Some background: Most of you probably don’t know this because I haven’t talked about it much, but before I got involved with music on YouTube, I used to be a sample library programmer/producer (proof). A pretty good one. My products were given high star ratings and reviews in Electronic Musician magazine, Sound on Sound and elsewhere.
My first sample library release was so successful in fact that the Garritan company – makers of the Garritan Personal Orchestra, which is shipped with every single copy of Finale Music Notation software – flew me out to Seattle for a week-long interview. I was given the job, but unfortunately had to turn it down due to the amount of unforeseen traveling that would have been involved.
During my time working in the sample library field, I got to learn and collaborate with some very knowledgable music makers and music tool makers. I was an official product tester for Native Instruments. I was on a beta team for GigaStudio. But most importantly, I got to have regular phone calls with Tom Hopkins. Tom is the man behind the programming of the Garritan Personal Orchestra, which was an industry changing product when it was released.
Tom and I got into a number of discussions about composing, arranging, and performing music using nothing but sampled instruments. His feelings were strong that sample library technology had given us an opportunity to finally make music “perfect”.
What’s a sample library? Let me explain. Typically when you play a keyboard, you are hearing a synthesized (fake) sound. That sound resembles a piano, or a violin, or a harp, or a trumpet. But it is not a recording of any of those things, it is synthesized.
On the other hand, a sample library is an actual recording of every single note on any given instrument. The more expensive sample libraries include not just every single note, but also every single note played in a variety of ways (articulations), and at a variety of volumes (velocities (different velocities are important because some instruments change sound (timbre) dramatically when they are played softly compared to when they are played loudly).
So, when using a sample library, when you hit the A above middle C, you trigger an actual recording (or a “sample”) of someone playing an A note on a violin (or any other sampled instrument you may be using). And most violin sample libraries aren’t just recordings of random violin majors, no, national orchestras have been sampled, one player at a time, to build massive orchestral sample libraries. They are recorded in state of the art performance halls for reverb-baked-in libraries, or in professional isolation rooms for completely dry libraries that you can add your own ‘verb to in post production.
SIDENOTE: The most impressive orchestral sample library is the Vienna Symphonic Library. A special facility was built in Vienna solely for the purpose of recording, editing and programming this library. The full library consists of 1,143,482 individual note recordings and costs $13,990 to purchase. The total package clocks in at 746 GBs of data. But it sounds amazing! ;)
Okay, back to the individual sample recordings…
Through the use of MIDI programming and a DAW (Digital Audio Workstation, like Logic, Cubase, ProTools, Sonar, et al.), you can actually control whether each note is bowed up on the violin, bowed down, sliding to the note from a higher note, sliding to the note from a lower note, playing the note harshly, playing the note gently, playing with vibrato, playing without vibrato. Some libraries even let you control which string the note is being played on.
Even if you’re not a music composer, you can clearly see how powerful and revolutionary and… cool this technology is. You basically have a 100 piece orchestra dangling from puppet strings tied to your hands, mouse and keyboard.
You control all of the above with MIDI data. MIDI records and plays back a musical performance, but it’s not an audio file. It’s a list of commands. A single MIDI data file will tell your sampling software to play an A note, slid in to from a lower note, with vibrato, for two beats. That same data file will tell your sampling software how loud and hard to play the note, and will include any minute pitch bends or muting or other playing/performance cues you can think of. You get up to 127 different commands for each and every note on each and every instrument!
Want to learn more about MIDI? I made TWO separate VIDEOS about MIDI data a few years ago.
An additional benefit, because all this MIDI data is not an audio file, you can edit every single command independent of the others. Let’s say I program in a perfect pitch bend, vibrato, volume and slide for my note, but then, “oops, that was supposed to be an A sharp, not an A!”. No problem, I change just one of those 127 commands, and now all that perfectly programmed data plays back just the same, only now it’s a recording of an A sharp, not an A.
If your mind’s not completely blown right now, you’re not human.
MIDI and Controller Data in Cubase
With all of this insane note and performance control, you could compose an entire original symphony, performed by a first-class orchestra, one note at a time, completely from your desktop. The resulting recording would be made up of thousands of individual actual recordings of individual notes, no synthesized sounds. And, most importantly for my friend Tom Hopkins… you could make it perfect. Not one mistake, not one early note, and zero “swing” on any of the notes.
Tom went on to mention some of the other exciting aspects. Composers could now write pieces that could never be performed before. Computers can play notes faster than any human ever could. Computers can sustain those faster tempos indefinitely. Computers could sustain notes longer than is physically possible in the world we inhabit. No longer would composers have to search endlessly for a virtuoso who could handle their works. Their computer could do it, just as easily as it could check their email for them. And given the recording quality of the individual sampled notes, very few would know the difference.
So is all this perfection really better than a human performance? Sure, humans make mistakes. Humans are not metronomes. Humans add swing to triplets and other note patterns. But is the alternative not stale? Is the alternative not… soulless?
This is the question I am currently struggling with as my band The Caulden Road begins production on our first album. Sure Christian and I can program in every drum and percussion hit. We could program in pianos and violins and choirs and harps and anything else that’s ever been sampled. And it can be perfect. And it can sound like a million bucks. But will it still have life? Will it still have passion? Hiring choirs ain’t cheap. And an orchestra won’t fit in my office/home studio. And maybe perfect is what the modern music fan wants anyway. Quantized rhythm sections, pitch-corrected vocals.
It’s just, something about it all might be too synthetic for me.
PS. Wanna hear a sample library in action? This is an original piece I composed and programmed using a “lite” version of the VIenna Symphonic Library mentioned above, along with a few sampled instruments from my own private collection I’ve produced:
Lullaby by Alan Lastufka
{ 2 comments }
