In this video I discuss some of the considerations in place when building my template for composing in Cubase 6.
If you are a working composer in today’s industry, time is a big concern. And it’s a big WASTE of time to always be setting up new projects, especially larger ones. It can be great boost to your efficiency to utilize a template or multiple templates in your DAW. Thats what I’m going to be talking about today, so let’s get started.
Today I am going to walk you through the (rather large) composing template I built in Cubase 6. I originally built this template in Digital Performer, and then replicated in Logic, but now that I have some free time I have gone ahead and recreated it in Cubase as well. This template is geared very much toward composing for video, and you’ll see how in a moment. So lets go over the essentials.
Before looking at anything in the edit window, lets review how I’ve configured cubase. I come from a background of digital performer and pro tools, and have gotten used to certain key commands. Almost all of the functions from these other programs exist in Cubase, but do not have logical key commands, or even key commands at all. Luckily Cubase excels at configuring and saving these, so I have gone ahead and manually entered all of the most useful and common commands into an xml file which you can import and use to enhance your own cubase workflow. Now lets get started.
Up top I have a marker track, signature track, and a tempo track. When I’m working with video, I can hit control+M to place a locked marker. I can then use the time-warp tool to make microscopic adjustments to the tempo on either side of the marker so that certain musical elements sync visually. This in turn will alter the tempo track, creating a long tempo map of ever-shifting tempos.
Below my global tracks, I have MIDI tracks organized into folders according to traditional score order. All-in-all my template is comprised of 84 individual MIDI tracks, which I have reduced to as few as possible by using program changes. The program changes act as keyswitches for instruments that have many articulations, such as any of the strings. Where now I have one track for Violin 1 A, i would need to have about 5. 84 tracks may seem like a lot, but in reality it has been reduced by almost a factor of 5.
The strings in my template are particularly flexible. They are divided into divisi, Violins 1 FC, A, B, and C. Each track is loaded with about 5 different articulations including legato+portamento, pizzicato, spiccato and tremolo. Rather than playing the strings like a keyboard patch, I manually record each track separately to capture unique performances which adds a more lifelike and dynamic quality to string passages.
The brass I’m using is Cinebrass Pro from Cinesamples, a fantastic company that puts out the highest quality virtual instruments. I’ve organized it in a similar way to the strings, though rather than divisi I’ve simply separated the soloists from the ensemble for each instrument.
Below all of the midi tracks are VST Instrument tracks, and group channels which I use to export audio stems, and reverb on its on own audio track. More on that later. If you haven’t noticed by now, ultimately the template is organized according to signal flow, from top to bottom. We start with the midi notes, which trigger the VST instruments, which then end up as audio submixes or “stems” for further mixing in pro tools.
Each group channel saves as an audio stem. On each channel is a reverb insert, which has only a small amount of predelay to function as the early reflection. Each of the channels then send to varying amounts of tail reverb, which is a long echoing plate reverb. Plate reverbs are good for the tail, because they are very transparent sounding, and since the audio signals all add up, its good to stay as transparent as possible to avoid a murky sound. This is one of problems with convolution reverb, and I why I generally avoid it. Convolution reverb is an audio sample of a spaces reverberation, which is recorded by using a starter pistol or sine wave sweep. But it is just ONE audio sample. Physical spaces are very dynamic, and treating every single sound in your mix with the same impulse yields a very flat “low-contrast” sound as I like to call it. To solve this problem, I choose to use multiple algorithmic reverbs in tandem with each other, which yield reverb that feels more realistic. I have taken the time to program presets for the different depths in 3d space that players will sit in on the stage, using more or less pre-delay. You can take it a step further to add a shimmering quality to reverb tails that make it feel “alive” by adding some light chorusing. I don’t find it necessary however.
Finally lets look at the engine that powers this monster template. If I hit F11 and bring up my VST instruments rack, you can see I have 12 instances of Vienna Ensemble Pro 5. Each of these is connected to a server instance on another computer I have under my desk, which has 24gb of memory and over 6TB of hard drive storage. This was the part that took the longest to set up. Inside most of the server instances are Kontakt 5 instrument banks which I’ve created for the purpose of using program changes to toggle between different articulations on the same midi track back in Cubase. The latency is non-existant, and performs without error. I’m only using one computer screen, so I setup Windows Remote Desktop on the server which also has very low latency.