Production Expert

View Original

How To Compose Music In Pro Tools

Composing in Pro Tools, what is it like? In this article, composer Walter Everton, shares some of his tips and tricks he has found out the hard way. It includes a detailed look at his own template, and why he thinks you should build your own template.

There is a lot of information out there on MIDI composition and orchestration on topics from the technical minutia of writing CC automation more effectively to broader scopes on general techniques and approaches to scoring in a DAW.

What is lacking though is how to do this more effectively in Pro Tools. The reason for this is because most composers who have graciously shared their time in creating content work in either Steinberg’s Cubase or Apple’s Logic Pro, both of which I will admit seem to be far more ergonomic than Pro Tools, is at this specific task of MIDI/DAW composition/orchestration.

I have used Cubase in the past, and this is certainly one of the biggest things I miss from my days using Cubase. The workflow of MIDI inside Cubase is much simpler to use.

Here though, I’d like to share some of the techniques and tips I have “discovered” when working in Pro Tools for MIDI orchestration, and some big benefits of working in Pro Tools that you may not be able to have in other DAWs.

Create A Template

When starting a composition project, I start with my template. This is of course not unique to Pro Tools, but I find it integral to creating quickly and efficiently in Pro Tools (or any DAW for that matter). This template frees me from needing to worry about a lot of both tedious and complicated setups that I would otherwise need to recreate perfectly every time.

Record A ‘Temp’ Piano Part

Once the template has generated my new session, I will import whatever video/picture I am working against (if any) and start temping out on piano and finding a flow to whatever I am working on. Starting by temping with a piano has two benefits. If I am working on a project that doesn’t have pictures, this lets me more quickly compose without the need to fully orchestrate things out. I can leave that until later when I have the time to fully realize the orchestration I’m hearing in my head.

If I’m working on a project that does have video, then I can easily move the piano around and change just this one instrument rather than trying to fix a whole orchestra of MIDI all in one go. Changes to how things are structured and where things are against picture most assuredly will change. So temping it out in piano gives me more flexibility by simply not having as much to worry about changing as changes are guaranteed to come.

Check With The Client

Once things are temped out on piano, this is the first time that a producer, music super, or director will listen to what I have written. I find this is a crucial step as it has helped reduce the number of changes I have to deal with later on because the rest of the team really did not like the idea I am working with.

There is usually a bit of unease around the piano temp because it usually doesn’t sound quite right (or in some cases really not right), but I let the team know that I’m looking more for their thoughts on the melodies, accompaniment, and structure rather than the way it sounds.

Sometimes ‘Temp’ Orchestrate It

Once they are happy with the way the ideas are sounding (with some trepidation), I will spend the time to properly orchestrate it out. Notably, sometimes the post team will not really grasp what I am trying to convey to them musically from just a piano. In those cases, I will spend the time orchestrating it out for them so they get the idea that I am trying to convey.

Get The Tempo And Hit Points Right

The reason, however, for not spending the time orchestrating it in the first place is because if they don’t, most importantly, like the tempo I am working at, it is easier to change that by temping it in a piano since I don’t use a tempo ruler while temping. So changing tempo is as simple as playing something new at a different tempo and not worrying too much about precisely timing note starts etc. This lets me figure out what feels right to hit “hit points” and figure out the precise tempo I need later to achieve this.

So to summarise the past couple of paragraphs: temp things out; get your post team to approve your temp; make changes; get approval; lock it in; orchestrate and tempo-rule out the temp piano.

Orchestration

Orchestrating is next. This is quite a tricky thing for me to make suggestions on as it is quite an art form, and as such, it takes a lot of practice to find what works best for you and what approach is best for what you are working to create.

One of the best suggestions I can give you is to get a copy of Samuel Adler’s: The Study of Orchestration. This is the book used by my university in the Composition Department and seems to be one of the more popular books for orchestration out there, and for good reason. Mr Adler does a fantastic job at condensing a lot of information into easily digestible chunks that are both easy to use and retain the essentials for each element.

Now, this doesn’t deal with the topic of synth use, but the ideas used across the text can be extrapolated to any use of synths that you wish to use. The building blocks of an orchestra aren’t that different from a synth texture but simply a different way of creating large musical works or complementing traditional orchestration. Here are some highlights from my interpretation of Adler’s text…

  1. Know how best each section of an orchestra functions and what they sound like individually and as a group:

    • Strings: homogeneous sound, technically very nimble, can create a huge sound or a very small one (by writing for just section leaders).

    • Winds (aka Woodwinds): heterogeneous sound, technically very nimble, have a more limited dynamic range (can only get so soft or so loud without the instrument not working as intended), great for adding character and embellishment to string orchestration in addition to functioning as a Woodwind choir. Make sure you leave space for a performer to breathe and for their embouchure to recover.

