avesk / echo-synth Goto Github PK

Given a signal, return an dict of envelope values. Each value represents the duration of each feature.

Input:

• signal - array of signal data points

Output:

• attack - float
• decay - float
• sustain - float
• release - float

Maybe wrap this into a dictionary

envelope = {
    "attack": 0.00,
    "decay": 0.00,
    "sustain": 0.00,
    "release": 0.00,
}

Given a signal and an amplitude track construct the resulting signal. Multiply each signal point with corresponding amp track value

Input:

• signal - array of signal data points
• amp track - array of amplitude data points

Output:

• signal

given attack, decay, sustain, release, produce an amplitude track that models this envelope

Input:

• envelope - probably a dictionary containing 'attack', 'decay', 'sustain', 'release'

Output:

• Amplitude track of envelope

Function to generate signal data given a freq, amplitude, and osc type

Input:

freq - int between 20 - 20K
amplitude - float between 0-1
type - string ['sinusoid', 'square', 'triangle', etc...]
duration - float

Output:

resulting signal

Given a signal return the top (10?) frequencies that are present in the signal. Make sure these frequencies are percievable by the human ear and regular hardware(within 40hz-4Khz maybe?)

Input:

• Signal - array of signal data points

Output:

• List of most relevant frequencies in the signal with an amplitude

output = [(440, 0.8), (250, 0.3), (200, 0.1) ... ]

For these note books to be usable by others we will need some level of interaction. Sliders will be used to modify the state of the envelope for example. Implement sliders into jupyter notebook so we can use the sliders to adapt the values of our envelope interactively.

docs