mc.pfft~ Reference

Spectral processing manager for patchers (multichannel)

mc.pfft~

Description

The pfft~ object is designed to simplify spectral audio processing using the Fast Fourier Transform (FFT). In addition to performing the FFT and the Inverse Fast Fourier Transform (IFFT), pfft~ (with the help of its companion fftin~ and fftout~ objects) manages the necessary signal windowing, overlapping and adding needed to create a real-time Short Term Fourier Transform (STFT) analysis/resynthesis system.

Discussion

The number of inlets on the pfft~ object is determined by the number of fftin~ and/or in objects in the enclosed subpatch. Patchers loaded into a pfft~ object can only be given signal inlets by fftin~ objects within the patch. See fftin~ and in for details.

Arguments

subpatch-name [symbol]

Name of patcher to load

FFT-size [int]

Optional
Unitssamples

Specifies the FFT size, in samples, of the overlapped windows which are transformed to and from the spectral domain by the FFT/IFFT. The window size must be a power of 2, and defaults to 512. (Note: The size of the spectral "frames" processed by the pfft~ object's subpatch will be half this size, as the 2nd half of the spectrum is a mirror of the first and thus redundant, unless the full-spectrum-flag is present.)

overlap-factor (hop-size-denominator) [int]

Optional

The third argument determines the overlap factor for FFT analysis and resynthesis windows. The hop size (number of samples between each successive FFT window) of Fast Fourier transforms performed is equal to the size of the Fast Fourier transform divided by this overlap factor. (e.g. if the frame size is 512 and the overlap is set to 4 then the hop size is 128 samples). The value must be a power of 2 and defaults to 2. A value of 4 is recommended for most applications.

start-onset [int]

Optional
Unitssamples

The fourth argument specifies the start onset in samples for the Fast Fourier transform. It must be a multiple of the current signal vector size and defaults to 0.

full-spectrum-flag (0 or nonzero) [int]

Optional

A non-zero fifth argument may be used to specify "full-spectrum mode". In this mode, the pfft~ object will internally compute a complex FFT and process full DC to SR mirrored spectra (instead of simply eliminating the redundant half of the spectrum). This takes extra computing power but may be potentially useful in some of the more esoteric spectral processing applications.

'args' and list-of-argument-values [symbol]

Optional

Use the argument args followed by an argument value to initialize any pound-sign arguments in the loaded patcher (e.g., args #1). If used, the args argument must be the last argument word used; everything which appears after the word args will be treated as an argument value.

Attributes

Common Box Attributes

annotation [symbol]

Sets the text that will be displayed in the Clue window when the user moves the mouse over the object.

background [int] (default: 0)

Adds or removes the object from the patcher's background layer. background 1 adds the object to the background layer, background 0 removes it. Objects in the background layer are shown behind all objects in the default foreground layer.

color [4 floats]

Sets the color for the object box outline.

fontface [int]

Sets the type style used by the object. The options are:

plain
bold
italic
bold italic Possible values:

0 = 'regular'
1 = 'bold'
2 = 'italic'
3 = 'bold italic'

fontname [symbol]

Sets the object's font.

fontsize [float]

Sets the object's font size (in points). Possible values:

'8'
'9'
'10'
'11'
'12'
'13'
'14'
'16'
'18'
'20'
'24'
'30'
'36'
'48'
'64'
'72'

hidden [int] (default: 0)

Toggles whether an object is hidden when the patcher is locked.

hint [symbol]

Sets the text that will be displayed in as a pop-up hint when the user moves the mouse over the object in a locked patcher.

ignoreclick [int] (default: 0)

Toggles whether an object ignores mouse clicks in a locked patcher.

jspainterfile [symbol]

JS Painter File

patching_rect [4 floats] (default: 0. 0. 100. 0.)

Sets the position and size of the object in the patcher window.

position [2 floats]

g/s(set)

Sets the object's x and y position in both patching and presentation modes (if the object belongs to its patcher's presentation), leaving its size unchanged.

presentation [int] (default: 0)

Sets whether an object belongs to the patcher's presentation.

presentation_rect [4 floats] (default: 0. 0. 0. 0.)

