text
stringlengths
1
112
the user and then render the audio to a pair of wired headphones it could equally transmit
the streams using bluetooth either bluetooth classic audio or bluetooth le audio although
that is unlikely to be an efficient use of airtime
chapter isochronous streams
figure using a broadcast assistant and past
figure shows what is expected to be the most common use case where a device like a
phone watch or remote control acts as a broadcast assistant to scan and present the choice
of broadcast audio streams to the user then use s past to allow the us ers headphone or
earbuds to connect to the selected stream as we will see the broadcast assistant can also be
used for volume control and mute allowing a user to find select and control the rendering of
a broadcast audio stream it is illustrated in figure
although there is no connection between a broadcast source and a broadcast sink for the
audio stream devices can use acl connections to help synchronise to the bis so that a tv
could automatically synchronise to an audio stream when you enter a room in this case the
broadcast source would normally also contain a broadcast assistant which would be paired
to a users headset when the user chooses to connect to the broadcast assistant it would
use past to provide details of how to synchronise to the b ises and bass to transfer the
broadcastcode
figure collocation of a broadcast source with a broadcast assistant
this is one of the topolog ies that form s the basis of audio sharing where a number of friends
can share music from one phone chapter explains these options in greater detail
section isoal the isochronous adaptation layer
isoal the isochronous adaptation layer
isoal core vol part g is one of the most complicated aspects of the core isochronous
stream s feature the good news is that its taken care of for you in the bluetooth chips but a
knowledge of why its necessary and what it does is useful isoal is used for both broadcast
and unicast audio streams
isoal exists to solve the problem of what happens when there is a mismatch between frame
sizes the most common reason for this is that an acceptor only supports a ms frame size
but an initiator needs to use ms frames as it is running connections with other bluetooth
classic devices such as older mice and keyboards which are only capable of running at a
ms timing interval that will change in time as new peripheral devices become more
flexible but until then isoal provides a means to accommodate them while the whole
bluetooth ecosystem migrates to a ms timing interval
the isoal layer sits in the data path between the codec and the link layer encoded sdus
from the codec may be delivered via the hci if the codec is implemented in the host or
through a proprietary interface regardless of where the codec is situated as figure
shows isoal provides fragmentation and recombination or segmentation and reassembly
and is responsible for sending the encoded audio data in either framed or unframed pdus
figure architecture of the isochronous adaptation layer isoal
classic bluetooth i s based on a ms interval but many applications standardised on ms which
causes this problem for dual mode devices
chapter isochronous streams
depending on the setting provided by the host layer the controller can decide whether to
use framed or unframed pdus for unframed pdus if an sdu can fit within a single pdu
the isochronous ada ptation manager may fragment them but sends them without a
segmentation header thats the most efficient lowest latency way to transport isochronous
data if the sdu is larger than the maximum pdu size it will be segmented and sent in
multiple unfram ed sdus the recombination or reassembly process es reassemble the m into
sdus unframed pdus can only be used when the isointerval is equal to or is an integer
multiple of the sduinterval which is itself equal to or an integer multiple of the sampling
frame this means that generation of the sdus need to be synchronised with the transport
timing so that they dont drift with respect to each other otherwise you need to use framed
sdus
for framed sdus the isochronous adaptation manager adds a segm entation header and an
optional timeoffset the timeoffset allows multiple sdus to be segmented across
pdus providing a reference time that maintains an association between the sdu generation
and transport timing if you need more detail the full spe cification of isoal is contained in
vol part g section of the core
the main aspect of isoal that designers need to be aware of i s that it contains the set of
equations which define the transportdelay transportdelay is the time between an sdu
being presented for transmission and the point where it is ready for decoding at the
appropriate synchronisation reference
for framed sdus the cig transport latencies are
transportlatency cigsyncdelay ft isointerval sduinterval
separate calculations need to be made using the respective values of cigsyncdelay ft
and sduinterval for the central to peripheral and peripheral to central directions
for a bi g using framed sdus
transportlatency bigsyncdelay pto nsebn irc isointerval
isointerval sduinterval
for unframed sdus the calculations are slightly different for a cig
transportlatency cigsyncdelay f t isointerval sduinterval
again you need to use the respective values of cigsyncdelay ft and sduinterval for
the central to peripheral and peripheral to central directions
section isoal the isochronous adaptation layer
for an unframed big
transportlatency bigsyncdelay pto nse bn irc
isointerval sduinterval
ooo
that concludes the basics of isochronous stream s now we need to look at the lc and see
how qos choices influence robustness and latency
chapter lc latency and qos
chapter lc latency and qos
introduction
two of the most debated aspect of wireless audio particularly amongst audiophiles are audio
quality and latency in its early years bluetooth technology was often criticised for both
although in most cases the audio quality probably had more to do wit h the state of the
transducers rendering the audio than anything to do with the bluetooth specifications
transmitting audio over any wireless connection involves compromises interference is a fact
of life which means that some of the audio data will be lost that results in gaps in the audio
unless you take measures to add redundancy typically that inv olves transmitting the audio
packets more than once so that there are multiple chances that one of them will get through