Re: [linux-audio-dev] VST link (open?)

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: Re: [linux-audio-dev] VST link (open?)
From: Dustin Barlow (duslow_AT_hotmail.com)
Date: Wed Jan 23 2002 - 05:33:46 EET


> >I know this can be done now with MTC and or STMPE sync between two or
>more
>
>as mentioned elsewhere, no it can't be done with MTC or SMPTE sync.

What I was referring to by "this" (at least in my head and not the words I
typed) was syncing two computers both running an MTC or SMPTE capable
sequencer each of which are doing separate tasks and can communicate on an
analog audio signal path (ie resampling) back and forth. If the desire is
to sync digital signals, then yes, I agree that it cannot be done with MTC
or SMPTE alone.

I know a word-clock sync connection is normally used to digitally sync two
audio cards that live in the same box. I don't know if it can also be used
to slave a card that lives in another box. Cubase can be set to use the
clock of the audio card as it's internal clock, so I presume one could do
that as well to get the SP/DIF and TOSLink ports running in sync and well as
the sequencers. Unfortunately, I have only one Aardvark Q10 (which I love
dearly), so I can't try it.

>this isn't at the level of JACK. such things would probably be
>user-space ALSA PCM devices. Routing audio data back and forth is a
>transport issue and doesn't have so much to do with synchronization as
>it might at first appear. You might say that the transport mechanism
>becomes the sync mechanism, rather than it using a sync mechanism. Perhaps.

I understand and concur. Although JACK could, I suspect serve, to simplify,
as it does with digital audio.

I did follow the thread a while back about using TCP/IP and got the
impression from the discussions that if it were used as the transport
delivery mechanism for communicating with other real-time audio systems,
guaranteed packet delivery is an issue (obviously less of an issue on a
private network). Which leads me to my next set of questions...

If one were to use SP/DIF, TOSLink, or Firewire in place of an ethernet card
to distribute audio and sync, are the same packet delivery issues still
there? Is this a hardware problem, a protocol problem, or both. Does this
problem also effect SP/DIF, TOSLink, and etc. hardware and/or protocols?

Also, how do they compare as far as max bandwidth is concerned? I know that
SP/DIF can do at least 24/96 DTS 5.1 evident by my DVD player when hooked up
to my digital amp using SP/DIF (and yes I normally use the TOSLink
connection, but I was curious if SP/DIF was up to the task and it was).
But, I have no clue as to how close that is to maxing out a SP/DIF based
connection.

>Some people (Universal Audio, TC) think you should be adding DSP cards
>rather than building a distributed network over ethernet or
>whatever. Its actually supposed to be quite hard to max out a
>Creamware SCOPE system, for example, and likewise for a Kyma/Capybara
>system.

Agreed...but it appears that Steinberg has other plans...which is what
prompted this discussion in the first place. :} I presume at some point
that even with these DSP host cards, you still can run into the glass
ceiling of the PCI bus if you wanted to have cross card digital audio
communication. It also seems in the spirit of linux to revive those old
machines in the closet to do something other then run as ftp servers :)
Believe me though, I have no notions of grandieur for my old 486.

db

_________________________________________________________________
MSN Photos is the easiest way to share and print your photos:
http://photos.msn.com/support/worldwide.aspx


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Wed Jan 23 2002 - 05:26:16 EET