My friends at Hollywood Scoring wanted to test out bringing their own recording gear into the iconic United Recording Studio A, which has been 'closed' since the beginning of 2023. They invited myself and a few others to write for a small orchestra. I was also on hand all day to conduct or help in the booth for other composers who conducted themselves. I consulted with contractor Noah Gladstone and engineer Adam Michalak on what ensemble would be best for the day. We came up with flute, clarinet, bass clarinet, three horns, and strings (18/6/6/3). This is a great ensemble as you have a decent string count, making a full sound; three woodwinds that can be used for solos, lines, or chords; and three horns that also can play a chord for a pad, unison lines, or a solo. Some people wrote for just this group, others used it to beef up samples or play along with synths. We had every possible combination covered.
I decided to take the opportunity to write something that would demonstrate some of my favorite orchestration techniques. In this two-part post I'm going to go deep into my process, show you the scores, and let you hear the result. Some of what I discuss here is also covered in more detail in other dedicated posts on the blog, so there will be plenty of links to follow. For other parts I'm going to go into more detail for this specific context, and maybe on some tangents. Some new stuff I'm just going to make up on the spot and as always I will talk about the notation just as much as the orchestration as you as their connection is something I feel is not talked about enough in other resources. So grab a glass (or bottle) of wine and some vegemite on toast – we are going to break this apart and plunge deeper than a drop bear’s free fall from a eucalyptus tree. (That was ChatGPT's suggestion, my editor overruled what I originally wanted to say.)
Why Error Code 7 I hear you ask? If you are a finale user you will understand. If not, lucky you.
Before reading on, check out this score video so you know what I am talking about.
- PDF of the score used in the video.
- PDF of the score used at the session.
- Video to download, easier to refer to as you go along. (Click the 3 dots bottom right to download if it does not automatically.)
The piece starts with what we call a 'wire' in the scoring world, a long held note that often in demos will come in from nothing and swell. As you can see I just have them start at pp. They will start at the tip of the bow and you will naturally get a swell as they move the bow. 99% of people would write it from niente or ppp and have a swell, but I think this is redundant as the physics of the bow and the mental game of chicken the musicians play about their soft entrances will lead to that effect with no notation needed. In fact, when I have written an actual swell, they often go too far and then I have to waste time telling them not to. It is important to note that the larger the number of players on it, the more the natural building effect is present.
I have also been on many sessions where the engineer prefers not to have them all attempt to come in from a true niente, this can waste a bar before making a proper sound. Instead, with a clearer start, the engineer can use volume automation to achieve the effect, and with better control, if that is what is needed. What is harder to do is go the other way and make something out of nothing. So as you can see, for several reasons, writing from niente or instruction to 'come from nothing' are best saved for instagram posts, not the real world.
A favorite trick of mine to keep these wires subtle and under control is to have it played by players at the back of the section. These players are further from the Decca Tree so the sound that's picked up is less direct. There will be microphones near the players should we need to bring it out more.
For every project I do I define three or four stands as 'Back Stands' and they can play anything like this. I will make sure that it is an even number of players from both firsts and seconds so that the rest of the parts can still function and balance as usual. In this particular section I had eighteen violins (ten firsts and eight seconds) so I had four from the firsts and two from the seconds peel off, leaving me with six players on each part. To get them back into the section, my short hand is 'div. ord.' I use it to cancel any fancy divisi I might have specified and return to standard firsts and seconds. Sometimes I might use an arrow pointing them back to their old staves as well. I find I do not have to do any fancy orchestrating to get them out or back into the section, if a part is coming up for 'Back Stands' they will just tastefully end whatever they were doing and be ready for their new parts. Once done, they do the opposite.
For the very opening, I instructed the violins to use 'circular bowing'. Instead of the bow staying in one position in relation to the bridge and moving from side to side (whether normal, sul pont, or sul tasto), it circles, covering all of those spots. A section will not sync this up by default, so you get a whispy sound with subtle random color changes. This is the opposite of the technique I use in the other parts of this introduction where they all move at the same time to pont and back. In this piece it is only played by six players so each color change is more pronounced. The more players you have on it, the more the sounds blend together and the less obvious the effect will be, though it is still a cool sound.
