(Click on a section. . .)
A Little Primer on Sampling
Sampling
What is a "sample"? In simplest terms, it is a snapshot of an audio waveform, which can be stored in digital form. Analog recording changes the voltage patterns in the magnetized oxide particles on the recording tape according to the signal coming in from a microphone, while digital recording turns the voltage changes into a series of 0s and 1s which are saved onto a disc at the current Sample Rate and Bit Resolution.
| Analog Audio Recording | ![]() |
| Digital Audio Sampling | ![]() |
The rate at which the data is saved determines the quality and accuracy of the sound. This is called the "bit resolution" and the "sample rate",
Bit Resolutions The higher the resolution, the smoother the resulting sampled waveform. |
![]() |
Sample Rates Again, the higher the rate, the smoother and more accurate the sample. |
![]() |
Once a Sample has been recorded, it is processed to make it usable for a keyboard instrument to use. The original .wav (wave) file is loaded into an audio editing program, and extraneous noise is edited out. Then a crucial operation makes it suitable for use:
Selecting the loop portion of an audio sample |
![]() |
HOW IT WORKS . . . 1. When a key is depressed, the computer in the console locates the appropriate audio sample for the stop and note. 2. An attack waveform derived from the sample is played once. 3. The loop waveform plays over and over and over until you let go the key. 4. A release waveform, derived from the sample is played once. |
What is Real Time Sampling
Sampled audio is just what its name implies: it is a reproducing system - that is, it plays back a recording of an already existing
pipe sound, as contained in a single audio file. It has had these major drawbacks:
| The only way the basic sound within a sample can be changed, is by loading in a different sample. There is no way to change the basic sound except by manipulating the waveform before it is placed in the organ's memory or by applying filters to remove certain parts of the waveform.. |
| The crucial attack sounds which give vitality to the individual notes, cannot be changed dynamically. |
| Because it is a reproducing system, the various individual notes cannot interact one with another as do wind-blown pipes. |
| Because of memory and speed requirements, most stops use only 3 - 5 samples, which means that groups of anywhere from 5 to 25 contiguous notes will have exactly the same tone color and attack. |
One can, however, change the general characteristics of a stop's sound by making adjustments to the entire stop at once outside of the sample. These changes include:
Most manufacturers have software routines wherein the volume level of each individal note can be altered individually. Some have stereo speaker systems to give the illusion of spatial independence.
But, when all is said and done, sampling has remained a reproducing system in which one cannot create new sounds, nor dynamically alter the sound contained within each sample.
And that's not organ building . . .that's manufacturing one heck of a big interactive iPod...
However, with the decrease in memory costs, the increase in CPU speeds, and the rise in the level of audio technology, the advent of real-time sampling has turned this staid system into a versatile interactive means of creating sound. This is done by storing the original waveform - but now in greater detail than ever before - in the organ and doing the processing required for looping and other functions as the organist is playing!
The use of velocity-sensitive keyboards allows us to control various aspects of the sound in direct response to the organist's playing.For example, you may know that when you repeat a note on a pipe organ (both on tracker and electric actions), depending on how much time elapses between keystrokes, the attack of the pipe changes due to the varying amounts of pressure inside the pipe foot. we can now have 5 or 7 samples of various attack sounds directly taken from various samples, so the ictus (attack) of the note changes from each keystroke to the next.
The difference in attacks - slow, faster, and fastest - can be plainly seen in the first three waveform samples shown below. Of even greater interest is the fourth waveform, which shows the response of he waveform transients to three repeated notes.
What does Innova do with all this?
We apply our experience with the almost infinite nuances of the winded pipe organ to the digital instrument. While we make use of the technology involved, we always attempt to, first and foremost, keep it human. The technology is not our raison d'etre...it is the music.
To be continued . . .