11 Nov
2004
11 Nov
'04
2:21 p.m.
I've recently been conscripted into the role of backup to our primary
phone system admin with an eye towards leveraging my network admin
background in our coming VoIP deployment. Due to this I've been cramming
as much telephony information as I can into my brain over the past 3-4
weeks. Our PBX (private branch exchange, i.e. the switch for our office
phone system for those unfamiliar with the term) is a proprietary system
from NEC. Having to deal with systems like this is contrary to my
nature.
To sooth the irritation of being rubbed the wrong way by this
proprietary technology I've signed on to the asterisk users and dev
lists with a few aims. First, as I expected, I've found the members of
the asterisk community to really know their stuff when it comes to the
standards and protocols that make phone systems work. In a few days of
reading the lists I've already learned much that I haven't gotten out of
the NEC manuals, but that helps me to better understand their closed
system. Second, I'm hoping that down the road I can get an asterisk
system into this operation to provide additional services and
functionality that would be more expensive to purchase from NEC.
I'm writing to LAU to get some feedback on a number of possiblities that
come to mind for cross fertilization between this community and the
asterisk community. Also I'm hoping there are some here who have
experiences with asterisk they would be willing to share.
As I understand it asterisk can use ALSA supported full-duplex cards to
provide voice i/o. An asterisk server with a number of connections to
the phone network and several RME HDSP or other such high channel count
multi-channel cards would seem to be a very useful, cost effective and
high quality solution for supporting call-in shows and telephone
interviews for a radio station. Such a settup could also provide a nice
platform for an intercom system for a business or even a home. Is anyone
here doing such things?
Apropos my recent inquiry regarding bats, telephony has traditionally
saved bandwidth by limiting the frequencies transmitted to a roughly
4kHz band since the information important to intelligible speech can be
conveyed without the sounds outside that band. Are the concepts used to
capture bandlimitted audio for speech the same, or similar to, what
would be used to capture the interesting information from sounds
produced by animals who hear above the human hearing range?
There are a variety of audio data compression and synthesis/resynthesis
schemes in use in the telephony world. Have any of these been repurposed
for use as effects, perhaps wrapped up in LADSPA plugins?
Are there similarities between jack and asterisk in what they need to do
to provide audio routing and scheduling? Perhaps this has already
occured or perhaps their needs are too different, but could the two
projects benefit from sharing ideas or even code? If these are naive
questions and the two domains are orthogonal I'm interested in knowing
why. Hmm ... as I write I realize a big difference is that many phone
conversations happen at once and have no need for synchronization.
jack's typical application space involves keeping many channels of audio
in sync. So, I guess I've largely answered this one for myself. But,
still input from the system programming gurus like Paul and Jack would
be most welcome and surely enlightening and educational.
I had a few other ideas and questions, but they've slipped away from me.
Anyway, this has gotten long enough.
Thanks in advance for reading and for any feedback.
-Eric Rz.