When I started working on large orchestral scores, the string section sizes were pretty standard, and similar to the concert hall where each lower section shrinks by two players. So for example, a standard orchestra might be 14 first violins, 12 seconds, 10 violas, 8 cello and 6 basses, or “14/12/10/8/6” for short (in the film world where we lump all the violins together, so this is often condensed further to “26/10/8/6”). After many years of orchestrating, conducting, and recording orchestras I have come to the conclusion that those numbers are not the best for the studio. My standard lineup now is 24/8/10/4. This gets me through most scores. That is 46 players. Of course on some projects budget considerations mean I can’t have everyone, and during the Covid period caps on numbers meant we could only have a smaller ensemble. Either way, there will ideally be more cellos than violas, in contrast to the standard symphonic ensemble.
Violins
My preference is for 24. That is enough to make a full sound. You have an even number of firsts and seconds (I do not see why there should be two less seconds). One of the advantages of writing all the violins on the one part in the studio is that we can easily divide into three even parts. If you try this with separate Violin I and Violin II parts, it gets complicated as you need to write lines one and two in the Violin I and say ‘divisi 8/4,’ then lines two and three in Violin II with ‘divisi 4/8’! Instead, with combined parts, we just write the three lines and say ‘div. a3’ or ‘8/8/8’ and we are ready to go. Also if we need to rebalance the parts, perhaps with more people on the melody, we can easily move some seconds up.
24 also easily goes into four parts of 6/6/6/6 and so on. It is really easy to work it all out. I like the violins all set up in as tight of a formation as possible. There is no need to have a super defined first and second section as a lot of the time we are not divided like that anyway. I will always show my plans to the engineer so they can make sure they have all of the section mics in the right places to catch each group.
I find that in the studio, with microphones everywhere, this is enough to balance. If the brass are really blasting, we will need to stripe it anyway. I have noticed that with more players, the back chairs end up so far away that I do not feel connected to them, and can often see from how they play and act, they are not feeling connected to me and the rest of us. Timing on tight rhythms can get sloppy due to the distance and sheer number of players ‘widening’ each attack/articulation.
I recently posted this layout diagram for the updated 100th Anniversary Disney Logo session I orchestrated and conducted. As previously mentioned, I prefer to use 24 violins. You really don’t need any more in the studio. But sometimes, as in this case, I am on a gig and there are already 26 or more booked before I am involved. I am not going to cancel some players just to make my maths work out easier, so where are the extra players going to be of more use? Is it on top where the line will already project? Or in the 2nds where the support could help and you are most likely to divide? Honestly, I have never understood why a standard orchestra has two more first violins than seconds (once you get up into symphonic numbers); the instruments are the same and they produce the same amount of sound. A viola is bigger and produces more sound mass, so you need fewer to balance, and so on for cello and bass.
This thinking applies when you have plenty of violins anyway; like I have said before, 24 is my perfect number. Under that, you can benefit from having extra bodies on top to give the impression of a bigger section.
I way prefer having all the violins seated together on my left. It seems like every time I go and see an orchestra in the concert hall they are set up in some different way. The most common alternative, having the firsts on the left and the seconds on the right, is not a favorite of mine nor of any players I have spoken to. It makes it hard for them to communicate because they can’t see, feel, or hear each other well. The back desks of each section are so far apart they need binoculars to see each other. Of course there are some cool effects you can get with the section split. Chris Beck likes chords to swell in and out from side to side sometimes. We will have a discussion before each project and work out if that is needed. It is important that I know the seating before I start orchestrating. That said, if it is just a small section of the score that needs that, you can always stripe it on one side and then flip it in the computer!
Violas
I like eight. Normal numbers would dictate ten, but I have found eight works great. Dividing into two parts is fine. I like them positioned in the middle as they are the glue that holds the string section together.
Cello
Budget permitting I will have extra cellos, ten or twelve. There are several reasons for this.
Modern scores have a lot of low material, be it chords or riffs in the bass clef range. While the viola can play a lot of those notes, they are way down on their C string which is quite loose and does not resonate on long notes or project on short ones like the cello can. When I had eight cellos, I used to have two play with the violas on their low riffs to give it some bite. Now I go the other way, I have six cellos play that, then maybe a few violas can play along to thicken it. If I have ten I divide it 6/4, dividing in groups as opposed to on the stand. I will have the six at the front play the upper part and the four at the back play the lower, as they are usually with the basses. Not only does this work for riffs, but if I ever need the cellos to go high on a melody, I can have a few left to stay low on the root or fifth.
So since it always seemed that the cellos ended up in two parts a lot of the time anyway, adding a few players helped the sound of the section. Plus a single cello can project and produce more sound than a single viola, so you get a little more band for your buck. And even if they are in unison on the bass part or a melody, it never hurts to have more! So boosting the size of this section just makes more economic and musical sense.
