![]() |
|
![]() |
INTRODUCTION
The CDP suite is bewilderingly large, and it is hard to know where to begin. The simple answer is "anywhere", but that is not very helpful. It all depends on what you might want to do with your source sound, or what sort of sound you want to achieve from which you need to work out how to get there (not always so easy!). Much of the CDP suite is the work of the English electro-acoustic composer Trevor Wishart and reflects his priorities, especially in regard to vocal or speech processing and segmentation; nevertheless pretty well every possible technique of sound transformation is there somewhere.
The main division in CDP is between Time-domain processes and Spectral processes, operating in the Frequency-domain, and Soundshaper preserves this distinction in its Process Menus. This section is a brief overview of CDP processes in the context of Soundshaper's menus.
Time-domain processes operate on a sound waveform: a typical source might be a 16-bit .wav file, but other types can be used. Soundshaper splits these into two menus: the Edit-Mix Menu and the Soundfiles Menu.
(Soundfile types: CDP supports 16 and 24-bit integer and 32-bit floating point soundfiles, with a wide variety of sampling rates ranging from 8KHz to 96KHz. In addition, ambient and WAVE-EX format soundfiles are supported. Both wav and aif file types are supported, but not mp3, flac or Ogg Vorbis.)
EDIT-MIX PROCESSES
The Edit-Mix Menu has the fundamental sound-editing operations that you find in visual editors, such as basic editing (cutting and pasting), spatial distribution and channel operations, level and balance, and mixing. In the classic analogue studio these functions either involved cutting and splicing tape, or operating a mixing desk. The Edit-Mix menu also has a growing number of functions that extend these principles into relatively uncharted areas, such as multi-channel spatial distribution.
Soundshaper does not currently have a built-in editor, but links to any of several free editors and can import cue points from these (see Markers).
The Edit-Mix Menu has the following divisions:
- Edit Cut and paste sounds, including top-and-tail based on level, cut at zero; make multiple cuts to different files, extracted randomly or by gate level; insert silences and join files; also mask portions with silence and switch between soundfiles.
For a simple cut, insert the time as a marker in the Markers edit box and click the CUT button. You can also process a portion of a sound and paste that back into the original (see Edit at Markers). All edits in CDP/Soundshaper are non-destructive, i.e. they do not actually overwrite the original file.
Note also Soundshaper's separate EDIT button, which brings up a menu with a selection of the main editing and mixing functions.
- Channels Split a sound into its separate channels (or selected ones); interleave files to make a stereo or multi-channel file; convert between mono and stereo, or mono/stereo and multi-channel; manipulate channels in multi-channel space: rotate, renumber, etc.
- Spatial Pan sound across stereo or multi-channel space, including ambisonics. Pan values can simulate space beyond full left or right, by causing the sound to be attenuated. The multi-channel functions either distribute a mono sound across multi-channel space, or re-order the spatial distribution within an m-c file (e.g. by changing channel order or rotation).
- Mix CDP mix functions are best for mixing small sounds to create larger units; although larger mixes are perfectly possible, a visual editor is more intuitive and tends to have better controls. There are CDP functions to mix two files and multiple files, to crossfade between two sounds, and to generate a transition set of mixes between sounds.
The main process, however, is Mix, for which Soundshaper has a dedicated Mix Page. Mix uses a mixfile, a text data file of soundfile names, times, level and pan position. This is a very flexible format, with the exception of having fixed pan positions (to vary pan you have to apply pan to each sound before putting it in the mix). There is also a multi-channel mixfile format, in which pan is replaced by channel routing.
- Mix Files On Soundshaper's Mix Page, you can create and edit mixfiles, or re-call and modify previously saved mixfiles. You can also run a number of CDP functions for manipulating mixfile data; these can also be accessed on the Mix page.
- Level Various functions for adjusting gain, inverting phase and gating sounds (zeroing what falls below a given level). See also Envelope functions which do the same sort of thing. Soundshaper has a number of preset gain and normalising menu-items, and a separate menu for these, accessed via the LEVEL button:
.
