In this second post, I'm going to walk you more or less through the process of mixing the Etudes. I recorded each piece as a "chunk" in Digital Performer. To simplify the mixing process, I created v-racks and routed each of the tracks through an auxiliary channel. I also setup aux channels for two different reverbs, one for handling the mix of all the different sources, and a master track. All of these aux tracks were in one v-rack. Here's a screen shot of the mix window:
My goal was to simulate different perspectives on the guitar - One right up front, another further back, and a third from rather far away. By blending these different perspectives, we get something that's closer to what the ear actually interprets than if we depend on only one.
For the room mics, I'd originally labeled and panned them from the perspective of the performer. I decided to mix from the perspective of a listener, so I had to switch the L and R channels for the room mics. Rather than go through 29 different chunks and re-pan everything, I used the Waves S1 Imager to switch L & R. I also decided to close the stereo field up by about 25 or 30%. I prefer, when recording, to get as wide a spread as possible. It's always easier to close it up (bring the channels closer to center) than to artificially widen it.
The next thing was to EQ out the frequencies I didn't want. For the most part, this just meant rolling off some of the sub frequencies (mostly below 100Hz), but I did pull out just a bit around 5k on the room mics. For this, I favor the UA emulation of the Cambridge EQ:
The close mic still had more of these really low frequencies than I wanted. You wouldn't even hear them without a subwoofer, but they're obnoxious at certain volumes with one. Of course, they also lend to the warmth of the sound, so I didn't want to just cut them even further across the board. I decided to use a multiband compressor to reduce the very low frequencies at higher volumes. For frequency dependent compression, the Waves C4 and C6 are hard to beat. I used this on the close mic and again on the aux for the mix of all mics. If you're familiar with the interface for this compressor, you'll notice that the compressor only triggers when the level at those frequencies is fairly high, but that the gain reduction when it does trigger is significant. Here's the C6:
The slight compression was applied to the room mics because dynamics tend to have less of a range from further away and this helped to reproduce that.
The next step was to apply some reverb to all of the tracks. I used a fairly short and warm reverb to do this using the Waves IR-L since the room I record in is (intentionally) almost entirely dry. If you look at the screen shot of the mix window, you can see that very little is applied to the close mic, a little more to the far mic, and even more to the room mics. This helps to create a sense of depth in the mix.
I also applied a bit of a longer reverb to the room mics using Waves TrueVerb. You can see that I first experimented with the UA EMT140 and decided against it. Again, the idea here is that a listener much further away from the performer will hear more room ambience than a listener right up close. I wanted the room mics to sound much further away than my physical space would allow and the extra reverb helps create that effect.
On both reverbs, I used a bit of compression to the signal going in. This is because most digital reverbs (including convolution reverbs like the IR-L) tend to really jump out in the high frequencies at higher volumes and be more subdued at quieter volumes. Real rooms aren't nearly as dramatic at the high end, so a compressor that could filter out (not respond to) lower frequencies was useful. The UA Precision Bus Compressor was just the tool for the job. I just used it to trim 1-2 dB off when the guitar got louder:
The last step in mixing was to try and "glue" these three different sources together. The first thing to do was balance the levels of the three sources. Since I wanted a fairly up close and intimate sound, I used more of the close mic, less of the far mic, and even less of the room mics. I then applied the C6 to tame the really low frequencies of the composite sound from all three sources. (See screen shot above.)
I would have then typically run everything through the UA LA-2A - a leveling amplifier that really does a brilliant job of pulling different sources together and which I use religiously on drums - and done some post EQ. I found, however, that the occasional fortissimo rasgueado (right hand strumming really loud) blew the meters on everything and needed to be brought under control if I wanted to have a decent over-all level and consistent sound across all 30 etudes.
For that, I used the Waves L3 Multi-Maximizer, a multi-band limiter, which I generally only use as a mastering tool. In this case, it was exactly what I needed to reduce the signal on those very occasional really loud parts without affecting everything else.
I found two interesting things after that. The first was that I didn't like what the LA-2A did to the mix! It's uncommon that that happens, but it did. I opted for using the C1 again instead. A barebones compressor with no "color" was, in fact, all this needed for finally blending my close, far, and room mics together. The second thing was that any little tweaks I tried to make with various EQs took away from the sound rather than adding to it, so I left the mix without any further adjustments.
Oddly, EQ was hardly a factor in creating these recordings. Most of it was finding different ways of using compression (which is often avoided in classical recordings) and using different mic "perspectives" to create the illusion of what the human ear interprets.
If there's one thing to keep in mind with the use of most effects, it's that less is more. You'll often get better results from dialing back a particular effect and applying it again at a later stage in the process if necessary than you will by simply dialing it up in the first place.
Take notice of the number of times that compression of some kind is applied, but just a bit. Not more than 1 or 2 dB. There are two different reverbs in use and each is used differently on each source. I used EQ to roll off some of the sub frequencies, but I didn't cut them out when that wasn't enough. I moved to a frequency specific compressor after that and ended up using that idea subtly twice rather than dramatically once.
Here are some effected recordings for you to listen to. Compare these to the dry ones in my previous post. Consider how the three sources relate to one another and help build an amalgam illusion of a real guitar hitting your ears instead of a recorded guitar.
close wet
far wet
room wet
And the end result:
mix
Brant Grieshaber - guitarist
Guitar Teacher
