Please consider my proposal for array extension of the LADSPA:
---- cut ----
/* This flag indicates that a port is an array of LADSPA_Data. The size
of the array, which is the number of LADSPA_Data elements in the array,
is stored in the index [-1] of the passing pointer. Further implication
on how to interpret the data is provided in the hints. */
#define LADSPA_PORT_ARRAY 0x0100
/* The port is an array of LADSPA_Data representing an envelope in which
each value is a parameter that directly applies to a sample or a set of
samples. The size of this array is indicated in the index [-1] of the
passing pointer. Using LADSPA_HINT_ARRAY_ENVELOPE implies a size such
that ArraySize * n = SampleCount, where n is a positive integer. When
LADSPA_HINT_ARRAY_ENVELOPE_INTERPOLATE is also specified and when an
envelope applies to more than one samples, the plugin should do
interpolation of these envelopes whenever possible. It is up to the
plugin whether linear or second-order spine interpolation is to be
performed.
Note: since the granule of an envelope defined here is really
microscopic, it is only useful when the host calculates different sets
of these micro-envelopes for each input buffer, so the plugin can be
properly instructed in the right scale of time. */
#define LADSPA_HINT_ARRAY_ENVELOPE 0x0100
#define LADSPA_HINT_ARRAY_ENVELOPE_INTERPOLATE 0x0200
/* The port is an array of interleaved pairs of LADSPA_Data. The number
of pairs is ArraySize / 2, so ArraySize has to be a multiple of 2. The
array is then laid out as the following ordered set: { key0, data0,
key1, data1, key2, data2, ... }. One useful example for this data
structure is an equalizer parameter. */
#define LADSPA_HINT_ARRAY_PAIRS 0x0400
/* All the above array-exended hints can and should be used along with
the other LADSPA_HINT_* flags. However, for LADSPA_HINT_ARRAY_PAIRS, it
is defined that all original LADSPA hints apply only to the data. To
make hints for the key, use the following hint flags instead. These
flags are just the LADSPA_HINT_* flags shifted left by 16 bits. */
/* Note: this will void ladspa on 16-bit architectures. Do we want to do
this? */
#define LADSPA_HINT_KEY_BOUNDED_BELOW 0x010000
#define LADSPA_HINT_KEY_BOUNDED_ABOVE 0x020000
#define LADSPA_HINT_KEY_TOGGLED 0x040000
#define LADSPA_HINT_KEY_SAMPLE_RATE 0x080000
#define LADSPA_HINT_KEY_LOGARITHMIC 0x100000
#define LADSPA_HINT_KEY_INTEGER 0x200000
/* When one makes use of the LADSPA_HINT_KEY_BOUNDED_BELOW or
LADSPA_HINT_KEY_BOUNDED_ABOVE, you need to use an extended
LADSPA_PortRangeHint structure defined as the following:
typedef struct _LADSPA_PortRangeHint {
LADSPA_PortRangeHintDescriptor HintDescriptor;
LADSPA_Data LowerBound;
LADSPA_Data UpperBound;
/* This is the extended part, note this should preserve binary compatibility. */
LADSPA_Data KeyLowerBound;
LADSPA_Data KeyUpperBound;
} LADSPA_PortRangeHint;
---- uncut ----
This is a morning of work, so please proof read for me if you find any
errors.
Likai Liu
