The Sound library is a Haskell library for creating sounds or series of sounds.
Refer to the Installation.txt file provided.
The main functions of the Sound library are makeSound and makeSoundTitled. These functions take a Sound Sample (aka Sound [Int32]) object and a double indicating the length of the sound to be produced. makeSoundTitled also takes a string that is the name of the file produced. makeSound uses a default name for the file produced. The produced sound is saved as a .wav file inside the sounds directory.
There are 9 constructors for Sounds:
SSolid (solid) is a Sound a that contains an a.
STime (t) is a Sound Double that returns the current time.
SSound (sound) is the constructor for a basic sound. It takes a Sound Double indicating the frequency, a Double indicating the length, and an int indicating the volume.
SStart (delay) is a Sound a that takes a Double and a Sound a and delays the Sound a by the Double's amount of time.
SConcat (-->) concatenates two Sound Samples together, so that the first follows the second.
SSpeed (slow, speed) takes a Double and a Sound a and makes that Sound a faster by a factor of its Double.
STrans (trans) takes a Double and two Sound as, then transitions from the first Sound a to the second Sound a the Double amount of time before the first Sound a ends.
SMap (mapSound) maps a function and a Sound a and maps the function to the Sound a.
SZip (zipSound) maps a function between a Sound a and a Sound b.
Sounds also support basic arithmetic operations such as addition and multiplication, and basic trigonometric functions.
By using these operations in conjunction, one can create long, complex sounds.
The interpreter takes a Sound a and an Int indicating the starting time for the Sound a and returns a WAVEFunc a. A WAVEFunc a which is the type (Int -> a, Int), where the first item in the tuple is a function from the time to an a, and the second item in the tuple is the length of that Sound.
To evaluate a Sound Sample, as many times are used as the length of the sound in seconds * number of frames per second samples are evaluated. A WAVEFunc Sample is an (Int -> Sample, Int), so we can concatenate all of the Samples from applying the time to the function inside the tuple together to get a list of Samples. The WAVE library then takes this list of samples and creates a .wav file.
Example sound samples are near the bottom of Sound.hs. The Sounds produced from these examples are in the examples directory.
An obscene amount of conversion between ints, int32s, and doubles was required.
It was tedious to write (maxBound div 2) or (maxBound div 3) to write the most common volumes for sounds, so full, half, third, and quarter were created to simplify writing sounds.