Assuming this an open question, I'll share my thoughts...
I love Linux audio for its ideology and the huge efforts of its
contributors (something beyond my capabilities). But I do find I'm
constantly' battling against the elements'. Ardour 3 closes spontaneously
during sessions, xruns are unavoidable under certain circumstances no
matter what settings I use and there seems to be a limit as to which
plugins will work alongside one another (e.g. QSynth and JAMin can only be
used with a very light Ardour load).
I understand many won't have these problems, but that's the point -Linux
audio's stability varies from system to system and that's the only negative
thing about it compared to corporate OS/Audio. The engineering process
sometimes sucks the life out of my creativity because I spend more time
searching for solutions to tech problems than revelling in the creative
musical process.
But I love it. When working well it has everything I need, the design of it
is great and I love the jack system. It will, with a few hitches, allow me
to record my music without having to take out a bank loan.
Dan
--
*Dan Capp*
~ Design & Illustration ~
www.dancapp.com
I am currently creating midi files using Noteworthy Composer under Wine. I
allocate sound banks like so: asfxload -b [bank no.] [font] and staves in
Noteworthy can be allocated to particular sound banks. I then create wav
files to use use a backing tracks (for practice) using fluidsynth having
first mapped soundfonts to presets in fluidsynth (I like to use skifta as a
remote which is very handy). All of this works well but I would like to
switch to a native Linux notation editor. My preference would probably be
denemo or musescore, rosegarden has a lot of functionality i don't really
need. I am struggling to find how I can call multiple soundbanks in a single
score and export those as a midi file so fluidsynth can utilise the relevant
soundfonts. Grateful for any advice on resolving this -- thanks
--
View this message in context: http://linux-audio.4202.n7.nabble.com/Need-to-create-midi-files-using-multi…
Sent from the linux-audio-user mailing list archive at Nabble.com.
some people here & there writes in this thread:
> Linux is THE hobbyist operating system
i work in a big fancy datacenter. so, the question is:
should i treat about 3/4 servers there as 'hobbyist' ?
so. i suspect that the real problem is somewhere else.
> Here, on Linux, there is no such thing as market competition.
> And thus - no natural selection of software, so to speak.
there IS natural selection on linux, since there are basically two
kinds of linux audio software: 1) nearly-working & 2)
sometimes-semi-working never-finished pre-alpha snapshots. first of
these two is MUCH better.
> Most good music stores will let you try before buying. I take my computer
> right into the store and plug things in. It does mean I don't buy stuff
> that requires extra tweaks... just stuff that "just works".
so, at last the community MUST start some kind of online list or
database of class-compliant devices that just work (or work with
limitations that should be listed there too).
> Interesting, I have been thinking that making music (or video, graphics,
> whatever art form) is another form of development.
the problem is the fact that marketologists don't think so.
they think that techinical side should be hidden as much as possible,
and the user should interact with one or two magic buttons only.
On Wed, Feb 6, 2013 at 10:52 AM, cunnilinux himself wrote:
> the problem is the fact that marketologists don't think so.
> they think that techinical side should be hidden as much as possible,
> and the user should interact with one or two magic buttons only.
"I'm an artist. I speak in hyperbole" (c) The Oatmeal
Right? :)
Of course, you can always try to convince Louigi that he needs
separate channel ports in gladish on stage. Because, dontcherknow,
half the live session should be spent on connecting each L and R
output separately.
