Recently my friend Clare Kavanagh came to me asking if I could put together an arrangement of Puccini’s “O Mio Babbino Caro” for L’Oreal Ireland’s Colour Trophy Awards 50th anniversary show. They had very specific timing requirements so I would need to work to those constraints, extending the aria itself by adding extra sections, a new intro, and ending.
I started by mocking up the track in Logic Pro, using some string libraries to write parts that matched ‘placeholder’ mp3s I had for the aria itself.
Much of the track was mocked up using a combination of LA Scoring Strings and the Spitfire solo string libraries, with Maria Mason playing the lead violin part. I recorded Clare’s soprano part in the morning, Maria arriving later that evening to lay down the violin part.
Recording an operatic vocal part is a very different job to that of recording a standard rock or pop vocal. The dynamic range present in a classically trained performer’s vocal is much wider than that of your average singer, and operating pieces widen this range even further.
When we think of an operatic vocal, the sound we are imagining is formed very much by the response of the room or hall to the singer “exciting” that space. For this reason, you want to mic up the performer from a healthy distance to allow the vocal to mingle with the room sound. You’ll want to be in a nice room too of course.
I used a Blue Bluebird pointed straight at Clare for the direct sound, and two AT-4033s in XY to capture the room, with the mics setup around 3.5m from Clare. All three microphones were patched into our Universal Audio Apollo Quad. The UAD plugin signal chain was the same for all three for tracking – Neve 88RS channel strip, into the Studer A800. I like to emulate traditional large console/analog tape signal chains for tracking, working within those limitations.
I used the Neve sparingly, doing some gentle high pass filtering and adding a smidge of ‘air’ with the High shelving EQ. The 88RS tries to be as linear its phase shift with EQ boosts/cuts as possible, making it very transparent. The Studer I set up with the Quantegy GP9 tape at 30ips to minimise high frequency loss. I went for the default calibration for this tape type, as I was looking for an honest representation of what was coming in, with the gentle compression that tape offers. No saturation wanted here!
Getting the correct gain levels can be tricky, you’ll need lots of headroom and very quiet microphones and preamps. On the plus side though, the standard of performance is probably going to be very high (at least it was in Clare’s case!) so you won’t need to do many takes.
With care taken to position the anger and the mics just right, jobs like this can be really rewarding. You can check out Clare’s page here.