The first chord is Fma7, followed by an Ebma7. The wire starts on an A, but then at the chord change I have half of the players slowly side up to a Bb while the rest maintain the A wire as a pedal point. Then the chord changes to a Cma7; the Bb slides to a B, notated as a Cb here so they can see the portamento line. (I know that will drive some purists nuts, notating B as a Cb or visa versa, it works just fine!) The default for players when they see a portamento line is to wait until the last minute before they slide. In this case I want them to move right away, so I had to add the text to adjust their behavior. Unfortunately, this goes against all of their training and I often find that even with the text I have to ask them to not hold the note before they slide.
A common comment I might then get from players when the slide is only a half step or even a whole step but up high is that they have such a small distance to move their finger they need to wait. I will usually say I prefer they move early and arrive early, over waiting. I will also often have to ask them to not slide the last bit rapidly to land it on the beat where it is notated. This gives a comical sound and ruins the effect. I tell them that if they are in that situation, it is ok to be late arriving at the target pitch, as that is better for the effect. In this example I have them all move at the same time, that is their default. Sometimes it is nice to have them move separately, giving a blur to the sound. The easiest way I have found to do that is to ask the inside stands to move later. If you just ask some to do it, no one knows who is going to do what without having a discussion and you might get all or none!
Seating and session organization - Tangent alert!
Over the years I have developed ways of organizing divisi that are simple for everyone to follow, foolproof and adaptable should the section size change between when you wrote your score and the session. I always avoid specifying exact numbers. Not only can things change, but it gets confusing as to who is going to do what. String players always have a reference for who will do what, I have just built on this.
You start with standard firsts and seconds. By default they will divide on the stand, no special instructions needed (I don't even use the words 'div' or 'unis' in this situation). If you need three parts, in most cases I divide the seconds, I will explain why in part two. All strings players share a stand and thus work in pairs. In a simple two part divisi the top voice goes to the player sitting on the outside of the stand, closest to the 'audience' and the bottom voice will go to the player on the inside. Check out Divide and Conquer for more information on divisi in strings. If you want three voices, my first choice is usually to keep the firsts all together and divide the seconds.
As all violins read off the one part in the studio, it is super easy to write three lines and have them split into equal groups. I mark this 'div. a3' and for clarity I most often use three staves. Players also prefer it copied this way too. I do mark this with text as it is not standard or a default behavior. In order for this to work, they need to know who does what, so I draw a nice picture for each session and give it to the stage beforehand to set up the seats. Then it is very easy for me to show to the concert master and then we can explain it to the players before we start the first cue.
If they are in three sections, and then extra notes appear, the default is to just divide on the stand, as they would in the more usual Violin I and Violin II sections.
This is why I always prefer 'div. a3' in 'groups'. Some leaders will have them 'number off': 1,2,3,1,2,3, etc. This can get confusing and is limiting and I am convinced every time it is done there are at least two players who have no idea what part to play. By doing it in groups and keeping each stand/pair on the same 'part' you can then divide them further. Thus I can easily get them into six parts with minimal text and discussion. This technique also keeps each 'line' in blocks than can easily be pulled out by spot microphones. If you number off, each part is spread out. I don't mind that sonically, but it is harder to then dig a part out if needed when mixing.
With this system, the same score will work with eighteen or twenty-four violins. Speaking of which, I have always said my perfect number of violins is twenty-four. But you can't always afford that. In my experience, if you want to keep a full sound and be able to divide into six parts and not get too weak and thin, eighteen is the minimum magic number. This gives you ten on the top which gives a nice full sound and you can divide into three even groups of six, or keep six firsts and six seconds and have six peel off at the back and maintain balance. As you will hear in this piece it works well and does not sound thin. Likewise, I think six viola, six cello, and three basses balance eighteen violins. It's true that I like more celli than violas in my large sections, but that is partly because you are usually up against more low end brass in a large orchestra. In the case of this session and music, there was no low brass. Check out my post on section sizes and layouts for more information.
In this example from my orchestration of Chris Beck's arrangement of the Disney logo I have five parts. To be clearer I label each staff with the group name, but if it was just one note in each I would just say div a3 as I do in Error Code 7. I dare anyone to come up with a better way to get a five part divisi! Note that I could have also had the third stave played by the back stands, and then the front firsts and seconds would divide as they do in Error Code 7, but I did not want the Eb to be that distant.