- Sequence CDP has a Sequencer function (called Sequence2 in CDP docs.) which can play multiple sound sources using a file of: Sound no., time, pitch (MIDI values), level and duration. This is close to a standard MIDI file, and Soundshaper has a MIDI to SEQ utility in the Tools menu. Three other Sequence functions use less detailed data files.
- Cleanup & Utils Finally some utilities, most of which are never needed. A notable exception is CopySFX, which Soundshaper uses to convert sources to temporary files before processing. This can convert between a large number of soundfile formats, including Microsoft WAVEX and ambisonic files.
~~~~~~~~ For more detailed information on EDIT-MIX functions , see: EDIT-MIX MENU. ~~~~~~~~
SOUNDFILE PROCESSES
The Soundfiles Menu deals with all sound-processing functions which are not in the edit-mix category. The envelope and filter categories will be familiar to synth players, while pitch functions are mostly speed-changing operations, copying the bahaviour of tape recorders and record decks, in which speed-change meant pitch change).
The other categories all involve making copies of (parts of) sounds, or segmentation, including finding and exploiting suitable divisions, as in the grain, rhythm and distort menus.
Soundshaper's Soundfiles Menu has the following divisions:
~~~~~~~~ For more detailed information on SOUNDFILE functions , click here: SOUNDFILES MENU . ~~~~~~~~
- Envelope Here you can alter the overall loudness contour of the sound: create fades and swells, warp the envelope in many ways, or totally re-draw it, create new envelopes, and impose or replace the envelope of one sound with that of another.
- Filter Filtering changes the tone-colour of the sound; if this a new idea, try singing the word "miaow": as your mouth changes shape, you are filtering out some of the harmonics in the sound and emphasizing others. CDP has all the classic filter types and has particularly effective filter banks (c.f. Graphic EQs), which can be used to "tune" the sound by strongly emphasising certain frequencies.
- Pitch Most processes here change the speed of the sound, which also changes its pitch (and time). Quite different are the two modulation processes RingMod and CrossMod, which can radically change the frequency content; Degrade is in a class of its own.
- Reverb/Echo Reverberation is a matter of echoes within an enclosed space; CDP's Reverb and Roomverb implement classic reverb. algorithms, while convolution uses impulse files or soundfiles to mimic any acoustic space. Delay lines similarly produce echoes and there is a wide range of functions covering every aspect of delay and echo. Loop is particularly important as an introduction to delayed echoes as a process of segmentation.
- Extend/Segment The first of several submenus dealing with segmentation. Here the sound is chopped up or re-ordered in many ways, ranging from simple Reverse to sophisticated zig-zagging through the file.
- Grain The GRAIN process (CDP: Brassage / Sausage) extends the principles in Loop to create either thickly textured sounds out of overlayed segments, or separately segmented sounds. The rest of the Grain suite processes "grainy sounds" - i.e. sounds with silences separating the sections / grains. Any sound with gaps (of minimum 32msecs) can be regarded as a "grainy sound" and the function Granulate can be used to create such sounds. Grains can be treated in many ways, e.g. duplicated, omitted, repitched, reversed in order.
- Pitch-Sync Grains CDP's PSOW suite is rather experimental and works best with vocal sounds, which have strong formants. It segments the sound according to grains it finds that might correspond to "FOFs", and then processes these grains. FOFs (from French 'Forme d'Onde Formantique') are formant waveforms - grains - used to synthesise the singing voice.
- Rhythm The Retime program deals with the retiming of amplitude peaks or silence-separated events. It promises a lot, but can be hard to operate effectively. As a producer of experimental retimings of events, it has potential, but is less successful in emulating commercial "beat-slicing" functions.