Sets the x and y position and width and height of the object in the patcher's presentation, leaving its patching position unchanged.

rect [4 floats]

g/s(set)

Sets the x and y position and width and height of the object in both patching and presentation modes (if the object belongs to its patcher's presentation).

size [2 floats]

g/s(set)

Sets the object's width and height in both patching and presentation modes (if the object belongs to its patcher's presentation), leaving its position unchanged.

textcolor [4 floats]

Sets the color for the object's text in RGBA format.

textjustification [int]

Sets the justification for the object's text. Possible values:

0 = 'left'
1 = 'center'
2 = 'right'

varname [symbol]

Sets the patcher's scripting name, which can be used to address the object by name in pattr, scripting messages to thispatcher, and the js object.

Multichannel Group Attributes

chans [int]

The chans attribute sets the number of channels and instances in the MC wrapper object. To define a fixed number of channels regardless of what is connected to the object, set chans via a typed-in argument, for example typing mc.cycle~ @chans 100 would create 100 instances of a cycle~ object inside the MC wrapper. If chans is 0, the wrapper object will auto-adapt to the number of channels in its input multichannel signals (using the maximum of all connected signals). If an object does not have any multichannel signals connected to its inlets, the chans attribute will need to have a non-zero value if you want more than one instance.

If chans is changed while the audio is on, the number of instances will not change until audio is restarted. However, if chans is reduced while the audio is on, any extra channels will no longer process audio and will output a zero signal.

initialvalues [list]

The initialvalues attribute only applies to object creation time so it must be set via a typed-in argument. initialvalues sets the first (and only the first) initial argument for successive instances in the MC wrapper. For example, typing mc.cycle~ @chans 4 @initialvalues 50 60 70 80 would assign an initial frequency to the cycle~ instances inside the wrapper. The first instance would be assigned a frequency of 50, the second a frequency of 60, the third 70, and the fourth 80. Note that initialvalues does not determine the actual instance count; this can be done using the chans attribute. If there are more instances than elements for the initialvalues attribute, those instances are instantiated with the default value.

To set a default value of an argument for all instances, type it as an argument before any typed-in attributes. For example, modifying our example above: mc.cycle~ 100 @chans 10 @initialvalues 50 60 70 80. In this example, the first four instances are set as before, but the next six are created with a frequency argument of 100.

To change instance values or attributes after the wrapper object has been created, use the setvalue, applyvalues, or replicatevalues messages.

values [list]

You can use values as an alternate name for the initialvalues attribute.

target [int]

The target attribute sets an index for targeting specific wrapper instances. Subsequent messages are directed to an individual instance instead of all instances. It is strongly recommended you use the more reliable setvalue message instead of the target attribute. The voice index of setvalue will override the current setting of target. When target is 0, incoming messages are sent to all instances. When target is -1, incoming messages do nothing. Note that target only affects messages, not setting attribute values.

usebusymap [int]

When usebusymap is enabled, the MC wrapper controls whether individual instances process audio using a busy map maintained by either an mc.noteallocator~ or mc.voiceallocator~ object. When a channel in the busy map is marked as "free" or "released" no audio processing occurs by any instance on the channel corresponding to the voice index. When usebusymap is disabled, instances in the MC wrapper process audio at all times. This will also be true if usebusymap is enabled and there is no local or named busy map available. (See the busymapname attribute for a description of local and named busy maps). For brevity the name bz can also be used.

zero [int]

When the zero attribute is enabled, channels in the MC wrapper due to the use of a busy map output zero signals. To save a small amount of CPU at the risk of loud and unpleasant noises due to uncleared signal data, you can disable zero. In this case, disabled channels in the MC wrapper do nothing to their output channels. If usebusymap is disabled or there is no active local or named busy map available, the setting of the zero attribute has no effect.

