1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
You can get more inofrmation on the expr object at
http://www.crca.ucsd.edu/~yadegari/expr.html
-----------
New in Version 0.3
-Full function functionality
------------
The object "expr" is used for expression evaluaion of control data.
Expr~ and fexpr~ are extentions to the expr object to work with vectors.
The expr~ object is designed to efficiently combine signal and control
stream processing by vector operations on the basis of the block size of
the environment.
fexpr~ object provides a flexible mechanism for building FIR and
IIR filters by evaluating expressions on a sample by sample basis
and providing access to prior samples of the input and output audio
streams. When fractional offset is used, fexpr~ uses linear interpolation
to determine the value of the indexed sample. fexpr~ evaluates the
expression for every single sample and at every evaluation previous
samples (limited by the audio vector size) can be accessed. $x is used to
denote a singnal input whose samples we would like to access. The syntax
is $x followed by the inlet number and indexed by brackets, for example
$x1[-1] specifies the previous sample of the first inlet. Therefore,
if we are to build a simple filter which replaces every sample by
the average of that sample and its previous one, we would use "fexpr~
($x1[0]+$x1[-1])/2 ". For ease of when the brackets are omitted, the
current sample is implied, so we can right the previous filter expression
as follows: " fexpr~ ($x1+$x1[-1])/2". To build IIR filters $y is used
to access the previous samples of the output stream.
The three objects expr, expr~, and fexpr~ are implemented in the same object
so the files expr~.pd_linux and fexpr~.pd_linux are links to expr.pd_linux
This release has been compiled and tested on Linux 6.0.
--------
Here are some syntax information: (refer to help-expr.pd for examples)
Syntyax:
The syntax is very close to how expression are written in
C. Variables are specified as follows where the '#' stands
for the inlet number:
$i#: integer input variable
$f#: float input variable
$s#: symbol input variable
Used for expr~ only:
$v#: signal (vector) input (vector by vector evaluation)
Used for fexpr~ only:
$x#[n]: the sample from inlet # indexed by n, where n has to
satisfy 0 => n >= -vector size,
($x# is a shorthand for $x#[0], specifying the current sample)
$y[n]: the output value indexed by n, where n has to
satisfy 0 > n >= -vector size,
I'll appreciate hearing about bugs, comments, suggestions, ...
Shahrokh Yadegari (sdy@ucsd.edu)
1/29/02
|