But don’t get me wrong, I am not saying the violas do not have their place. Cellos in their upper range will not blend into the string section on chords, and moving parts in that register feel much more flowing on the viola. I have done sessions without violas, and it really does feel like there is a hole in the middle of the sound.
I like cellos on the right, split up into three rows of stands (six players) on the outside, and two rows (four players) on the inside.
Basses
Four or six are great. Any more and it just gets too thick down there in most studios. I like them on the right, behind the cellos. If it is a string-only session, some people have them in the middle. I still prefer them on the right as they work with the cellos so being close to them makes sense. I have also heard from engineers that the sound is better on the right as they are then usually close to a wall and this helps the low end.
A Scoring Example
- 12 Violin I
- 12 Violin II
- 4 Violas – this line is just gluing it all together, and will be fine with four players.
- 4 Violas – this is the most important part. In this range the viola will not project well, but it will thicken the sound.
- 6 Cello – This will pop, it is in a great range for the cello to project. Just think about how much longer and tighter the cello string is for these notes compared to the same notes on the viola.
- 4 Cello – Will have lots of bite in this range.
- 2 basses – Adds weight to the cello, projects as well as it can for a bass
- 2 basses – will be more felt than heard. The string is so long and loose that it does not bite at all like the cello, but it works well with the basses and cello in the octave above.
Some Seating Examples
These diagrams were drawn for projects done during the pandemic, so the groups were spread out more than they would usually be, but this still gives you an idea of each section’s layout. For more in-depth discussion of dividing sections and an explanation of the color-coding going on here, check out my post Divide and Conquer.
I Like It Close
I like to set up the strings as close as possible to each other, as tight as I can get them before they complain! This does several things. First, they can feel, breathe, tune, and communicate with each other much better. Second, the sound is more concentrated when it hits the Decca tree (the main microphones above the conductor’s head). You would think you will get a bigger sound by spreading out the players, covering a larger footprint. But in my experience the perceived sound doesn’t increase and it makes it harder to play as a unit. This was very evident to me with our Covid setups, where we had to set up on single stands spaced apart. The sound is more diffuse and wispy, and timing, tuning, and communication are all more difficult.
Bigger Is Not Better
One would think having a huge string section, more than the numbers above, would be advantageous. But in my experience excessive size creates some problems. Firstly, the physical space needed means people are too far apart to know what is going on with everyone else. Secondly, we rely on some amount of humanization of rhythms to give a natural feel, but too much humanization can make things worse. A large section is just too smooth sometimes. Short notes can also get too wide and ‘heavy’. A small section can sound quite gritty, but a perfectly sized section smooths things out in a beautiful way.
You can also overload a room with too much sound. The low end in particular can get out of control if you are not careful. The last thing you want, after you have told the composer and studio how awesome eight or ten basses will sound, is to have the engineer ask to have only four or six play because they are overloading the room. I know this from experience!
Small Sections
Orchestras for TV shows and indie projects are usually smaller, for example on Empire we had 13/5/4/3. I still like them set up as tight as possible.
One interesting phenomenon I have noticed is how the size of a string section affects its performance and the recording. In small sections, the players tend to play softer, and the feedback from the booth is always to play softer as well. There are also way more noise issues when recording smaller sections. I have several ideas about why this may be the case.
Firstly, the larger the section, the more the sound will naturally balance and even out. It is harder for individual players to ‘stick out’ in large sections and so there is safety in numbers. But in smaller ones, it is easy to stick out, so players hold back a little to avoid this. Also the psychological effect of being in a smaller group where it is harder to hide can cause players to play more conservatively.
There is also the issue of responding to what the booth hears. The sound of a smaller section in the booth is generally harsher as you hear more ‘grit’ in the sound; it is just not as round. Think of how gritty and big just a quartet can sound, and it is because you hear every hair on the string. As mentioned, the more players, the more even the sound gets. A smaller section will have a harder time achieving the mellower, rounder sound of a large section, and so the primary way they take the edge off is to play softer. One solution to address this situation is to play with mutes, as then they can play out a little more without it getting louder, thus taking the edge off the sound and making it easier to blend.
As for the noises, with fewer players, playing softer, there is less sound mass hitting the mics, so the inputs get turned up more to compensate. This then causes all the other noises in the room to be more audible.
Conclusion
That’s my take on strings in the studio! Check out the rest of the blog for more details related to string writing. I already mentioned my post Divide and Conquer. I also have a two-part post about a piece I wrote and recorded to demonstrate many of my favorite orchestration techniques, which includes a discussion of the section size and divisions as well as audio demonstrations.