Conveniently, when usebusymap is enabled in mc.mixdown~ object, disabled channels are not mixed to the output. When unused signals from wrapped objects with zero disabled feed into mc.mixdown~, they will be ignored, reducing the risk of unpleasantness getting past the mix output.

busymapname [symbol]

When the usebusymap attribute is enabled, an MC wrapper object uses the local busy map of any mc.voiceallocator~ or mc.noteallocator~ in the same patcher by default. To use a named global busy map instead, set the busymapname attribute to the desired name. For brevity the name @bzname can also be used.

op [symbol]

Sets the function that will be used when the generate message is set. If you use attrui set to edit the op attribute, you can see a handy menu of the 40+ possible functions, so you don't have to memorize their names.

voiceprob [float]

The voiceprob attribute is used when employing the $ or * arguments to the setvalue message. It determines the probability that the setvalue message will be sent. For example, if voiceprob is 0.9, there is a 90% chance the setvalue message will be sent to a randomly chosen voice.

Messages

bang

Patchers loaded into a pfft~ object can only accept bang messages by in objects within the patch. The number of inputs is determined by the in objects in the enclosed subpatch. See in for details.

int

Arguments

input [int]
Integer values sent to the pfft~ object cause the object to act according to the user-defined functionality within it.

float

Arguments

input [float]
Floating-point values sent to the pfft~ object cause the object to act according to the user-defined functionality within it.

list

Arguments

input [list]
Lists sent to the pfft~ object cause the object to act according to the user-defined functionality within it.

anything

Arguments

input [list]
Messages sent to the pfft~ object cause the object to act according to the user-defined functionality within it.

clear

Clears all of the pfft~ object's internal buffers.

(mouse)

Double-clicking with the mouse on the pfft~ object opens a Max patcher window containing the patcher loaded by the object.

mute

Arguments

mute-flag (0 or 1) [int]
The word mute, followed by a 1 or 0, will mute or unmute the pfft~, turning off signal processing within the enclosed subpatch.

open

Arguments

subpatch-filename [int]
The word open will open the subpatch loaded into the pfft~ object.

wclose

Arguments

subpatch-filename [int]
Closes the enclosed subpatch if it is open.

Multichannel Group Messages

deviate

Arguments

range [float]
message-name [symbol]
center-value [float]
upper-range [float]
Generate a random value for each channel around a center value. An optional number after the center value specifies the upper range size so it can be different from the lower range size.
Example: deviate 100 cutoff 1000 will generate random values for the cutoff attribute of the objects in the wrapper centered around 1000 Hz (between 900 and 1100 Hz). deviate 100 1000 200 sends float messages to the objects in the wrapper with random values between 900 and 1200.
If no message name is provided, a float message is used by default.

exponential

Arguments

exponent [float]
message-name [symbol]
multiplier [float]
The exponential message generates an exponential series. The first argument is N and the second (optional) argument is K in the following expression:
K * exp(-1 * N * channel) where channel starts at 0 for the first channel.
If the second argument is not present the default value is 1. Example: exponential 1 10 would generate, for four channels, values of 10, 3.678, 1.353, and 0.498. exponential -1 2 would generate 2, 5.437, 14.78, and 40.17.
If no message name is provided, a float message is used by default.

scaledexponential

Arguments

exponent [float]
message-name [symbol]
base [float]
The scaledexponential message generates an exponential series with the exponent scaled by the total number of channels. The first argument is N and the second (optional) argument is K in the following expression:
K * exp(-1 * N * (channel / num_channels) where channel starts at 0 for the first channel.
If the second argument is not present the default value is 1. Example: exponential -1 2 would generate, for six channels, values of 2, 2.363, 2.791, 3.297, 3.895, 4.602. scaledexponential -1 2 for four channels would generate 2, 2.568, 3.297, 4.324. scaledexponential provides a way to keep the range of the exponential series roughly the same independent of the number of channels.
If no message name is provided, a float message is used by default.

increment

Arguments

increment-amount [float]
message-name [symbol]
start-value [float]
The increment message generates a range of increasing values for each channel. The range starts at the second argument and increments each channel's value by the first argument. If no message name is provided then a float message is used by default.
Example: increment 5 2 for four channels would generate 2, 7, 12, and 17.
If no message name is provided, a float message is used by default.

