SoundStation and Sabre
DAR have chosen Ethernet since it is commonly available, it works, and it saves
them having to reinvent the wheel (or the hub). It doesn't mean that Ethernet
is the perfect network system, but the advantage of buying in standard technology
from the computer industry at large is that when a better standard – in all
respects – comes along, DAR will be able to adopt it easily, and hopefully the
new standard, whatever it may be, will be able to use the existing Ethernet
cable so that upgrading doesn't mean ripping out the entire network and starting
again (perhaps you should install extra twisted pairs in advance). The DAR network
will run on SoundStations and Sabres as peer to peer groups quite happily, so
several operators will be able to access each others' files as and when needed.
This would obviously be useful, but perhaps not such an extraordinary benefit
since bulk data storage will still be necessary at each location. The real advantage
of having a network will come when DAR introduces a specialised server product.
You could use a SoundStation or Sabre as a server, but the server doesn't need
to have replay facilities since it will be a data repository for everyone to
use. A typical example of how this would work would obviously be to store a
central sound effects library, or perhaps rushes could be loaded onto the server
so that editors can access what they need when they like. DAR are developing
an interface for the server which will run under Windows where you can see lots
of files on a high resolution screen, top and tail files, and arrange them into
DAR-style directories and groups.
The editor will be able to appreciate the benefits of the network without
any overhead of new things to learn, since he or she will be able to connect
the machine to the network, when it will recognise the server in much the same
way as it recognises a new optical disk being inserted in the drive. The contents
of the server will appear as just another group in the top level of the directory,
prefixed by a symbol that identifies the server as the source of the material.
Assuming that there is some edited audio on the server, the operator will be
able to download an EDL, or 'reel' in DAR's terminology, which is a relatively
small file containing all the edit information, and the local SoundStation or
Sabre will access audio directly from the server as the reel plays. Yes, of
course there is bound to be some limitation on the possibilities of working
in such a way. The limitation is that it is only possible to access two channels
in real time over the 10 megabits/s Ethernet link. But this might be enough
for editing since these can be any two tracks from the reel, and segments can
be edited and slipped in exactly the same way as if they resided on the local
machine. Or you might want to work the other way round and keep the audio on
your local hard disk and when you have finished editing, upload the EDL to the
server for someone else to merge with his/her part of the project. A third possibility
that avoids the two channel restriction is to download the audio files themselves
from the server, which will take a little time perhaps. A fourth would be to
use two channels from the server in conjunction with locally stored material.
Of course, the existing Ethernet standard, as I said, is an interim solution
and the bandwidth of the network will increase in the not-too-distant future.
One of the ways the network will immediately be put to use is in backing up
material, for safety copies and for freeing up disk space to work on another
project. At the moment, editors can back up their work from a standalone machine
onto optical disk, but the capacity of optical disks is limited and taking a
snapshot of the state of a project at the end of the day would be a time consuming
process. With a network and a server with plenty of storage space, copying can
be done in the background while you are still at work, or you could set it to
back up the entire project last thing before you set off home. When a project
has been backed up once, backing up again is much quicker because most of the
audio files will be exactly the same as they were when you started, so they
don't have to be copied again. Once you have a server, backing up to a removable
medium is also more practical because one Exabyte tape drive can be connected
to the server rather than having one for each workstation.