20 Oct
2002
20 Oct
'02
4:41 p.m.
dgm responds (embedded comments inline):
Mark Knecht wrote:
I've been staying on the sidelines of this thread up 'til now. I'd like
to add a users' perspective. I am not a programmer at all. I come at this
totally from the point of view of a user.
So do I, for the most part.
There is a really GREAT reason for people to start migrating somewhere
though - 96KHz support.
The biggest problem with 96 kHz is that at the moment there isn't a
really great way to get a 96 kHz
digital stream from one device to another. Double speed AES/EBU is
probably the best solution, but
not many people have multiple channels of AES. You can send 96 kHz over
lightpipe or TDIF, but you
lose half the channels. Lots of cards (like my nifty Delta 1010)
support it, lots of apps support it, but once
it's inside the computer, there's not an easy way to get it out. Not
only that, but it's going to be awhile before
everyone has DVD-A, so most people won't be able to listen to 96 kHz
recordings anytime soon. The point:
44.1 and 48 kHz are going to remain standard for awhile, and the
presence or lack of 96 kHz at this point
doesn't seem to make or break a platform for most people that I know.
Everyone is worried about getting trapped in a
close, expensive, buggy solution when they more to 96K. DigiDesign is really
messing with the semi-pro guys. They know it, but they don't know what to
do. That's why I'm here.
Digidesign is a pretty arrogant company, as I see it. They just
unveiled all this nifty new hardware, like
the Digi002 and the ProTools HD systems (96 is one thing, but who really
needs 192 kHz?), but they're
still committed to Mac OS 9. I have the feeling that the only reason
Apple still puts OS 9 on their
boxes is for the people who want to run ProTools. (I cite the continued
preference of many musicians
for classic MacOS as proof of the inherent conservatism of many of the
people we're trying to reach.)
That problem is
the lack of MIDI integrated to work with audio. Today we have some great
MIDI apps coming along (Muse, Rosegarden, Midi Mountain) Ardour on the audio
side, soft synths galore (although they are not all usable in an integrated
way yet) loads of interesting plugins and jack to allow the audio to talk
between apps.
Fair enough criticism, I suppose. The list reflected my own personal
needs and priorities, and I don't
often use or care about MIDI. (I like old-school text-driven synthesis
languages and graphical synthesis
environments like PD and jMax.)
But, unfortunately, there's no way I know of yet to
talk about
all of the Alsa apps to users who want to replace Pro Tools, which is where
I'm coming from. This needs to be handled in a convincing way before talking
to people, or the negative press will be very hard to overcome.
Here comes my little rant. The 'problem' here is that Linux apps are
developed by very intelligent, well meaning, technical people who work on
what they want to work on and not necessarily what's wanted and needed by
the 'marketplace'.
True enough. One sees this in Linux across the board. I do, however,
think that this is changing. Geeks and
non-geeks can blissfully coincide. One does not need to give up emacs
to use OpenOffice.
I think that some real program management is required.
Look at what the market wants and needs, (audio, MIDI, soft synths, plugins,
scoring, automation, etc.) write it down somewhere and then do an honest
assessment of where Linux audio solutions are.
I don't think lack of awareness is the problem. The problem is that
most Linux audio apps are developed by
people who have full-time jobs doing other things. The problems
involved in designing audio apps are so great
that even those people who are able to work full time on Linux audio are
often stupmed as to how to implement
the desired solutions. Everyone knows we need a decent midi sequencer,
and I for one would love to have a truly
usable notation program. Great. Just about everyone who knows how to
develop such an app is working for one
of the big companies, and you can bet they ain't going to open-source
their code anytime soon. Guys like Paul Davis
and Bill Schottstaedt do not have the benefit of a team of full-time
codes who can help them solve the sorts of problems
they face.
If we have what the users
need, then we talk. If we don't, then we put a program in place to make that
stuff happen fast,
Suggestions as to how to make this happen?
so that when we do have it we can talk. Unfortunately,
this would require the developers to respond more like work and less like a
hobby. I don't know that they want to do that. <EOR>
Beyond that, we need to handle the installation stuff much more cleanly.
PlanetCCRMA is doing a great job of that for the Redhat platform.
Yes they are. Nando rocks.
I know
that many people wouldn't want to be forced into a distribution long term,
but in the beginning many people who will come and try out this stuff won't
be Linux users and should be directed towards a COMPLETE solution. Currently
I think the Planet is the best I know of.
AGNULA is also a possibility.
As I don't want you to think I'm negative on what's going on here, I
think there are MANY things that Linux could offer, but isn't even trying to
talk about yet. Things like:
1) Real multi-processor support
2) Distributed processing where different apps are on multiple machines all
working together
3) Remote access - all the application computers are in a different room and
one very quiet PC is used in the studio to display their screens. No noise,
but water cooling not required.
4) Much more stable platform. No reboots, no sad Macs, no BSODs.
5) More open hardware support, presuming someone ever makes the multicard
thing understandable by those of us that don't have a PhD. in Alsa.
Unfortunately Linux is no longer the only platform to offer all that.
Mac OSX has all of the above, as wel.
At this point, I highly doubt that anyone is going to chuck their ProTools
(TM) system in favor
of Linux anytime soon.
[MWK] I'm trying!!!!!!!!!!!!
Great! But are we talking ProTools LE or a big ol' nasty TDM system? I
could see someone ditching
LE, but TDM users (and the highest echelon of the market is exclusively
TDM, even if the software is not
ProTools) have too much invested in the hardware.
The thing that could do it is to have some killer apps - 90% of the
functionality of say, Logic would be good enough - with free redistribution.
[MWK] Yep! The caveat is that you have to have the right 90%!
As I said before, the problem isn't that the developers are ignorant as
to what the users want, it's that they
don't really know how to implement it.
[MWK] Great examples, but I'm not sure this morning what 'SC' is. However,
not a MIDI app in the bunch yet, as per my comments above.
SC = SuperCollider. Formerly a proprietary MacOS app, recently GPL'd
and ported to OSX, and
there's a project underway to port the OSX version to Linux.
-dgm