Dialogic 6.2 - Manual (Page 51)

The Bfv API Functions
November 2009 51
Infopkt Structure
Every infopkt consists of a header and data. The 4-byte header
consists of a type code and a length. The type code defines the
infopkt type, and the length field indicates the total length of the
infopkt, including the header. Data consists of raw data, formatting
parameters or, in the case of indirect infopkts, pointers to other files.
The maximum length of an individual infopkt is 30,000 bytes, but
Dialogic recommends limiting the size to approximately 1K. This
limit affects the size of infopkts only and has no effect on the size of a
fax document, since large fax documents are simply converted to
multiple infopkts within an infopkt stream.
When indirect infopkts point to other infopkt stream files, the
maximum nesting depth is three. All file names that occur in
indirect infopkts must be 0-terminated.
The infopkt.h file, located in the inc subdirectory, contains the
definitions of the infopkt structures.
See Appendix E in Volume 6 of the Bfv API Reference Manual for the
parameter values and defaults of each tag type infopkt.
Speech Infopkt Parameters
The INFOPKT_SPEECH_PARAMETERS infopkt defines the speech
parameters for infopkt-formatted speech files. These parameters
include the coding format, the compressed sample rate, the number
of bits per sample, the analog front end (afe) rate, and the data
format. In infopkt streams made up of simple speech files, this
infopkt begins the infopkt stream. In infopkt streams made up of
indexed prompt files, it begins each new speech file embedded in the
stream.
Applications use the BfvSpeechRecord function to record speech
in infopkt format (see the Bfv API Reference Manual, Volume 3 for a
complete description). Valid settings for the data coding format,
compressed sample rate for playback, and number of bits per sample
are found with the BfvSpeechRecord function.