3 Ways Of Treating Room Mics – The Making Of “Merrily We Roll Along”

So I finished the “Merrily We Roll Along” record a few days ago. I really like recording Broadway cast albums, especially Stephen Sondheim’s stuff. They’re always big, involved, and each one presents its own set of challenges. This production had 20+ people in the cast, a fairly large orchestra, and a huge amount of music to be recorded. Yes, this was another double album. There’s one thing I wanted to talk about that was a bit unusual this time around. The instrumentation on this album was somewhat different than your usual cast album, and I got to experiment with a few different mixing techniques necessitated by this unusual arrangement.

This show is very brass heavy. We had a bunch of trumpets, bones, french horns, and a tuba. They were busy and they were loud. This in itself isn’t unusual.  A big brass section is a fairly common occurrence on Broadway, and it’s usually balanced out by a big lush string section. The two sections orchestrated well make for a very impressive sound. Jonathan Tunick, who arranged this show, decided to take the string section in the other direction this time. He scaled down not only in the writing style, but also the instrumentation: He got rid of the violins. We just had a few violas and celli, and their playing was understated.

With the whole band in the same room, and heavy isolation not an option, how do you make a string section like this shine without making the brass take over the mix? I could just really work the string section’s close mics, and refrain from featuring the room mics, but that makes for a boring sound. The richness of a string section really comes through the room mics a lot more than the close mics. I wanted to be able to have the room mics nice and loud, especially since Avatar’s room sound is amazing, and I wanted to make the room sounded nice and bright, since there was no violins to contribute to the brightness. But I needed to do all this while making sure that the big brass sound doesn’t get too wild. I tried three different mix techniques to accomplish this. Some were more successful than others:

First, I tried using a dynamic EQ on the room mics. This is an effect which can apply different amounts of boost or cut to an EQ band depending on how loud that frequency is. It’s similar to a multiband compressor, but the heart of it is an equalizer, not a compressor with some crossovers. Brainworx and Sonalksis are two companies that make this type of plugin. I put one of those on the room mics, and set the effect to have a big high frequency boost when the incoming signal was low, and no boost when the incoming signal was hot. This worked fairly well, but I couldn’t find setting that would globally work most of the time. This technique also didn’t have a sound that was drastic enough for me. I actually ended up using this effect on the close string mics to control the high end a bit, but I moved on to trying a different trick on the room mic.

My next experiment was keyed compression. First, I applied a very hefty high-frequency EQ boost to the room mics. Then I put a compressor on the room track, set it to be keyed by a bus input instead of the regular signal, and fed the direct brass mics to the key bus with a send.  This way, every time the brass would play loudly (especially the trumpets), the compressor would kick in. This is a great technique. It’s not a multi-band type of sound, but it’s very effective. I could really dial up just the right threshold and ratio to control the brass sound when they play out while preserving the volume of the bright room mics when the brass section is quiet. I used this technique on the Follies album, and was happy with the sound, but I did find that when applied drastically, it can result in a “pumping” sound. It’s not the most transparent technique in the world, regardless of the attack/release settings. I figured I’d try one more thing.

The most successful technique was also the simplest one. Keeping the EQ boost that I used in the previous keying technique, I applied a fair amount of compression, but instead of keying the compressor with the brass mics, I simply used a sidechain filter to filter out all but the highest frequencies. This way, the compressor ignored how loud the low and mid friequencies were, and only did its thing when the high frequencies got unruly. Since the loudest high frequencies were produced exclusively by the brass, the effect was very similar to the keying technique. But, for some reason (which I have yet to figure out), this technique sounded a lot more transparent to me. I could apply massive amounts of compression, and while the room mic sounded ridiculously compressed when soloed, I didn’t hear much of that pumping sound when playing the full mix. I ended up using this technique throughout the album. Also, whenever I wanted the brass to get really big, instead of boosting the brass mics or the room mic fader, I simply backed off the threshold on the compressor. This sounded pretty cool to me.

So there you have it.  Three different techniques that could accomplish the same task, some more successfully than others. I ended up getting exactly what I wanted: A small string section that didn’t get swallowed up by the big brass section. Check out the PS Classics website to hear clips or order the album.