I've seen it suggested that compressing the stereo bus is the key to getting a mix to come together and sound 'like a record'. Is this really the case, and if so, at what point in the mixing process should I be adding bus processing?
SOS Forum Post
Reviews Editor Mike Senior replies: I find that bus processing of various kinds does wonders in pulling together a mix, and I do usually mix through a selection of bus processors for that reason. But let's not get carried away — a great mix will benefit as much (or even more) from bus processing as a mediocre one, so having access to decent bus processors doesn't really let you off doing a decent mix, because these days most serious engineers have access to decent bus processing!
If you start with all your 'polishing' bus processes in place at the start of the mix, you're likely to work less hard at getting the basic mix right in the first place. I think this was one of the lessons to be learnt from the On-line Mastering Shootout listening tests we did here in the SOS office for our March 2006 issue. Trying to fix mix problems with bus processing is fantastically difficult — improving any element of the mix usually involves compromising some other part. So, if you don't do the mix properly to start with, you'll find it very difficult to sort out any problems later using mastering-style processing. It's important to do the very best you can with your mix before you switch in any bus 'polishing', otherwise your final results will suffer.
My general advice would therefore be to avoid bus processing for as long as you can with a mix, so that you get it sounding as good as possible without any extra help. However, there are a couple of exceptions I would make to this.
The first is that, speaking personally, I usually patch in a full-band bus compressor over the mix while I'm creating my opening balances. Whether you would find this suitable as well will depend on whether you use a pumping compression sound as I tend to. If you plan to, then I'd suggest mixing with the compressor switched in — your balance decisions have to be different if you are intending to hit a bus compressor in this way, so you need to be able to hear what you're doing.
The second exception concerns using EQ rather than compression on the mix bus. It's a little trick I learned from our interview with Spike Stent, one of my mixing heroes, in SOS January 1999 (www.soundonsound.com/sos/jan99/articles/spike366.htm). He patches in a really high-quality EQ over the whole mix, boosting the 'air' frequencies so that he doesn't need to do this using lots of individual lesser-quality channel EQs. I use my Drawmer DC2476 mastering processor for this, so that I can stay in the digital domain. Spike was using a Massenburg EQ in preference to his 'low-quality' G Series SSL channel EQ, so the quality difference for him is probably less than for the rest of us — the difference between a really nice EQ processor like the Drawmer and the built-in digital channel EQs in a digital multitracker or software sequencer is much bigger.
What I would say though, even with regard to both these exceptions, is that you should always make sure to record completely unprocessed versions of mixes along with the processed versions. That way, if anything goes wrong with the bus processing, you don't need to completely redo the mix — you can just reprocess the unprocessed versions.