15 Apr
2003
15 Apr
'03
1:16 p.m.
Has anybody investigated the Gibson MaGIC protocol for use on Linux? It looks like a
perfect match: lots of audio channels and control data over CAT5. Just buy a separate
network card, port their protocol to Linux, and then you could use their 16-in 16-out box,
or get input directly from one of their new guitars with the RJ45 port on it. The The
only scary bit is that the license is free "for the next10 years" only.
-Ben
15 Apr
15 Apr
1:37 p.m.
Has anybody investigated the Gibson MaGIC protocol for
use on Linux? It
looks like a perfect match: lots of audio channels and control
data over
CAT5. Just buy a separate network card, port their protocol to Linux, and
then you could use their 16-in 16-out box, or get input directly from one of
their new guitars with the RJ45 port on it. The The only scary bit is that
the license is free "for the next10 years" only.
I may be wrong but I don't think this protocol will work with a
normal NIC. I'm not sure what (if any?) the advantages of this
protocol are over IEEE1394.
--ms
2:14 p.m.
Has anybody
investigated the Gibson MaGIC protocol for use on Linux? It
looks like a perfect
match: lots of audio channels and control data over
CAT5. Just buy a separate network card, port their protocol to Linux, and
then you could use their 16-in 16-out box, or get input directly
from one of
their new guitars with the RJ45 port on it. The The only scary
bit is that
the license is free "for the next10 years" only.
I may be wrong but I don't think this protocol will work with a
normal NIC. I'm not sure what (if any?) the advantages of this
protocol are over IEEE1394. 
2:29 p.m.
On Tue, Apr 15, 2003 at 09:04:53AM -0700, Mark Knecht wrote:
(cut)
2) CAT5 cables would be superior to 1394 as they are
locked into the
connectors and wouldn't fall out of your guitar. They are longer and more
flexible. On the downside, CAT5 connectors quickly wear out so what works
well today may not work well after a few evenings of heavy gigging.
(cut)
They should try it with Bluetooth :)
v
2:51 p.m.
> > Has anybody investigated the Gibson MaGIC
protocol for use on Linux?
It
> looks like a perfect match: lots of audio
channels and control data over
> CAT5. Just buy a separate network card, port their protocol to Linux,
and
then you could
use their 16-in 16-out box, or get input directly
from one of
their new guitars with the RJ45 port on it. The The only scary
bit is that
the license is free "for the next10 years" only.
I may be wrong but I don't think this protocol will work with a
normal NIC. I'm not sure what (if any?) the advantages of this
protocol are over IEEE1394.
I've looked at the protocol a very little bit. (I work in the 1394 area.) have not seen a detailed spec though, so I only have
impressions. Take the
following with a grain of salt.
1) I do not think the packets that are sent in this protocol will
necessarily coexist very well with standard Ethernet, so the observation
of
doing it with a second NIC is probably correct from
what I can tell.
That's
really no different than replacing a 1394 adapter with
a second NIC, which
is probably less expensive if it works.
It will _not_ coexist with standard Ethernet and I doubt a NIC can be
used because the timing is very critical. You'd probably have to buy a
dedicated MaGIC interface.
2) CAT5 cables would be superior to 1394 as they are
locked into the
connectors and wouldn't fall out of your guitar. They are longer and more
flexible. On the downside, CAT5 connectors quickly wear out so what works
well today may not work well after a few evenings of heavy gigging.
FireWire will be able to use CAT5 also IIRC.
--ms
5:22 p.m.
On Tuesday 15 April 2003 19.40, Martijn Sipkema wrote:
[...]
It will _not_ coexist with standard Ethernet
Right; it *can't*, unless all machines on the network have a hard real
time layer between the NICs and the protocol stacks. There are
protocols that would work for running RT streaming together with
normal network traffic, but there's no way around the fact that all
nodes must be hard real time. (Well, except for building it into the
hardware, perhaps...)
and I doubt a NIC can
be used because the timing is very critical.
That's not a problem normally, as most NICs have hardware latencies in
the microsecond range. All you need to implement hard QoS guarantees
is an RTOS, RT safe drivers and a suitable protocol.
//David Olofson - Programmer, Composer, Open Source Advocate
.- The Return of Audiality! --------------------------------.
| Free/Open Source Audio Engine for use in Games or Studio. |
| RT and off-line synth. Scripting. Sample accurate timing. |
`-----------------------------------> http://audiality.org -'
--- http://olofson.net --- http://www.reologica.se ---
9:01 p.m.
On Tuesday 15 April 2003 19.40, Martijn Sipkema
wrote:
[...]
Actually, from the specs:
"The default MaGIC frame timing is 48 kHz with an acceptable tolerance
of 80 picoseconds."
I think this needs to be implemented in hardware.
--ms
It will _not_ coexist with standard Ethernet
Right; it *can't*, unless all machines on the network have a hard real
time layer between the NICs and the protocol stacks. There are
protocols that would work for running RT streaming together with
normal network traffic, but there's no way around the fact that all
nodes must be hard real time. (Well, except for building it into the
hardware, perhaps...)
and I doubt a NIC can
be used because the timing is very critical.
That's not a problem normally, as most NICs have hardware latencies in
the microsecond range. All you need to implement hard QoS guarantees
is an RTOS, RT safe drivers and a suitable protocol. 
8:39 p.m.
On Tue, Apr 15, 2003 at 06:40:23 +0100, Martijn Sipkema wrote:
It will _not_ coexist with standard Ethernet and I
doubt a NIC can be
used because the timing is very critical. You'd probably have to buy a
dedicated MaGIC interface.
Standard ethernet or standard TCP? One means you can (maybe) use standard
NICs, one means you definatly can't. Most NICs are hardwired for ethernet
AFAIK.
- Steve
7921
days inactive
7921
days old
linux-audio-dev@lists.linuxaudio.org
9 comments
6 participants
participants (6)
-
Ben Loftis -
David Olofson -
Mark Knecht -
Martijn Sipkema -
Steve Harris -
Vincent Touquet