harmonic

Arguments

multiplier [float]
message-name [symbol]
fundamental [float]
The harmonic message generates a harmonic series using the second argument as the fundamental frequency ( F ) and the first argument as a multiplier ( N ) in the following expression:
F * (1 + N * channel) where channel starts at 0 for the first channel.
Example: harmonic 1 440 for five channels would generate 440, 880, 1320, 1760, and 2200. harmonic 0.5 440 for four channels would generate 440, 660, 880, and 1100.
If no message name is provided, a float message is used by default.

subharmonic

Arguments

multiplier [float]
message-name [symbol]
fundamental [float]
The subharmonic message generates a subharmonic series using the second argument as the fundamental frequency ( F ) and the first argument as a multiplier ( N ) in the following expression:
F / (1 + N * channel) where channel starts at 0 for the first channel.
Example: subharmonic 1 440 for five channels would generate 440, 220, 146.7, and 110.
If no message name is provided, a float message is used by default.

spread

Arguments

boundary-value [float]
message-name [symbol]
other-boundary-value [float]
The spread message generates a range of values distributed to each channel. The first boundary value is included in the range outputs, but the second boundary value is not (see spreadinclusive, spreadexclusive, and spreadincludesecond for other options).
Example: spread 0 10 for four channels would generate 0, 2.5, 5, and 7.5.
If no message name is provided, a float message is used by default.

spreadinclusive

Arguments

boundary-value [float]
message-name [symbol]
other-boundary-value [float]
The spreadinclusive message generates a range of values distributed to each channel. Both the first and second boundary values are included in the range outputs.
Example: spreadinclusive 0 10 for four channels would generate 0, 3.33, 6.66, and 10.
If no message name is provided, a float message is used by default.

spreadexclusive

Arguments

boundary-value [float]
message-name [symbol]
other-boundary-value [float]
The spreadexclusive message generates a range of values distributed to each channel. Neither the first and second boundary values are included in the range outputs.
Example: spreadexclusive 0 10 for four channels would generate 2, 4, 6, and 8.
If no message name is provided, a float message is used by default.

spreadincludefirst

Arguments

boundary-value [float]
message-name [symbol]
other-boundary-value [float]
The spreadincludefirst message generates a range of values distributed to each channel. It is the same as the spread message. The first boundary value is included in the range outputs, but the second boundary value is not.
Example: spreadincludefirst 0 10 for four channels would generate 0, 2.5, 5, and 7.5.
If no message name is provided, a float message is used by default.

spreadincludesecond

Arguments

boundary-value [float]
message-name [symbol]
other-boundary-value [float]
The spreadincludefirst message generates a range of values distributed to each channel. It is the same as the spread message. The first boundary value is not included in the range outputs, but the second boundary value is included.
Example: spreadincludesecond 0 10 for four channels would generate 2.5, 5, 7.5, and 10.
If no message name is provided, a float message is used by default.

decide

Arguments

probability [float]
message-name [symbol]
value [float]
The decide message generates a uniformly distributed random value between 0 and 1 for each channel; if the value is less than the probability value set by the first argument, the second argument is assigned to the channel. If the random value is greater than the probability value, 0 is asigned to the channel. (If a second argument is not present, 1 is used by default.)
Example: decide 0 10 for four channels would generate 0, 0, 0, 0 because the probability of generating a 1 is zero. decide 0.5 10 could generate 10, 0, 0, 10 if the randomly generated values exceeded 0.5 for the first and fourth channels.
If no message name is provided, a float message is used by default.

randomrange

Arguments

low-value [float]
message-name [symbol]
high-value [float]
The randomrange message generates a uniformly distributed random range of values for all channels between the first argument and the second argument.
If no message name is provided, a float message is used by default.

generate

Arguments

low-value [float]
message-name [symbol]
high-value [float]
The generate message runs the function whose name is stored in the op attribute. Arguments passed to generate will be given to the function that is called. Example: if op is set to deviate, generate 50 440 is the same as sending the message deviate 50 440.

