foscil and foscili implement composite units for FM in the Chowning setup.
crossfm, crossfmi, crosspm, crosspmi, crossfmpm and crossfmpmi are different units for frequency and/or phase modulation.
distort and distort1 perform waveshaping by a function table (distort) or by modified hyperbolic tangent distortion (distort1).
powershape waveshapes a signal by raising it to a variable exponent.
polynomial efficiently evaluates a polynomial of arbitrary order.
chebyshevpoly efficiently evaluates the sum of Chebyshev polynomials of arbitrary order.
flanger implements a user controllable flanger.
harmon analyzes an audio input and generates harmonizing voices in synchrony.
phaser1 and phaser2 implement first- or second-order allpass filters arranged in a series.
doppler lets you calculate the doppler shift depending on the position of the sound source and the microphone.
partikkel is the most flexible opcode for granular synthesis. You should be able to do everything you like in this field. The only drawback is the large number of input arguments, so you may want to use other opcodes for certain purposes.
You can find a list of other relevant opcodes here.
sndwarp focusses granular synthesis on time stretching and/or pitch modifications. Compare waveset and the pvs-opcodes pvsfread, pvsdiskin, pvscale, pvshift for other implementations of time and/or pitch modifications.
pconvolve performs convolution based on a uniformly partitioned overlap-save algorithm.
ftconv is similar to pconvolve, but you can also use parts of the impulse response file, instead of reading the whole file.
pvsanal performs a Fast Fourier Transformation of an audio stream (a-signal) and stores the result in an f-variable.
pvstanal creates an f-signal directly from a sound file which is stored in a function table (usually via GEN01).
pvsynth performs an Inverse FFT (takes a f-signal and returns an audio-signal).
pvsadsyn is similar to pvsynth, but resynthesizes with a bank of oscillators, instead of direct IFFT.
pvsfwrite writes an f-signal (= the FFT data) from inside Csound to a file. This file has the PVOCEX format and its name ends on .pvx.
pvanal does actually the same as Csound Utility (a seperate program which can be called in QuteCsound or via the Terminal). In this case, the input is an audio file.
pvsfread reads the FFT data from an extisting .pvx file. This file can be generated by the Csound Utility pvanal. Reading the file is done by a time pointer.
pvsdiskin is similar to pvsfread, but reading is done by a speed argument.
pvsbuffer writes a f-signal to a circular buffer (and creates it).
pvsbufread reads a f-signal from a buffer which was created by pvsbuffer.
pvsftw writes amplitude and/or frequency data from a f-signal to a function table.
pvsftr transforms amplitude and/or frequency data from a function table to a f-signal.
pvsinfo gets info either from a realtime f-signal or from a .pvx file.
pvsbin gets the amplitude and frequency values from a single bin of a f-signal.
pvscent calculates the spectral centroid of a signal.
pvscale transposes the frequency components of a f-stream by simple multiplication.
pvshift changes the frequency components of a f-stream by adding a shift value, starting at a certain bin.
pvsbandp and pvsbandr applies a band pass and band reject filter to the frequency components of a f-signal.
pvsmix, pvscross, pvsfilter, pvsvoc and pvsmorph perform different methods of cross synthesis between two f-signals.
pvsfreeze freezes the amplitude and/or frequency of a f-signal according to a k-rate trigger.
pvsmaska, pvsblur, pvstencil, pvsarp, pvsmooth perform other manipulations on a stream of FFT data.