You can see how I switch to standard firsts and seconds on beat three of bar 12, the firsts are divided and the seconds are unison. Click here for further discussion of divisi, including this example.
Returning to Error Code 7...
The initial harmony is a simple Fma7 chord and I use orchestration techniques to make it interesting. Usually that chord has a lush, sweet character, but I decided to orchestrate myself away from anything lush or cute. I have the notes arpeggiated in the strings and woods and held in the horns and some other strings.
I borrowed a gesture I used in A Spell, A Prayer for Corrine Bailey Rae and the Metropole Orkest.
You can check out the full performance from the North Sea Jazz Festival here.
It starts with a simple arpeggiated line, that I then split into four parts, where each part just repeats a single note. This gives a nice blur as each note sustains. It is a trick I use all the time, though not always with every note of the group, and it works really well across all sections of the orchestra. As you will see later in the chart you can get a cool effect just by splitting into two parts and having one or two notes hold in places.
I first worked out this technique when I was coming up with creative ways to split woodwinds on fast passages, usually pasted from the strings. I think I will call it E.A.D., Extreme Australian Dovetailing. You can read more about that in It's Staggering.
As you can see, It is a tricky one to notate as you need to give breaks so they can articulate the tremolo. You could just make it continuous and have accents, but each note would be late and lack definition. I made it cleaner by only using the trem articulations on the first time, then I used simile. I will say though that there is no perfect way to notate this. I tested several ways including cleaner looking full durations with commas for the lifts before each articulation. In general I love commas; rather than specifying an exact duration, you let the players use their musical instinct, which results in something that always matches the context. But in this case looking at all the commas after every note looked ridiculous. Like notating modal music and enharmonics, sometimes there is no perfect way that will please everyone.
To add interest, I have them move from sul pont to sul tasto and back with each swell. You can notate it with funky lines, but I just did it with some text as I already have hairpins giving the shape of what I want. Either is fine, just don't do both, you know how much that would annoy me. I added the word 'extreme' to the sul pont as I have found that in the studios players never go as close to the bridge as I would like, often for fear of sliding right over it. In both examples I get extra mileage out of the idea by having it repeat up the octave. In "A Spell, A Prayer" you can also hear a simple but very effective technique I used in the clarinets by having them all play ascending scales at different times, creating a percolating murmur.
In Error Code 7 it is a different kind of murmur in the woods, using the same E.A.D. technique I used with the strings. I put them in triplets this time to create a blur against the sixteenths of the strings. The default for any flute is to play with vibrato. Clarinets do not play with vibrato in legit music, only in jazz, so in order to have the flute blend with them for this texture, I marked it senza vibrato.
I compliment the sound of the sul pont by using stopped horns. As they swell, they sound more metallic, exactly the same effect as going from sul tasto to sul pont. Gluing all of this together I also have some strings holding the chord also moving to pont and back. Notice in the picture the horns are using stopping mutes instead of their hands. The mutes create a more stable and even sound than the hand. The choice of stopping with the hand or the mute is always left to the players. In general if there is a single note or a few in the middle of non stopped ones they will use the hand. If it is a whole phrase, they often use the mute. The mute is on a string that they can put around their wrist to make it easy to get out in a hurry. You can see the string in the photo. I notate the stop with a + above the note, nothing else is needed and it is just magic that if there is no + above a note they won't stop it! I know the convention is to place a circle above the next open note, but I have never done that and not once has a player kept playing a stop or asked. And don't get me started on adding text as well.
In order to pull off all of these string parts I had to really think about the divisi as I was already down six players to do the wire. I have the violas and the cellos divided into two (they will divide on the stand by default). Having six people on the wire left me with six firsts and six seconds. If I need four parts then each of them just divides on the stand. In order to not use five staves to pull this off I got creative and labeled the top staff as outside firsts and outside seconds, then used a staff each for the inside firsts and the inside seconds (all violins play off the one part in scoring). While it is still a little complicated, I think it is the best solution to get the parts I need with minimal text, explanations, and trees killed.
I repeat the same figure in Eb then C, totally cheating by just pasting it and transposing. Notice how I made the notation simpler and less complicated on the repeats. No need for all of the obvious information to be repeated. It also makes it simpler should we decide to change the dynamics; it is just the first figure that would need tweaking, then the rest just follow.
