[linux-audio-user] Rewrite sfxload for native ALSA

Robert Jonsson robert.jonsson at dataductus.se
Thu Dec 4 08:33:38 EST 2003


hursday 04 December 2003 13.45 skrev Mark Constable:
> On Thu, 4 Dec 2003 12:34 pm, LinuxMedia wrote:
> > > My thoughts are to try and create a reasonable piece of music
> > > that *I* find listenable and not too embarrasing, mainly so there
> > > are no copyright issues, and create some oggs then write up
> > > (heh, sure pal) a HOWTO and how I created the end result (that
> > > will never be "ended" because it could always be reused and
> > > re-released as another version).
> >
> > ...
> > Actually, there's much more to it than the above stuff. Of course,
> > there's the ongoing building of soundfonts and the extra stuff like
> > that. In fact, there's a lot of things like that I'm doing in along with
> > the above to keep inproving the situation. But aren't we all (-:
> What I lament the lack of is some system whereby there is a
> coordinated effort amongst a number of folks to come up with
> some kind of quasi-standard in a few areas, particularly with
> a "standard reference" soundfont so that when we play back any
> MIDI files, for instance, and they sound the same. I'm not
> suggesting some system that demands people do something in a
> certain way but allows "us" to build on top of each others
> efforts and explorations as ALSA and the apps that use it mature.

I don't know, it sounds hard to standardize sound. Isn't it a very private 
thing which sounds you wish to pick?  
There IS the GM standard... to me it's almost always just "in the way". I pick 
sounds much more randomly during a composition.
> For instance, it's taken me a week of effort to come up with a
> method of using bash shell scripts to be able to easily play
> any MIDI file with Timidity using the two guspat sets and 4
> or 5 soundfonts so I can get a better grip on the relative
> quality of each...

See below for some comment ;).
> . I'm sure someone has made this same effort in the past yet
>   I had to go through a week of hair pulling just to move a
>   half an inch forward... so I can _start_ to create some music
> . I'm equally sure _some_ other people in the future will want
>   to go down this same path and so they will also have to spend
>   hours working it out before they can start making music
> I'd like to see some structure/system whereby mine and other
> peoples efforts can build up "something" to eleviate us all
> (newcomers to linux audio) having to reinvent basic understanding
> before we all can really get on with what ALSA is mostly about,
> enabling people to not only just play, but create, music.
> Like I say, one little thing I hope to do is finally create
> some piece of music I am mostly happy with and then not only
> provide all the parts and components as examples but try to
> write up the whole procedure that went into creating this piece.
> To say the very least... if someone else had already done
> something like this, so I could simply follow the instructions,
> then I would be VERY happy not to waste my time reinventing
> wheels most likely already spinning elsewhere.

When it comes to music I'm also mainly interested in something that just 
works. As it happens I've gotten myself involved in various other ways 
including programming (it IS another interest of mine), but if I could choose 
I would use something that already worked as it should.

My solution is to use the internal FluidSynth of MusE, and it continues to be 
the solution since it is a working solution. 

<timidity rant>I really don't understand the fuzz about timidity, I tried it 
(again) a while ago, it may be true that it has good sound quality... though 
noone knows for sure since it's impossible to set up ;) (a bit sarcastic, but 
there is some truth to it).
Needles to say, I failed to set it up to do anything... anything at all. </>
With Fluidsynth there are several nice guis to choose from that improves 
usability no-end. 
Since I use the Fluidsynth that is available internally in MusE I also benefit 
from the added integration that this gives, patch selection being saved with 
the song file and so on... which basically brings me standing at the 
just-works goal.

My 2 cents

More information about the Linux-audio-user mailing list