.  It really is that powerful. (Of course that’s an oversimplification, but I’m trying to make a point! :-) ) And one of the best parts is that Waves has created a plugin version of this meter!  A single real Dorrough Meter costs about $500, but for about the same price you can get the plugin and have unlimited Dorrough Meters!  I’ll take it!

Of course if we actually want to be sure that our mixes are in the Network-Approved dialog normalization sweet spot, we’re going to need a more powerful custom-tailored metering tool.  Enter the Dolby LM100 and Dolby Media Meters.  The LM100 is a 19″ rack-mountable hardware unit, and the Dolby Media Meter is a plugin equivalent (also available as a stand-alone program).  Both are meters designed to measure dialog normalization according to the ITU BS 1770.1 specification.  In English, it basically measures how loud the dialog appears to be throughout the show.

This is a more useful tool than the traditional Leq(a) measurement scale, which simply measures average signal level over time.  This is because people tend to consider how loud or soft a show is according to the dialog level, not the music or sound effects.  Also, any halfway-decent mixer knows that dialog is King and has mixed the show appropriately in the first place to be sure that all dialog is clear.  Therefore, measuring the average level of the dialog over time is more useful than just measuring the average level of the whole mix over time.  An action show with lots of guns and explosions would have a much higher Leq(a) average level measurement than a courtroom drama, but both should be able to have the exact same dialnorm measurement as measured by the Dolby Media Meter.

By the way, as of the Fall TV season 2010 the dialnorm value most widely accepted by television networks is -24dB, plus-or-minus 2dB, with peaks limited to -6dBFS.  This may sound like an incredibly narrow range, but remember that it’s a measurement of persistence over time, which means it’s just an average.  You can have loud passages of dialog as long as they’re counter-balanced by softer passages.

So how does the LM100 or the Media Meter know when they’re measuring dialog instead of sound effects or music?  Dolby’s answer distills down to “Magic.”  It’s fair enough, I suppose, for them to want to protect their intellectual property.  Also, it’s almost a completely irrelevant point anyway.  This is because it’s an agreed-upon standard that broadcast networks require, and mixes that meet this specification can and do typically sound great.  So whether or not the LM100 truly can discern dialog from sound effects, the bottom line is that mixes that read -24dB on the LM100 over the course of the mix tend to be right in that “pocket” that we all want to be in.

So to sum it up, a strong metering ensemble in Pro Tools is a combination of the Waves Dorrough Meter plugin and the Dolby Media Meter.  With these two you can get your mix in that “sweet spot” and be sure not to be rejected by a network for being outside of spec.