User Manual
Crumar Seven User s Manual '
CHAPTER 5. ABOUT POLYPHONY, VOICES VS. NOTES, HOW VOICES ARE ALLOCATED IN DIFFERENT SYNTHESIS MODELS.
Ever wondered why some digital piano, despite having just 88 notes, offer a total polyphony of 128, 256 or even more voices? Here's
the answer. First, there's a difference between a note and a voice. A note corresponds to a key on the keyboard, it's what we play when
we put our fingers on the keyboard; a voice is an element of an instrument that is responsible of reproducing a single sound. Imagine a
small pipe organ, it has probably just one 61 note keyboard, but since it's capable of playing different “registers”, it will probably have
more than 500 pipes! Each single pipe is actually a voice, because a single note can play more than one voice at a time. Therefore, many
sample-based instruments need many voices to be able to offer an acceptable amount of polyphony. One more question: how are
voices allocated, or employed, each time a note is played? This varies according to how the instrument works. Some sample players
need two voices to play a stereo sound (left and right channels), other need more than one voice to play the basic tone plus some
details (an example could be an orchestra ensemble that uses two voices for the sustaining strings sound plus two more voices for the
attack transients). So, each time a new note is added to form a chord, new voices are allocated, until they reach the polyphony limit.
This leads us to the next question: what happens when the limit is reached? Again, this varies according to how the playback engine
was designed. The most common situation is the simple “round robin” cycle, where all voices are allocated starting from 1 up until the
limit is reached, then back to voice 1 and so on. Of course, if a voice is already playing and is now destined to another note, it gets
interrupted. This case is called “voice stealing”. Some players use intelligent methods to prevent frequent voice stealing occurrences,
some try to steal the older voice, other steal the quieter voice (based on the amplitude envelope). And now the last question: how are
voices allocated when the sustain pedal is depressed and the same note is being played over and over (like in a “ribattuto”)? Well, most
sample players just don't care of what note is being played, they just allocate the requested voices and do their job. Other players
allocate only a pair of voices and alternate between them, others allocate a limited number of voices, like 4 or 8, in order to prevent the
overlap of many voices playing the same sound.
This was how
other instruments
use polyphony. Now let's learn how Crumar Seven works.
Crumar Seven User's Manual - Page 10