That was great, now lets make it better
There is nothing like working with the greatest sight-readers on the planet. But even if they get every note right on the first run, the magic is not going to be there yet. My theory is that it takes at least three runs before the real magic happens, when the orchestra gets the correct tuning, timing and balance. For my example let’s assume this is an easy-to-medium difficulty cue. Harder cues would take more time, but this theory holds the same for the simplest of cues, even ones with footballs.
First run– No one has seen this music before; they do not know if they are playing the melody, harmony in thirds, or a counter line; if it is all consonant or if there are some rubs. There may be some questions about notes or dynamics to sort out. One awesome thing the players here do is sort most of that stuff out themselves, only asking or letting us know if there are real problems.
Second run– The player knows their part is right. They play it with no mistakes and now they are listening to the players around them more, noticing who they are with and how to balance the music. BUT, they are doing this as they go, so they have to play the note before adjusting, making mental notes for the next run.
Third run– The players see the whole picture now, who they are with, whether they are in the foreground or background, etc. It is not until this third run that the tuning, timing and balance really come together.
Of course there are many other variables that go into this and we often do way more than three takes especially for longer or harder cues. One thing to note, though, is that you eventually get to a point of diminishing returns. I have been on sessions where the director or composer for some reason wants another take, even after we have nailed it a few times. This can wear out the players. They are human beings, after all, and start to wonder why they have to keep doing the same thing over and over. As their mind wanders so can the playing. This is where the conductor can really help. If it gets to that point, I always ask the people calling the shots what we can do for them to make it better, and if they have no answer, sometimes I will suggest they just listen to the last take again. If they are adamant that we do it again, I will let the orchestra rest, then come up with something small that will breathe a bit of life into the cue but not change it too much, thus giving the musicians a reason to play it again. It could be something as simple as bringing a certain part out or asking that a phrase end a beat earlier. But if they have a reason to play, then they will play much better.
Depending on the budget, we may not be able to run every cue three times. With great players it will be OK, but I still feel that to get magic the above theory holds.
I find I spend a lot of time making sure the orchestra locks with the tracks. Even though I am dealing with session musicians, they are classically trained and tend to play rhythms in a classical way. That means, for example, they will ‘clip’ up sixteenth notes.
Without any intervention, it is highly likely that the orchestra could sound like this when playing the above rhythm, it is a little exaggerated, but you will see my point.
It is natural to bunch up groups of notes. It sounds fine when the orchestra is by itself, but when we try and line this up with a drum track that plays all the 16ths in the bar, it is not going to work.
In Example 3., they will bunch up, but will also get to beat 3 and 4 early. I will have to ask them to think about beats 3 and 4.
In Example 4., they will wait a little longer than an 8th on beat 4 and play the 16ths faster to still land on beat 1.
I don’t worry about intonation on the first read through. We are all sight reading, so no one knows what part they have or who they are with. Players will know when they are out of tune and usually do not need to be told. But if it is not working on the second run it is time for the conductor to step in. I am pretty direct when it comes to saying how it is, but I have found over the years that if you can make your point without putting anyone down, you get a better result. If the horns are not tuning with the celli, instead of telling them that right away, I ask them if they can hear the celli and have the monitor mixer put some cello in their headphones. ‘Let’s take an A’ is a universal expression for politely saying something was not in tune. The next trick if a passage is not tuning is to rehearse it without the click so the players can really concentrate on the pitch. You usually do not have to say anything, and they will work it out. If the problem is a timing one, I ask them if they have enough click.
Striping is when we record the orchestra in sections. There are two reasons for this; in movies it is for balance and mix control, and in games it can also be for layering. It does not matter how large the string section is, brass at fortissimo can drown them out. Sometimes we ask the brass to play a little softer, but if we want the sound of them at ff, we will stripe them.
Some scores, notably those from Hans Zimmer and other Remote Control composers, record different sections of the same cues on different sessions, usually strings during the day and brass in the evening. This is quite hard and can lead to some less musical results as the brass have no real idea what they are playing to. With no strings in the room, intonation can also be a problem. The conductor must really be alert and thinking ahead when recording like this. When I work like this, I have to think of what the brass will do and make sure the strings phrase in a way that works with the brass. If there is a big theme played by the celli and the horns, I have to make sure the celli know this and ‘breathe’ with the horns. If there are any changes made, good notes must be taken in order to relay them at the later session. Pacing then becomes an issue for the brass. On a normal session they have the opportunity to rest, but on a brass only one it is easy to push them too hard and blow their chops, or get them worried that they will. The strings may also play with a little less intensity when the brass are not blasting at them so I soften have to let them know what the brass are doing.
An ideal way to stripe is to have everyone in the room together, but do separate passes for each section. If there are woodwinds and keyboards they will play with the strings. Percussion might go with the brass or on their own. This way everyone knows what everyone else is doing and the tuning is not as hard to get together. A recent example of this was Chris Beck’s score for Ant-man and the Wasp. We had a large orchestra at Abbey Road. Many cues we recorded all together, but if it was a with loud brass and percussion we would rehearse them all together, then record in groups.
Read the conclusion in Part 4.