If you read various music recording and production forums, over time you'll notice that certain arguments are played out more frequently and with more passion than others. Does analog sound better that digital? Is recording at 192KHz really any better than 44.1? Are major labels worth your time any more? Actually that last one isn't much of an argument.

Another perennial favourite is whether or not your should process your mix bus. In other words, should you put EQ, compression etc. across your entire mix, or should you leave that for the mastering engineer? And if you do process the mix bus, should you do it while you're mixing or add it in at the end?

This is one of those ones that really just comes down to the way you work. There seem to be more people in the "dry" camp, who believe you should put nothing across the mix bus, and that any problem you are trying to fix by doing that should be achieved by working with the individual tracks instead. I actually fall into the smaller camp. In fact, I mix into a compressor and EQ right from the start, and I can't actually remember the last time I changed the settings on my EQ, so it's permanently set to the same curve.

Why do I do this? Well, I actually think it makes mixing quicker and easier, especially if you're mixing an entire album. I think mixing into a little compression (and we really are talking a little here - 1.5:1 ratio with no more than 2db or reduction) just helps to glue things together a little and stops you over compressing individual tracks. And mixing into the same EQ curve just helps to tie the mixes together a little, like a common running theme. My mix bus EQ is an API 5500, which is essentially a pair of 550Bs, but with an added "Range" control that lets you set the boost or cut in 0.5db increments (unlike the fixed 2db of the 550B). My curve looks like this:

Lo: +2db @ 50Hz

Lo-mid: -1db @ 500Hz

Hi-mid: flat

Hi: +2db @ 15KHz

Because I always mix into this curve (with rare exceptions), I know exactly what it sounds like on my source material. Of course, you could argue that this same curve could equally be applied during mastering, and I'm making changes that can't be undone later. Why do it early on in the process when it could just be added later if it needs it?

I think that's actually the whole point. I still believe that you should commit to sounds early on - make decisions early in the process. Back in the day, you spent a lot of time getting the sound right at source, because after it was committed to tape there really wasn't much you could do to fix it. Now we have so much incredible technology at our disposal that it can be overwhelming, and you can spend the rest of your life tweaking a mix or master, during all of which time no-one is listening to your music.

I'm not going to argue the point - it's just personal preference and neither way is wrong. But this is what works for me.