- Texture Texture processes repeat the input sound(s) in various ways to create a texture of events ('notes'). 'Notes' can be treated as simple repetitions, or repeated in groups, or with a timed rhythm, or as transposed ornaments or fully-defined motifs (timed and transposed). In each of the eight main functions, repetitions can be pitched at random within a defined pitch range, or restricted to a user-defined pitch-set or a "harmonic-field", which octave-transposes the pitches; this set/field can be time-varying.
- Distort The Distort suite has all sorts of ways of creating "grungy" sounds out of nice ones. It locates so-called "wavesets", or pseudo-wavecycles, based on zero-crossovers, and processes these by repetition, omission, replacing waveforms, adding "harmonics", and so on. You can process cycles individually or in groups: the longer the groups, the more you are processing recognisable segments and the less "grunge".
SPECTRAL PROCESSES
The spectral processes are perhaps the most distinctive group of CDP functions, offering many original approaches not found elsewhere.
Spectral processes operate on frequency analysis files (suffix .ana), converted using the public-domain Phase Vocoder (PVOC), which analyses the frequency spectrum using the Fast Fourier Transform, or FFT, at successive time intervals. These successive analyses are not "instant-time" snapshots of the spectrum, but use a short time-window, with the windows usually overlapping (in time) to some degree. The frequency information is analysed using fixed frequency bands. (For further details, see the Introduction to Spectral Processing and, for even greater detail, Richard Dobson's article The Operation of the Phase Vocoder in the CDP documentation.)
Note that in Soundshaper a soundfile can be used as the input to a spectral process, as converts soundfiles into .ana files automatically, using Auto-conversion.
Soundshaper's Spectral Menu has the following divisions:
- Emphasize and Reshape Both groups of processes alter the spectral envelope: that is, the distribution of amplitude across the frequency spectrum. As this is time-varying across successive time-slices (windows), it also incorporates the changing amplitude envelope as it evolves in each frequency band. The functions either focus on certain aspects of the spectral envelope or reshape it. Typical examples are Focus, which concentrates the energy around spectral peaks, Accumulate, which sustains frequency bands, and Average, which averages the energy across adjacent frequency bands.
- Filter These functions filter the sound by eliminating certain frequency bands and preserving others; they can either mimic traditional filtering or can thin the sound (e.g. Trace), or otherwise make selections according to frequency bands (e.g. Slice, which partitions it into separate files).
- Freq./Pitch These processes manipulate the frequency information directly, especially without altering the timebase. Transpose is the most obvious application (c.f. Speed in Soundfiles | Pitch menu); others stretch or shift the frequency information, while Pick and Tune tune it to a specified pitch template.
- Morph/Formants Morphing makes transitions between sounds, by interpolating amplitude and/ or frequency values. Vocoding (Vocode & Cross) creates hybrid sounds, by implanting the spectral envelopes of one sound onto another. The formant functions are concerned with the spectral envelope extracted as a separate "formant" file type (.for), which is used in the PITCH suite see below. The link is that PutFmts is a direct equivalent of vocoding, in which a spectral envelope is replaced or imposed, using an extracted formant file (.for).
- Combine These processes operate on two or more spectra: interleaving them, taking the band with the maximum amplitude, producing the sum or difference of their amplitudes, or the mean of their amplitudes or frequencies.
- Time Although in the spectral domain, these processes are all concerned with the evolution of sounds in time. They change the information across successive time-windows, e.g. by blurring it, freezing it, stretching sounds in time (preserving pitch), or selecting a different succession of windows (Drunk, Shuffle, Weave).
- Utils Finally, some utility programs, including Cut, Clean (orig. version -- new one under FILTER), Gate (as in Edit-Mix > Level, but done spectrally).
~~~~~~~~ For more detailed information on SPECTRAL functions , click here: SPECTRAL MENU . ~~~~~~~~
PITCH PROCESSES
CDP's Pitch processes are an offshoot of spectral analysis, but are put in a separate menu in Soundshaper, for convenience.