The next part of the piece helps transition us from the textural introduction to the harmonic main theme. You will see the back stands finish their note off while a new idea starts. Note that we are now in mutes. In the concert world you would have to have them all put them on in a rest and risk a lot of noise, or have them stagger them on, leaving it up to them as to when to do it. If you have read anything I have ever written you would know that is not going to be a solution I would choose. Being in the studio, we can cheat and stop recording, put the mutes on, and then start up again with no risk of unwanted noise or players randomly dropping out. I don't care how many leaders tell me they can do it, it only takes one person's mute to make a noise and the take is ruined. Same thing with page turns, I always stop and start. No one will know once it is all edited and mixed.
You will see I marked bar 19 as 'Pickup' and put a note on the back stands to finish their part on the previous pass. I like to plan how we are going to record and edit as I orchestrate, so I will mark the pickups in the score. We will often break it up even more once recording, but if I know ahead of time we will stop and start it is nice to have it in the score and the parts. This is not standard and I dream of the day when everyone knows what it means. In this case it is for mutes, but sometimes it is because a woodwind or brass player has to change instruments and there is no time, so stopping is the solution. There are many other reasons why we do pickups and I talk about them in Part 2 of my series on conducting.
This section employs two of my favorite string techniques, "stremolo" (my term) and portamento or sliding. You first hear the stremolo in the second violins in bar 19. If I want some interesting color on a wire but don't want it to stick out as much as with circular bowing, then stremolo is a good way to go. Players randomly speed up and slow down their bow, as if they are going to get to a full tremolo, but they never get that fast. One can ask them to go to a full trem and back, it’s also a cool sound, but that effect is more in your face and I prefer the more subtle strem effect. If you listen to the Back Stands 18-20, then 25-30 you can hear the difference. With Circular Bowing you hear the upper partials pop out as the bows get closer to the bridge. With strem you just get a shimmering quality to the sound, but it does not get in the way or draw as much attention as circular bowing. The effect pops out more if there is a swell, such as with the violins and viola in bars 27 and 29. The example to the right is the technique, but to appreciate the effect you need a section to even out the changes.
For some history on the name: I had already started using the 'S' on the note as it was not used for anything else and was clean and readable. I was recording in London and asked the orchestra what I should called it. Sonia Slany, a fellow Aussie, yelled out 'Stremolo' and it stuck.
The other technique I love is long slides in the low end. You can hear a ton of it in my Trollhunters and Maya scores (examples below). I already discussed the issues around timing and slides in my section about the introduction but another important consideration if you want a smooth and good sound is whether the notes can be played on a single string. The slides in the cello in bar 19 are fine, they can only play the F on the C string and they can slide up to the A. Cellos don't mind going into upper positions and the sound is fine. The bass is not quite as flexible, as the sound and intonation in upper positions is not as stable. However these notes are fine, the F starts on the E string and they slide to the A. One thing to note is that while every other section avoids open strings, bass players are in general a lazy bunch and will always choose the open string (this is acceptable due to the fact that basses have a softer tone than the other stringed instruments, so open strings don't poke out like they do on those others). Here that means in bar 23 they need to look ahead and play the D on the A string, and not the open D as they would by default.
Here are a few more examples of bends in my work as a composer on Trollhunters and Maya and the Three. The Maya cue takes it to the extreme; I also use some samples to beef it up, as is pretty normal these days. The scene is a battle with an Ouroboros, so extra sliding around was called for (note that the long fall at :36 is samples and was never live, nor intended to be)
At bar 27 I bring back the woods and horns with a tip of the hat to our chord shapes from the opening. This is also a bit of a nod to a technique Chris Beck uses all the time where chords come and go in different families, overlapping. In this case it is strings, then woods, then horns, then back to strings. I use mutes in the horns to take the edge off and help them blend, the exact opposite of how I use the stopped notes in the introduction. (You can find more information about muting and stopping horns in my post about the instrument, including a video of studio legend Jim Thatcher demonstrating.)
Looking at it now I wonder if I meant to have a C in the violas in bar 30 and forgot, or maybe I just liked the sound of the minor second in the violins, we will never know as I have manipulated so many more crotchets since writing this and can't remember. I know theorists for generations to come will come up with ideas about what is going on.