Broadcast levels workflow

Question

Hi there,
I teach Journalism production to undergrauate journalism students. We teach the Adobe apps: Audition, Premiere, InDesign, Photoshop.
In Audition, radio students construct stories with various recorded or sourced audio files in the Multitrack view, then mix them down to the Edit view for saving as MP3 files.
Naturally these source files vary in level and quality, depeding on where they came from.
I'm just wondering what other people's workflows are to get a neat, broadcast level mixdown with files of varying level, and how one workflow might be better than other workflows (or the ones we're using now). I'd like to settle on a "best practice" workflow to document in my teaching.
For instance, do you:
1) Apply normalising/compression to tracks in the Multitrack view?
2) Mixdown and then apply normalising/compression?
3) Mixdown and fix low parts "by hand"?
4) Something else?
Kind regards
Matthew P
Brisbane, Australia

Answer 1

Matt Petersen wrote:
4) Something else?
Oh, definitely that... In a nice, idealised student setting you can spend loads of time polishing everything after the event... but in the real world, stuff happens. For a start, if you have loads of different takes, all of different levels then what tends to happen is that each individual file is sort-of okay in itself, but not necessarily compared to any other file. So firstly you have to get all of these files to at least a vaguely consistent level, and that's where batch processing comes in. Assuming that the basic recording technique was okay, then batch normalize all of the individual files to about -1dB. This is way quicker than doing each one by hand, and it means that the next stage in the process you can generally whizz through a lot more easily without mucking about with levels all the time.
So now you can put stuff in the mix quickly and easily, and play it out directly if you are desperate - which at some stage you are going to be. If you want to make sure that the final output sounds a bit more homogeneous, then all you have to do is to put some multiband compression into the master stage, and hey presto, what comes out sounds pretty good. And when you have to redo things, or use different parts of what's there, it's dead easy to swap sections around from your files and still have instant, good-sounding output without doing anything else at all. And that's what most real-life radio production is about - good results fast, because the editorial production team won't wait.
So the basic rule is to get all your source material normalized first - and if anything else about it needs fixing, do that first too. And do as much of this stuff in batches as possible. One of the main reasons that Audition is used the world over for radio is because you can achieve a heck of a lot quickly in batch mode, and I'd say that any radio course that doesn't teach students what the real advantages of this method of working are isn't exactly doing their future employers much of a service at all.