ease.linear

Arguments

low-value [float]
message-name [symbol]
high-value [float]
mid-point [float]
The MC wrapper provides access to the easing functions found in the Ease Package. These are accessed with message names consisting of ease. concatenated with the easing function name. For example, to use the in_out_circular function, send the message ease.in_out_circular.
The ease messages generate an non-linear and inclusive range of values across the space of channels. When you use two number arguments, the first value will be the low end of the range and the second will be the high end of the range. For in_ and in_out_ functions, this means the low end value will be set for the first channel and the high end will be set for the last channel. For out_ function variants, the high end will be set for the first channel and the low end will be set for the last channel.
When the ease messages are supplied with three numerical arguments, the first two specify the range as in the two-argument case, but the third argument, which will be constrained between 0 and 1, defines a mid point. Between the first channel and the channel closest to the mid point, the entire range of the function is applied. Between the mid point and the last channel, the range of the function is applied with the values reversed, creating a mirror image. The mirror image is exact when the third argument is 0.5, otherwise it will be biased toward 0 or 1. With a mid point of 1, the result is the same as if the third argument was not supplied at all. With a mid point of 0, the result is the same as if it was entirely reversed. In other words, it's as if the out_ version of the function were used instead of the in_ version that was originally specified -- or vice versa.
Available messages are: ease.linear, ease.in_back, ease.in_out_back, ease.out_back, ease.in_bounce, ease.in_out_bounce, ease.out_bounce, ease.in_circular, ease.in_out_circular, ease.out_circular, ease.in_cubic, ease.in_out_cubic, ease.out_cubic, ease.in_elastic, ease.in_out_elastic, ease.out_elastic, ease.in_exponential, ease.in_out_exponential, ease.out_exponential, ease.in_quadratic, ease.in_out_quadratic, ease.out_quadratic, ease.in_quartic, ease.in_out_quartic, ease.out_quartic, ease.in_quintic, ease.in_out_quintic, ease.out_quintic, ease.in_sine, ease.in_out_sine, and ease.out_sine. Refer to the Ease Package documentation for details on these functions and demonstrations of their behavior.
If no message name is provided, a float message is used by default.

smoothstep

Arguments

low-value [float]
message-name [symbol]
high-value [float]
mid-point [float]
The smoothstep function works analogously to the ease messages to generate an inclusive non-linear range of values, but uses the smoothstep function to generate a non-linear ramp. Refer to the documentation of the ease messages for more information.
If no message name is provided, a float message is used by default.

setvalue

Arguments

channel [int]
message [symbol]
message arguments [list]
The word setvalue, followed by both a channel index (starting at 1) and any message that can be sent to the wrapped object, sends the message to an individual instance within the MC wrapper. setvalue 0, followed by a message, sends the message to all instances. The setvalue message can be used in any inlet.
Instead of a number, the setvalue message can also take a symbol indicating that the target channel index should be randomly chosen:
  • setvalue * will choose a channel randomly but avoid duplicate choices until all channels have been chosen (similar to the Max urn object). Before chosing a channel, * will also decide whether to send the message according to the current value of the voiceprob attribute. If voiceprob is 0.1, there is a 10% chance of sending the message. If voiceprob is 0.9, there is a 90% chance of sending the message.
  • setvalue + will choose a channel randomly but avoid duplicate choices until all channels have been chosen (similar to the Max urn object). Unlike * it will always send the message.
  • setvalue $ will choose a channel randomly (similar to the Max random object). Before chosing a channel, $ will also decide whether to send the message according to the current value of the voiceprob attribute. If voiceprob is 0.1, there is a 10% chance of sending the message. If voiceprob is 0.9, there is a 90% chance of sending the message.
  • setvalue # will choose a channel randomly (similar to the Max random object). Unlike $ it will always send the message.

setvaluerange

Arguments

low channel [int]
high channel [int]
message [symbol]
message arguments [list]
The word setvaluerange, followed by a low and high channel index (starting at 1) and any message that can be sent to the wrapped object, sends the message to the specified range of channels.
Example: setvaluerange 1 4 50, sends the message 50 to channels 1 - 4. If the second argument is -1, the message is sent to all subsequent channels. For example, setvaluerange 2 -1 50 sends the message 50 to all channels between 2 and the current number of voices.
Note: the random channel selection feature using *, +, $, and # does not work with the setvaluerange message.

applymessages

Arguments

messages [list]
The word applymessages, followed by one or more numbers and/or symbols, sends individual messages successively to instances in the MC wrapper, starting with the first instance. For example, the message applymessages 0 bang will send the '0' message to the first instance, and the 'bang' message to the second instance. If there are more instances than arguments to applymessages, the extra instances are unaffected.