    • Brass: homogeneous sound, slowest moving of all the sections (in general, French Horns and Tubas especially are not as technically nimble as their string counterparts), large dynamic range weighted towards the loud end of the spectrum (the other section, like Woodwinds that can have difficulty playing extremely soft)

    • Percussion: mostly heterogeneous sound (depending on the ensemble), technically the most nimble of the orchestra (in general, exceptions do exist), have a huge dynamic range as a whole. This ensemble in the orchestra is also the most varied with the potential for a huge range of sound in addition to an almost limitless timbre.

  2. Use choirs to either play the same thing or counterpoint/texture. Choirs in this case mean one of the sections listed previously.

  3. There are well-established practices, however, don’t let them dictate how you want to write. The “rules” are more guidelines but help you achieve a good sound faster. Learn the rules, so you can break them conscientiously.

Mix The Music Against A Guide Track

Once orchestration is done, the time comes to mix what you have written against any other material you have (i.e. dialogue, fx etc). If you are ‘one-man-band-ing’ this, then this will be entirely up to you. But if you are part of a larger team, then this could be down to a score/music mixer.

As I have only worked on small independent projects so far, I can only speak to the former workflow. However, what I can suggest at this point is that your music track shouldn’t require a huge amount of effort to mix. Ideally, you will have written something that complements what is happening on screen (or the other elements of your project), and so the more elements there are competing for attention, the lighter your music should be.

Note: ‘Should’ is a huge caveat here. There are a number of exceptions to this. Again this is more of a guideline rather than a rule. Let the story dictate how you score, but keep in mind that you are competing with possibly many other elements, all of which need to be heard clearly.

Ideally, you will only need to mix general sections up and down as the other elements of the project require (i.e. dialogue will require you mix your music down so the dialogue can clearly be heard). This, hopefully, won’t require too much fiddly work on your part as the orchestration and composition take care of most of the “mixing” that needs to be done.

Check With The Team

All through this process, I check in with the rest of my production team making sure as I get major things done that they approve of what is being done, so I can address any of their concerns or problems that they have with what they are hearing. Much better to do this in chunks if you can, rather than all at once at the “end”. So by the time you get to the end of this process, you don’t end up with a lot more work to do.

Bounce Out A Mix And Stems

Once I complete a final version of my music, I do a re-record pass of it in Pro Tools, rather than a conventional bounce. This lets me take a final listen in case I missed anything and to do a quality control (QC) check on what I am about to deliver.

It also lets me (because of my template) record all of the stems I’ll possibly need later down the road (as discussed earlier when mixing against picture).

With these re-recorded stems, I have everything I need to go into the final process of either handing off my work to another engineer to mix or to mix it myself in a different Pro Tools session with all of the other elements of the project.

My re-recording also generates lower-order final mixes as I work in a 7.1.2 bed format (for future flexibility), so during re-record I’m generating 5.1, and 2.0 final mix stems along with everything else which is also important.

Immersive Stems

In my stems, I generate 7.1.2 beds of, Strings, Woodwinds, Brass, Percussion, Synth among many other more granular possibilities down to the individual instrument as needed.

I do all of my routing inside Pro Tools rather than letting Vienna Ensemble Pro handle the sub-mixing and only sending the sub-mixed busses to Pro Tools. I prefer to keep that process in Pro Tools to retain the maximum flexibility for anything I need to do when creating MIDI scores. This also means I deal with my projects in a similar fashion to how I would if they were recorded with a real ensemble, meaning, when one day I get a project that has the budget for it, I’ll have little difficulty transitioning from MIDI orchestration to a live orchestra.

Some Pro Tools Specific Notes

I’m going to finish this article with some things that are more Pro Tools specific in regards to MIDI composition.

Many Voices

For my specific template and writing style, the added voices introduced in Pro Tools 2020 for non-Pro Tools Ultimate rigs has been essential for all the routing of VI’s in my template. Just be aware, if you are running any non-HD rigs, you will have a more limited voice count to work with than if you were using Ultimate.

Economical Use Of Voices

So it would be worth planning out what you will be using and trying to get an idea of how many voices you will be needing to do what you want. If you end up running over how many voices you have available, look for a place to consolidate instruments.

For example, I use a texture synth called Gravity made by Heavyocity where I have all of its patches loaded at once, instantly available via their individual MIDI tracks. This would be impractical and a huge waste of the limited resources in Pro Tools to have each one of these patches as a separate audio source in Pro Tools.

So I have consolidated each category of texture into two audio tracks, ending up with about 8 stereo audio tracks instead of hundreds. This is also why I compose using keyswitches with BBC Symphony Orchestra instead of splitting out short and long articulations like some people like to do. This means instead of having multiple Violin 1 tracks, I just have one where all of the articulations appear, just as if this were a real section playing.

The MIDI Edit Window

Another quirk of ProTools is the MIDI edit window. While, yes, there is a shortcut to quickly get to this window (control + “=”), I have found it cumbersome enough that it was a major consideration in my switch back to OSX last year. (In Windows Pro Tools the Mix, Edit, and MIDI Edit windows are stuck in what I like to call a “sandbox” meaning they can’t really easily be moved between computer monitors, especially in my specific computer monitor setup).

