clockblocks 0.2.1

A python library for controlling the flow of musical time.

clockblocks

clockblocks is a python library for controlling the flow of time, designed with musical applications in mind. In particular, it is a central component of scamp, a Suite for Composing Algorithmic Music in Python. It is described in detail in this paper.

A Clock acts like thread, but with the advantage that when multiple clocks are coordinated under the same master clock they remain precisely coordinated and do not experience drift. Furthermore, processing time is taken into account when "wait" is called in a given Clock. For example, the following program:

import clockblocks
import time
import math

clock = clockblocks.Clock()
start = time.time()

while True:
    print("Current time: {}".format(round(time.time() - start, 4)))
    # do some pointless and time-consuming calculations
    for i in range(1000000):
        math.log((i+1)**0.7)
    clock.wait(2)

... generates the output:

Current time: 0.0
Current time: 2.0001
Current time: 4.0001
Current time: 6.0
Current time: 8.0001
Current time: 10.0

Whereas a traditional thread:

import time
import math

start = time.time()

while True:
    print("Current time: {}".format(round(time.time() - start, 4)))
    # do some pointless and time-consuming calculations
    for i in range(1000000):
        math.log((i+1)**0.7)
    time.sleep(2)

...will gradually drift because of the intensive calculations, outputting:

Current time: 0.0
Current time: 2.3772
Current time: 4.7623
Current time: 7.1397
Current time: 9.5151
Current time: 11.893

In addition, clockblocks offers useful musical functionality, like sudden and gradual changes of tempo. Perhaps the most exciting feature of clockblocks is that clocks moving at different tempos can be nested within each other. In this case, each clock distorts time for those underneath it: a clock whose tempo is oscillating between slow and fast, nested within a clock that is accelerating, will generate a time stream whose tempo oscillates faster and faster.