I’m in Buffalo with my chamber group, the Meridian Arts Ensemble. We have a concert here and we’re making a recording. Actually, we’re making half of two recordings. The process is interesting.
The first CD of which we are recording half is a typical (for us) album of contemporary music. Of particular importance and pleasure to me is that we are recording a piece I wrote. That, however, will not happen until the spring. One of the other pieces is by a composer named David Sanford, and it is for our group plus a solo trumpeter. He wrote it for us to play with Dave Ballou, who is an old friend of the group, and Dave is here recording the solo part with us. We also have here some friends from Mexico, and we will be recording some brass choir music (Gabrieli and the like) for half of another CD. Fitting this project together was like doing a jigsaw puzzle, but everybody is in the right place at the right time.
To make a recording (and I am speaking here of classical music, broadly defined – pop music has a completely different process), you need: players, an engineer, and a producer. The engineer places the microphones and runs the equipment. Microphone placement is an art form: a centimeter can make a huge difference in the sound, and remember that we are in a 3D environment. In other words, the mic can go up or down, sideways, and can point in different directions.
The producer sits with the engineer and follows along with the score (the music, printed with all the parts one over the other) and makes sure that we have recorded what we need to record. Classical music CD’s are heavily edited (more on that later). If I miss a note in the third measure of the piece, we have to do another take to cover that note, and later an edit will be made. The producer has a huge responsibility, since the music is flying by and he or she needs to make sure everything is together, without missed notes. Further, he has to make sure that the character of what we are playing is right. Are the dynamics (loud and soft) coming through? Does that section that is light and bouncy sound light and bouncy? Is the trombone lining up with the tuba? Are all the parts balanced? Is the energy right?
For these sessions, we spent maybe an hour getting the sound right. It involves playing while the engineer and producer make adjustments according to their judgment. Some of the adjustments are done at the mixing board (which is actually a computer screen that looks like a mixing board) and some involving moving microphones. Then we record a little and go to the booth to listen. If we ask for more adjusting (as we did this time), then we have to record a little more to hear if we got it right. Need I mention that as soon as the microphones are correctly placed, everyone must walk around his seat with the greatest of care, not to disturb anything?
Once the sound is right, it is time to record. As with every other part of the production, this needs to be planned in advance. In what order should we record the pieces? Are there changes of setup, such as different percussion instruments for different pieces? Are there issues of fatigue or endurance? (Answer: yes, always.) Are there extra players who shouldn’t have to wait around for their piece? There is a complex choreography to the events of several days in order to complete the project on time.
We begin to record. We play either an entire movement or a section of a movement. In this case, we started with Dave Ballou’s piece, entitled For Brass Quintet and Percussion (Dave being the same trumpeter who will record another piece with us later, but he has also written us a piece). The Ballou is in one continuous movement, with no changes of percussion. All good. We play a hunk of music. Not the whole thing – every piece is conquered by the mouthful, not the meal. We need lips and brains and sticks firing at 100%, so it doesn’t make sense to record five minutes of music, losing our concentration after the first two minutes. We record the first section three or four times. Each take is preceded by a slate – the engineer announces “Take 1” over a little talk-back box we have on stage (of course he can hear us through our microphones, but we need to hear him, too, and he is two floors away from us). After the second or third take, we ask the producer what we need. In this case, the producer is the composer himself. He’ll tell us that in measure 9, the trumpets aren’t together, and that measure 27 must have a bigger crescendo (I’m making up the details). We’ll do it again.
Each take is recorded onto a hard disk using a program called Pro Tools. We need to be organized about the takes, because later they will be spliced into a final version. Pro Tools is actually an amazing piece of programming. Once the takes are stored on a drive, they are never changed, cut, or altered in any way. The software gives the user the ability to create a map that works something like this: “Start with Take 3 until beat 1 of measure 4. Then crossfade into take 7 for 3 beats. Then cut back to Take 3, and fade into Take 5 on beat 2 of measure 6.” Splices used to be made with a razor blade to physically cut a length of magnetic tape, but now it all happens on the computer. It takes a great deal of skill to do it seamlessly.
Once we have the first section recorded (“Are we covered?” we are continually asking Dave, or any producer we work with), we move onto the next. It is grueling work. I make sure to relax between takes, because I want full concentration when the tape is rolling (metaphor, since there is no tape). On the other hand, I don’t want to relax too much, because it is hard to get up, relax, get up, relax … Sometimes someone talks more than I would like between takes. Sometimes I talk more than someone likes between takes. We are constantly trying to get the job done, compromising what we each need to do our best, and there are 6 players, an engineer, and a producer, each with needs.
I once did a recording of a VERY complicated piece. We would do a few takes and ask the producer the eternal question: “Are we covered?” He would reply, “I think so.” OK, better do another take. We can’t move on until he KNOWS, not thinks.
We record the Ballou in about 3 hours of work time. The piece is about 10 minutes long. It will all be stitched up later. Here is our future: Jon, one of our trumpeters, will make a preliminary map, from take to take, of how he thinks the flow of the editing will go. He will then work with the engineer to actually make the edits. There is always a lot of improvisation here; the flow from Take 3 to Take 7 might just not work (maybe someone was a bit out of tune at the end of Take 3, and you can hear the edit point when Take 7 kicks in). Sometimes a take that seemed good actually has a wrong note, or a moment of poor ensemble (in other words, we aren’t together). Maybe the volume doesn’t match up from one take to another. Maybe we have unknowingly changed the tempo a bit. A great deal of art and skill goes into the editing process.
At any rate, a new day dawns, and we are ready to record our next piece, by composer David Sanford. This is the one in which Dave Ballou plays the solo trumpet part, so he is no longer downstairs producing – now he’s up on stage with us, producing trumpet notes. Horn player Adam Unsworth, who is with us this week playing brass choir music, has agreed to produce the Sanford, which is a daunting task because the music is so complicated. First, of course, we need to redo the placement of the microphones. Now we have a new player and a new setup, so it’s back to square 1, or maybe square 3, since we have the basic sound figured out already.
There’s a new complication in this piece: much of the solo part is improvised. This means that we lose the ability to edit. You can’t make a splice in the middle of a solo that is different every time. Also, David Sanford has given us some last-minute changes that are fairly extensive, so we are learning on the job. We’ve had one rehearsal on the changed parts. Luckily, we learn pretty quickly, but we are working very hard. So here we are, 3:30PM on day 2 of the recording, done for the day, and we have recorded 4 out of the 5 movements of the Sanford. We had hoped to be done by now. Just one movement to go, though the movement is part of the new version: it is brand new to us. I can’t remember how it goes. I guess I’ll know by tomorrow at 10:30, when it will be “in the can” – all recorded, all covered.
In a few months, we will listen to the first round of edits, and email our complaints, suggestions, fusses, whatever you want to call them. We’ll go through another round of edits and clean-ups, and then we’ll be ready to master and print the CD. We have our own label, called 8bells, and we will release the album and be very proud of it. Then, onto the next project.