My Preferred Screen Layout

I prefer to have the MIDI Edit window on my left screen sitting there permanently as I am constantly bouncing between the MIDI Edit window and the main Edit window while also having the video window on a third display centred above the other two monitors.

My suggestion is to have a second monitor as your MIDI Edit window, if possible. This will make things much easier (on both Windows and OSX systems).

Vienna Ensemble Pro And Pro Tools

In working with Vienna Ensemble Pro (VEPro) with Pro Tools, there are a few things I’d like to share that I have found out the hard way.

The most stable settings for using VEPro are in Pro Tools, a Hardware buffer of 256 samples, and in the VEPro plugin, a buffer size of 2.

Note: Don’t expect to be able to get your Hardware buffer below 256 samples, no matter what your VEPro buffer is set to. Either VEPro or Pro Tools doesn’t like this, but it hasn’t worked for me below 256 Hardware buffer size with a VEPro buffer set to 2. I know the VSL manual for VEPro says it can do whatever buffer size you want if you have enough CPU horsepower to deal with it, but I haven’t found this to be the case. So I would suggest these settings to you too as the most stable.

If you want to be able to record the audio coming out of channels 1 and 2 of a particular VEPro plugin, I would suggest this workflow for doing that…

  1. Instantiate the VEPro plugin (not the audio one) on a stereo Aux track

  2. Label it whatever instance you are connecting to.

  3. Label the output of that track whatever the VI/Instrument is coming out channels 1 and 2 of the VEPro instance.

  4. Create an Audio track as that instrument connecting the input of the Audio track to the output of the VEPro Aux track.

  5. This will let you record the audio from that track and do whatever processing to it you need to just like a normal audio track.

I found that instrument tracks don’t work for this as they only record MIDI and playback audio that is generated by MIDI through the VI on the instrument track. There is no reason to waste an audio track doing this.

You can’t record channels 1 and 2 from a plugin on an audio track as plugins occur after the input stage of an audio track in the signal flow of Pro Tools. This also means troubleshooting is a bit easier, as all of your VEPro plugins are isolated from the audio tracks, meaning there is a bit less to deal with and you have audio and VEPro separated.

Know Your Shortcuts

If you find yourself doing something semi-frequently, look for the keyboard shortcut for that thing. It will save you time, and, eventually, you’ll have the muscle memory for that shortcut. Just make sure you are using it, even if that means undoing what you just did to use the shortcut. That is how I learned them, even if I forget them occasionally.

When editing MIDI in Pro Tools and within the MIDI Edit window, F10 is your friend. That is the pencil tool. It lets you draw MIDI automation and notes quickly.

F6 + F7 gets you back to the Smart Tool.

Label Everything Clearly

Long labels are OK; it's just text on a computer screen. When labelling busses, it can be very helpful to create dummy busses and label them as divisions for different sections of busses (i.e. >>> VEPro Busses <<< is how I label my divisors) as shown above.

Busses don’t use up voices so go crazy with it! I actually picked this trick up from some of the post-work I have done and have found it very helpful for quickly finding things.

Get Picture Lock

Make sure you ask for ‘picture-lock’ before you start really doing anything too in-depth. There is no reason to spend a lot of time on a project that you very well may end up entirely tossing later on when you don’t need to.

You’ll be doing plenty of scraping work later even after you get picture lock because inevitably something will happen that means you have to backtrack to change something, or something ends up not working later on, or any other myriad of reasons.

All of the clients I have worked with so far have had no issue accommodating my request for picture-lock.

When I have done composition work before I got picture lock, it was usually a request to hear what I am thinking for a scene after we do a Previs (Previsualization) spotting session. So I’ll write that out quickly to give them an idea of what I am thinking. Later, I know I’ll have to redo all of that work, but that is OK because Picture will likely change by then anyway so I’ll have to redo it anyway.

Plan Ahead

Plan ahead as much as you can. For example, on an indie web series I worked on, even though I was the only sound guy, I planned to do a reel workflow with each episode on the hour. This gave me plenty of space around each episode to put miscellaneous work and gave me room and flexibility to move tempos around to figure out the best fit for a scene without disrupting any of the other work I had been doing.

One Session Per Reel Or All Together In A Single Session?

Now, you could have a separate session for each reel, which could be helpful for preventing anything from shifting as you work on composing for that reel. However, since I am working on a longer-form episodic program, I felt the benefit of having all the episodes in one timeline out-weighed the potential downsides.

I just have to be careful to make sure I don’t shift any of my other work for other episodes. If you are in doubt about a change you are about to make irreparably harming your work (especially with something that can’t be undone), “save as” your session and try out what you are wanting to do.

Remember that we are dealing with digital storage space, and Pro Tools session files are not terribly large. Do this as often as you need to.

Just make sure you go back to your original Edit Session once you have either confirmed the change works and/or is non-destructive or you have found a workaround.

See this content in the original post