Datasets:
kkboy1
/
LeAudio

Dataset card Data Studio Files
xet
 Community
1
text
stringlengths
1
112
code x x x x
value x x xcc xcc
supported max codec frames per sdu optional default
length x x x x
code x x x x
value x x x x
table example of pac records for mandatory khz sampling at ms with or audio channels
as an example of how they can be formed table contains t he sink pac record parameters
for the mandated codec requirements for all supported bluetooth le audio profiles ie the
and qos configurations from table and table of
bap to cover those
the supportedsamplingfrequencies cha racteristic has the values x for khz
and x for khz
the supportedframedurations is x to signify it only supports ms frames
the supported audio channel counts value s are x and x to support one
and two channels
the supported octetspercodec frame has values of x and xcc for
and octets corresponding to the n and n qos settings and finally
the supported maxcodec frames persdu indicate s that there is only one
codec frame per sdu with a value of x this could have been omitted as it is
the default value in this sink pac characteristic there is no metadata
chapter setting up unicast audio streams
this same information could be provided more compactly in the single pac rec ord of table
pac record record ltv
codecid lc xd
codecspecificcapabilitieslength xb
codecspecificcapabilities
supported sampling frequencies x
supported frame durations x
supported audio channel counts x
supported octets per frame x c
supported max codec frames per sdu x could be omitted as default
table the pac record content of table as a single record
these two representations are not completely equivalent pacs states that a n acceptor must
support all possible combinations of a parameter value with all other possible combinations
of a parameter value stated in a pac record that means that where multiple values are
included in a pac record as in the example of table then every possible combination is
valid as the initiator can expand the pac record into an array of all possible combinat ions
in theory that means that an acceptor exposing the single pac record of table should
if requested support a value of octets for khz sampling and octets for khz
sampling as well as any octet value between and although these are non standard for
the sake of interoperability there is an assumption that an initiator should confine itself to
using the specific configurations from tables and of bap which are tried and tested
but an initiator should choose another combination thats not good practice but it is
allowed if not all of those combinations are valid in a device then the pac records should
be expressed as individual pac records however that may result in the number of records
multiplying rapidly which is not a good thing the example above which is the baseline
support need ed for bap results in four pa c records which is manageable tr ying th e do
the same thing for tmap would require at least ninety six individual pac records as we will
see shortly the ase configuration process can guide the preferred options for configuring an
isochronous stream so individual pac records can be reserved for applications su ch as
vendor specific codec capabilities more details are provided in section of pacs
if an acceptor supports more than one codec type at least one sink or source pac
characteristic will be required for each as the codecid field of a pac record is unique it is
up to the implementation to decide whether all pac records for a codec are e xposed in a
single or multiple sink pac or source pac characteristics if there is a large number of pac
records an implementation may want to split them into separate pac record characteristics
to prevent the maximum att mtu size being exceeded when rea ding them
sink pac characteristics do not distinguish between unicast and broadcast streams so the
section pacs the published audio capabilities service
values apply to both source pac characteristics are ignored for broadcast
audio locations
both sink pac and source pac chara cteristics can have an associated sink audio locations
and source audio locations characteristic respectively although these are optional the
basic concept of audio locations was described in chapter the sinkaudiolocations
and sourceaudiolocatio ns characteristics specify which audio locations are supported by
the acceptor for received and transmitted audio streams each characteristic has a four octet
field which is a bitfield indicating which rendering or capturing locations it supports
if they are present at least one bit must always be set if theyre not present the device should
accept any audio l ocation proposed by an initiator in many cases the values are set by the
manufacturer and are read only a right earbud will only ever be a right earbud so has a sin gle
sink and or source audio location of front right a speaker is likely to have its
sinkaudio locations characteristic set to both front left and front right when an
initiator wants to establish a stream it will decide whi ch audio stream to send to match the
audio location if it finds two speakers it would normally send a left stereo stream to the
one exposing the front left location and a right stereo stream to the one with the front right
location if they both say th ey can support both front left and front right which most will
do the user would normally use an application to configure which is which the sink and
source audio locations characteristics may be writeable which would allow a configuration
utility to set speaker locations for use by other initiators because these are all interoperable
features in the bluetooth le audio specification it means that any compliant audio
application should be able to perform this sort of configuration alternatively a s peaker
manufacturer could provide a physical switch on its speakers to allow a user to specify whether
a speaker receives a left or right stream
in chapter we looked at how initiators and acceptors would deal with downmixing stereo
streams for mono r eproduction showing that this could be done either before encoding the
audio stream or after reception and decoding an acceptor can use its pac records to
indicate to an initiator whether or not it is capable of downmixing if it sets its audio
locati ons for front left and front right but only supports a sin gle bl uetooth le audio
channel then it implies that it is not capable of downmixing but will only render the stream
it has been supplied if it shows support for two channels the implication is that it can decode
them into separate streams to be rendered it is then down to the acceptor implementation
to render them as
separate left and right audio channels which it would do if it were a sound bar or
stereo speaker
a single one of the two audio channels it decodes which it might do if it were an
earbud or hearing aid or
downmix them to a single mono audio channel if it contained a single speaker
chapter setting up unicast audio streams