The pitch function GetPitch extracts time-varying pitch from a frequency analysis file (.ana), if the sound is pitched. It produces either a binary pitch file (.frq), or a breakpoint data file (.pch or .brk) of time - pitch (in MIDI values). There is no direct reverse function: to re-create an analysis file, the pitch file must be combined with a formant file (.for) & see Morph/Formants above, using the function MAKE. Typically, this would come from a different sound, but it could well come from the same sound if you have manipulated its extracted pitch.
Pitch files can be processed in various ways, and synthesized or manipulated directly (as breakpoint data), but commonsense has to prevail when re-synthesising with MAKE, as the formant files also contain much detailed frequency-related information. Only where there is a match between the energy in a particular frequency band and the specified pitch information, will there be a useful output.
Soundshaper's Pitch Menu has the following divisions:
- Get Pitch Extract pitch to a binary file (.frq), a breakpoint text data file (.pch or .brk), or set default values to be used where Soundshaper automatically extracts pitch.
- at root level: a variety of functions for generating a spectrum or synthesizing a pitch file, combining with formant data (MAKE), or imposing vowel shapes. Soundshaper also adds its own conversion of pitch-breakpoint data to transposition values as ratios, so that a pitch contour extracted from one sound can be imposed on another. (CDP can extract a transposition file (.trn), but seems to view the task differently.)
- Process A diverse group of functions operating directly on pitch data (exaggerating, quantizing, approximating etc.); also some functions relating to pitch vs. unpitched information in the pitch file (in speech: voiced/unvoiced). To an extent, the QUANTIZE function can be used to mimic the pitch-correction facilities of some expensive commercial applications.
- Convert and Combine Almost every conceivable combination of conversions and combinations between different forms, including the creation and conversion of (yet another) binary file type: transposition files (.trn). The most important functions are two converting Pitch files (.frq) to Breakpoint (.pch/.brk).
~~~~~~~~ For more detailed information on PITCH functions , click here: PITCH MENU. ~~~~~~~~
SYNTH & DATA FUNCTIONS
- Synth Menu Only a small number of synthesis functions are provided within CDP, which is predominantly a sound-processing system. You can create one of a number of traditional synthetic waveshapes, a chord, white noise, silence, or a spectral band around a given frequency. CDP7 added a powerful additive-synthesis program. Note that in Soundshaper, an output filename must be given. The output is sent to the Patchgrid as a new source.
In CDP, the traditional division between synthesis and musique concrète is heavily blurred, as the processing of synthetic sources produces, in effect, new synthetic sounds.
- Data Menu A number of functions generate data files for various CDP processes, such as filtering, reverb and pan.
- Data Edit Soundshaper's Data Editor page handles the CDP COLUMNS program, which has around 100 functions for manipulating ot generating data files.
~~~~~~~~ For more detailed information on Synth and Data functions , click here: SYNTH & DATA MENUS. ~~~~~~~~
INFORMATION FUNCTIONS
The Info Menu has a fairly large number of functions which either report to Soundshaper's Report Window, or produce an output text file. In each case, the output is not a file which can be processed further, so is not sent to the Patchgrid. If an input sound or analysis file is needed as input, it is assumed to be the file on the current cell. Auto-conversion of input file types is not fully supported for Info functions.
Soundshaper's Info Menu has the following divisions:
~~~~~~~~ For more detailed information on INFO functions , click here: INFO MENU . ~~~~~~~~
- Soundfiles Functions which give information about a soundfile. The general functions FileInfo and FileProps can report on most CDP file types. Some functions are specific to certain processes, notably GRAIN and DISTORT. Grain Count or Grain Assess are particularly vital for use in the Grain functions.
- Spectral Functions which give information about frequency analysis files (.ana). Relating channels (frequency bands) to frequency (and vice versa) is particularly important, as is the Window count. Several functions look for peaks in the spectral data.
- Pitch A few functions give information about pitch files (.frq), extracted from spectral files (.ana).
RETURN to top of page |