applyvalues

Arguments

message-name [symbol]
values [list]
The word applyvalues, followed by an optional message name and one or more message arguments, sends individual values in the arguments successively to instances in the MC wrapper, starting with the first instance. For example, the message applyvalues 4 5 6 will send 4 to the first instance, 5 to the second instance, and 6 to the third instance. If there are more instances than arguments to applyvalues, the extra instances are unaffected.

replicatevalues

Arguments

message-name [symbol]
values [list]
The word replicatevalues, followed by an optional message name and one or more message arguments, sends individual values in the arguments successively to instances in the MC wrapper, starting with the first instance. Unlike applyvalues, the replicatevalues message continues sending values to successive instances, restarting with the first element, if it runs out of arguments to send. For example, replicatevalues 4 5 to an MC wrapper object with three instances will send 4 to the first instance, 5 to the second instance, and 4 to the third instance.

applynvalues

Arguments

message [int]
values [list]
Whereas applyvalues can only set one value, the message applynvalues permits sending a message or setting an attribute with multiple values to instances in the MC wrapper, starting with the first instance. This is helpful for messages that require multiple values, such as the list message to wave~ to set start/end points. The message syntax is [applynvalues N value1, value2 etc.] where N is the number of values to set for each instance. For example, the message applynvalues 2 500 600 900 1000 will send 500 600 to the first instance and 900 1000 to the second instance. If there are more instances than specified in applynvalues, the extra instances are unaffected.

replicatenvalues

Arguments

message [int]
values [list]
Whereas replicatevalues can only set one value, the message replicatenvalues permits sending a message or setting an attribute with multiple values to instances in the MC wrapper, starting with the first instance. This is helpful for messages that require multiple values, such as the list message to wave~ to set start/end points. The message syntax is [replicatenvalues N value1, value2 etc.] where N is the number of values to set for each instance. Unlike applynvalues, the replicatenvalues message continues sending values to successive instances, restarting with the first group, if it runs out of arguments to send. For example, replicatenvalues 2 500 600 900 1000 to an MC wrapper object with three instances will send 500 600 to the first instance, 900 1000 to the second instance, and 500 600 to the third instance.

Output

message

Any messages received by an out object in a loaded patcher appear at the message outlet of the pfft~ object which corresponds to the number argument of the out object. The message outlets of a pfft~ object appear to the right of the rightmost signal outlet.

signal

The output is the result of the FFT-based signal processing subpatch. As with the fft~ and ifft~ objects, pfft~ introduces a slight delay from input to output (although it is less than half the delay than with an fft~ / ifft~ combination). The I/ O delay is equal to the window size minus the hop size (e.g., for a 1024-sample FFT window with an overlap factor of 4, the hop size is equal to 256, and the overall delay from input to output is 768 samples). The number of outlets is determined by the number of fftout~ and/or out objects in the loaded subpatcher. Patchers loaded into a pfft~ object can be given outlets by fftout~ or out objects within the patch. See fftout~ and out for details.

See Also

Name Description
cartopol Convert cartesian to polar coordinates
cartopol~ Signal Cartesian to Polar coordinate conversion
fft~ Fast Fourier transform
fftin~ Input for a patcher loaded by pfft~
fftinfo~ Report information about a patcher loaded by pfft~
fftout~ Output for a patcher loaded by pfft~
frameaccum~ Compute "running phase" of successive phase deviation frames
framedelta~ Compute phase deviation between successive FFT frames
ifft~ Inverse fast Fourier transform
in Message input for a patcher loaded by poly~ or pfft~
out Message output for a patcher loaded by poly~ or pfft~
poltocar Convert polar to cartesian coordinates
poltocar~ Signal Polar to Cartesian coordinate conversion
vectral~ Vector-based envelope follower
MSP Analysis Tutorial 3: Using the FFT MSP Analysis Tutorial 3: Using the FFT
MSP Analysis Tutorial 4: Signal Processing with pfft~ MSP Analysis Tutorial 4: Signal Processing with pfft~