So, basically,
s/the problem/presumably, the problem/
s/the fact/according to my assumption/
s/marketologists/some marketologists/
s/technical side should be hidden as much as possible/software should
not get in the way/
s/interact with one or two magic buttons only/spend as little time on
boring work as possible/
Is that UNIX enough for ya? :)
Personally, I salute to people who think that e.g. client/server
architecture of LinuxSampler is great and should be directly exposed
to users. I'll never be half as cool. I'm a complete and utter bore
who never gets any life. Unlike the sparkling minds of true UNIX gurus
who get all the girls/boys, because talking about low-level stuff is
so frigging sexy :)
Alexandre Prokoudine
http://libregraphicsworld.org
On Wed, Feb 6, 2013 at 11:38 AM, Alexandre Prokoudine wrote:
> Personally, I salute to people who think that e.g. client/server
> architecture of LinuxSampler is great and should be directly exposed
> to users. I'll never be half as cool. I'm a complete and utter bore
> who never gets any life. Unlike the sparkling minds of true UNIX gurus
> who get all the girls/boys, because talking about low-level stuff is
> so frigging sexy :)
P.S. Actually, one of the serious problems in software design is
understanding when low-level stuff is OK to expose, and what counts
for low-level. Historically free software developers solved this in
three ways:
1) dump everything into one huge dialog and let the geekest survive (FAIL)
2) put everything extra under "Advanced" tab and let users pick their
own settings (FAIL)
2) get rid of extra settings, streamline defaults (FAIL)
The 3rd should actually be WIN, was it not for the screaming majority of geeks.
Alexandre Prokoudine
http://libregraphicsworld.org
> I have 2.0.99+svn2019-1ubuntu4 here. Building FFADO would mean
> building JACK, and that would lead to a chain of rebuilds I'd rather
> not do. Although, maybe Kxstudio PPA already took care of that.
Hmm, I have built ffado many times on AVLinux and have never had to rebuild
Jack.
(though I had to do it a bit carefully because I could not remove all the
original ffado package because
then synaptic wants to remove gobs of packages, so I just made sure to
install to the
same location as used by the package, i.e. /usr, which overwrites the old
libraries. Also,
had to fix some links /usr/lib which I admit is a bit ugly - well I only
had to do this to build SVN 2260, otherwise,
I could have stuck with the AVLinux package which works fine, but lacks the
new mixer).
> (see: http://www.premiumorange.com/la-page-web-of-phil).
> I'm not sure what I'm supposed to see there. If you are referring to
>the podcast, je ne parlais pas francais tres bien :)
Sorry, I screwed up the link here is the correct one:
http://www.premiumorange.com/la-page-web-of-phil/index.php?page=P030601
Concerning this comment about the Focusrite Saffire 24:
> As a Focusrite Saffire 24 user who needs to start JACK twice every
> time and, since Ubuntu 12.10, restart every 30 minutes, I'd think
> before recommending it to a Linux user :) Absolutely love their
> hardware and their attitude to clients though.
The "start Jack twice" problem has been fixed for several months and if you
build ffado SVN 2260 or greater you will also get the improved version of
the mixer
(see: http://www.premiumorange.com/la-page-web-of-phil). I've never had
issues having to restart Jack every 30 minutes with the Saffire 24.
But, the point is taken that building the latest SVN of ffado is something
many audio users would care or know how to do
(though if you run AVLinux or kxstudio this has already be done for you).
Well this has been said before but the main problem with hardware on linux
is not that "linux does not support it" the problem
is that manufacturers do not support linux. Probably the best solution to
that is to get more linux users, which is slowly happening.
On Tue, 2013-02-05 at 10:37 -0500, Dave Phillips wrote:
> I've found a simple metric. Just open a Sweetwater catalog to the pages
> for computer audio interfaces and count the ones with Linux support.
How do you ensure that an audio device _really_ is supported?
well.
> So what do you think sucks about Linux audio ?
personally for me, that's the lack of contemporary pro hardware support.
right now i'm looking for a good USB audio intrface for my linux DAW...
and i feel like i'm back in the mid-90s again. ugh.
and the vast majority of supported hardware is low end. aaaarrrggh.
but that's all. for windows way of audio production, people have
windows, and for mac way, people have macs.
> Too difficult to set up audio system.
> JACK is a pain.
haha :) could anyone on earth PROVE me that windows is REALLY easier
to set up for ACCEPTABLE perfomance?? :)))))
> JACK is not a pain for me, but ask me about PulseAudio!!!
yep. terribly horrible thing.
> We of the LAU community are using a general purpose machine for a
> very specific purpose, and some of us do so very successfully. If
> someone is overwhelmed by the computer and the range of choices Linux
> gives, then he or she may be better off with a specialized machine.
in fact, linux has better potential for building _specialized_
environments for certain task(s), since linux is nothing else than a
construction set by it's nature.
for instance, mine is constructed for running software synthesis with
supercollider and realtime midi input.