While working on music for film, advertising, radio or TV, composers use a wide array of music sample libraries.
And no wonder.
Sample libraries are wonderful tools. They work perfectly for percussion sections or repetitive ostinatos with few articulation changes. Everything they can read from a MIDI file imported from Sibelius or Finale or what we can perform by ourselves on the keyboard significantly saves time.
Sample libraries also enable composers to introduce numerous changes to the project, which is both inevitable and beneficial in the production process.
Using computer to produce music have many benefits:
- It offers great possibilities of manipulating samples,
- It accelerates the production process,
- It is less expensive than traditional recording,
- It gives creators-producers the comfort of deciding about the final sound of the track by themselves.
There are, however, situations when sampling music just does not work, and achieving the desirable quality of sound would require many hours of tedious work or, is just impossible.
A better solution: Mixed Technique
If a piece of music contains an articulation, tempo or character changes, then achieving a good effect by means of samples alone can prove a very hard, time-consuming and sometimes impossible task.
We discovered that the best option is a mixed technique, where the power and diversity of sound are achieved by the use of samples, whereas solo parts and selected section parts are recorded by session musicians with real instruments.
Case 1: Art of articulation
Samples work great in repetitive, preferably percussive structures.
If, however, we want for instance a viola section to play a legato run followed by a pizzicato run, and then to blend via a long sul tasto note into a chord, then problems may arise. For each one of these articulations, we need to establish a subtrack or sub-line with separate sound parameters.
If a quintet is involved, we can expect further hours of clicking. The MIDI format cannot automatically adjust articulation based on marks in the instrumental parts. We need to set it up manually.
By contrast, a musician is able to sit down in front of the musical score and realize all the composer’s instructions at one go.
Case 2: Natural tempo changes
I once happened to work on a sample version of Johann Strauss’s waltz The Blue Danube. In order to render the swinging character of waltz, I prepared a tempo map based on a classical live recording. For the first minute of the piece only, I needed as many as 60 tempo changes (!), and the final effect still did not sound natural.
It may be an extreme example, and yet any speed-up, slowdown or suspension of tempo (all of which belong to the composer’s basic means of expression) introduces a good deal of complication in the DAW and significantly extends the time spent over a given piece.
Case 3: Masterpiece of solo parts
Sampled orchestra can be used to perfectly imitate some typical symphonic sounds.
But the moment the score features a part of “solo oboe”, or “solo trumpet”, or any other solo instrument for that matter, samples prove to be no solution.
Similarly, they fail to work for the imitation of solo human voice.
Libraries like solo violin or cello solo do exist, but being a professional cellist myself, I must frankly admit that they have just as much to do with the the sound of the cello as KFC has with roast chicken.
Case 4: Great wind sections
The expressive power of wind instruments relies in the fact that when a group of professional musicians plays a chord, they listen to each other, while particular chord components produce a different timbre.
Good instrumentalists know or intuitively feel whether they are playing the lead voice, counterpoint or only background for someone else.
In order to even approach this effect by samples, we would need to animate each “musician” separately and select the timbre, velocity and volume of sound for each particular moment of the composition.
The process turns out so arduous and time-consuming that it is indeed much simpler and faster to invite four good hornists to the studio.
Case 5: Beautiful string legato
Here we are facing a similar problem as in the case of wind instruments, only the instruments’ specificity differs.
String section produces its own unique timbre as several musicians play the same tone on different instruments with different bows – and every one of them plays it slightly differently. A broad, lush sound is achieved in this way, rich in various overtones.
In order to create something alike, we would need to use several various libraries and still individually equalize and “humanize” each one of them.