The CALM Act was passed in order to address the constant viewer complaints that commercials were louder than the programming, and that levels fluctuated wildly from channel to channel.  People were tired of constantly grabbing the remote to turn things up and down.

Now mixes delivered for broadcast have to meet a precise spec to address this issue.  (For the driest reading of your life, head over to the ITU’s website and read about the current international standard, BS-1770-3). While this spec is generally a good thing, it can be hard to precisely nail the exact required level with every mix.  This leaves us open to rejection at QC.  Not fun.

Today we finally have a variety of hardware and plugin options for automagically adjusting our final mixes to meet broadcast loudness specifications.  Options from iZotope, Flux, and Nugen are available to render our composite mix faster than real-time and make tiny level adjustments throughout to guarantee compliance.

But currently they have one fatal flaw…

These plugins only work on the full mix.  But of course, we still have to deliver split stems as well!  And the expectation has always been that the split stems, when combined, will exactly match the full mix.  But if our full mix has been auto-processed to meet loudness requirements, it now has lots of little loudness nips and tucks throughout that don’t exist in our splits.  And you can bet that network QC is rejecting split stems delivered in this fashion!

Of course, simply running the same plugin on each split stem is meaningless; the overall level & dynamics of any split stem bear little if any resemblance to the dynamics of the full composite mix in most cases, so necessarily any separate processing will yield completely different and unacceptable results.

What is needed is either:

1. a method of “side-chaining” the auto-leveling plugins from the composite mix stem while processing the split stems for delivery, such that the micro-level adjustments determined by the comp mix are exactly replicated in the splits
2. or a method of saving a “snapshot” of exactly what level adjustments the plugin applied to the comp stem while processing it, so that the same exact level tweaks can be applied to the splits

In the meantime, these plugins are only useful if you don’t have to deliver split stems, or if network QC doesn’t notice.  I’m not willing to take that bet, so unfortunately for now these aren’t an option.

Let’s hope that the software developers can address this issue and give us the full capability we need to deliver complete stems packages that meet spec easily and quickly!