PRE MASTERING
I export (from Reason, but any DAW will do!) an audio file for EVERY instrument..... the list may look like this:
Kick 1, Kick 2, Snare 1, Snare 2, HH1, HH2, Sub Bass, Warp 1, Warp2,
etc. etc.
So, when these tracks are loaded into a DAW the song file should play as it does in the sequencer (I use Sonar for this part). Please bear in mind that this is part of what I call my mastering process. The tune is already mixed appropriately.
I then apply Waves or Wave Arts plugins (usually a limiter) to sort out any transients that I couldn't see in Reason. It's important to see your audio as well as to hear it. Sometimes a hi hat or snare may have a very loud transient that doesn't necessarily show it's face until you're mastering.
Basically I put a limiter on all kicks and snares and set it so it's ONLY JUST working on the sound. Virtually no gain reduction. You may want to export all snares to one track if there are many layered snare sounds, this way you can give it one volume setting. It's difficult to get an idea of how loud your snare is if it's made up of three snare sounds....
I then use a hi pass filter on every sound (apart from the kick and sub) and set them so they are only just cutting the lowest audible frequency from each sound. For hi hats this may be around 1k, for a snare it's more likely to be around 150hz.
This process alone makes the mix sound better, and it also prevents me from 'tinkering' with synth sounds etc. It helps me 'draw a line' under the creative process, because I can get quite obsessive!
I then make sure the track is peaking at around -6dB, for no particular reason and then export it as a WAV at 16bit - 44.1khz.
MASTERING
I then load it up in Soundforge and have a look to see if there are any ridiculously loud transients, which by now there shouldn't be. If there are then I will find out what it is (maybe when two powerful hi hats or other percussive elements play at the same time or whatever. This has happened to me) I'll usually mix these sounds as a stem and limit the stem. (ie. do a sub mix of all the hi hats or percussion from Sonar and limit only the loudest part of the stem).
For mastering I use izotope ozone. The EQ usually shows a fairly big hump in the low bass area (which is fine!) and a drop off towards the top end (also fine). Sometimes my mixes can be a little barren in the low mids to mids so I may boost this frequency (never more than 1 or 2dB).
I usually remove a little top end and bottom end. The top end gets put back by the harmonic exciter in izotope and the bottom end gets put back by Waves Max Bass. Only a little, not all of it!
With the exciter I usually turn the mix all the way down and the amount all the way up, which gives me a slightly more gritty sound, I then add a very small amount, to replace the tops I reduced in the EQ stage. I usually EQ using a shelf so I'll add a little of the hi mids too.
I tend to use the multiband compressor in izotope very sparingly, just so it's only just touching the sound. I use the Waves multimaximizer for that.
The only other thing I do in izotope is the multiband stereo imaging. You might find that your hi hats or other top end instruments sound better with a little imaging enhancement. I leave the bottom end alone, add a little to the low mids, maybe 1, then 2 for hi mids, then 3 for top end.
I leave the mastering reverb and limiter well alone.
I then add a little Waves Max Bass to replace the tiny bit I took out in EQing. With the dynamics settings maxed out I usually set the Max Bass to a freq of 79, with the filter reaching all the way into the subs. I'll add a tiny amount, barely audible but noticeable when removed.
Then the Multimaximiser happens. I usually have to turn the priority for the top end and bottom end up quite a lot, so that the unit doesn't only compress the hell out of these bands. I then try to get it so that it's limiting most parts equally, then drop it back a little so it's not working ALL the time. We like dynamics, right?
Final stage for me is the ultramaximiser. No more than 3dB of gain reduction with the brick wall limiter set to an output of -0.2dB so that nothing can reach 0dB - that's the important bit here. I have converted stuff to mp3 before and it has then shown signs of clipping so maybe treat those slightly differently (apply a limiter to the wav, save it as MASTERED (or whatever!) then apply a different limiter after you've saved as mp3)
It took ages to get a work flow sorted properly so I hope this helps!!
And no, it's not proper mastering but it works for me. I am skint, this is just what I do.... I'm not trying to make any laws up here!
