[linux-audio-user] soundfonts and samplers

[John Check]

On Wednesday 11 August 2004 04:28 pm, Robert Jonsson wrote:
> Hi,
>
> onsdagen den 11 augusti 2004 21.30 skrev Iván Almada:
> > About SoundFonts:
> > (reply to robert's soundfonts rants)
> >
> >
> > i havent used soundfonts at all (never),  i was
> > planning to use them as a last resort not because
> > everybody has a bad attitude towards sf,  but what
> > youve said that theyre monolithic,  when im composing
> > i like to make new samples and load them in-the-fly to
> > a sampler and then process the sample more with fx and
> > stuff... and again... render to wave and then load- it
> > again... maybe change some rythmic structures and then
> > render again to audio and maybe the complete audio
> > chunk load it as a whole audio track in ardour or some
> > audio seq software with mixer to further process!!!
> > thats the way i like to make music,  so the point is
> > that soundfount could let me achieve this easily? or
> > is it difficult to compile your own soundfonts... the
> > other problem i think ill encounter is that sequencers
> > cant record the audio directly from your soundcards
> > wavetable... (please correct if im wrong)...
>
> Oh, I don't think that is a problem, most soundcards have a selectable
> input, you can generally set it to record from several sound sources,
> including the wavetable.
>
> > on the other hand ive been searching for sf and ive
> > realized that maybe they will be very useful to me
> > because you can load pre-made drum kits like a TR-909
> > for example and use them easily just as if you had the
> > real midi module...
>
> Yes, this is a very good use of soundfonts, there are some excellent fonts
> on the net for old drum machines!
>
> > another question: my soundcard has 4 synths built in
> > right? does this means i can load 4 soundfont banks at
> > the same time???
>
> I'm not sure you said what soundcard you had, but with four separate synths
> it sounds like an sb-live right?
> It was a while since I tried this (I generally run fluidsynth which also
> uses soundfonts), I think they use the same "synth engine" though, so they
> have the same sounds. You can have a different sound for all 16 midi
> channels though, I recall it is somewhat hard to load fonts on all channels
> though, quite a bit easier with a softsynth like fluidsynth (especially
> with a frontend like Qsynth or the one in MusE).

Power of the GUI right there. The main problem with soundfonts in the emu10k1 
chipset is no FX. There is a compiler for the DSP microcode, but the loader 
has yet to be ported from OSS.

>
> > on the other hand ive realized specimen is a great
> > sampler:  everything i need (preview .wav, map
> > different regions, loop point editor, filter, though
> > more different filters would be very very nice...) and
> > i didnt know you could use it directly from rosegardan
> > through alsa midi ,  that excelent,  just as if
> > specimen was a plug in, i can trigger samples from my
> > midi keyboard with an inperceptible latency,  now i
> > would like to know several more things,  like how to
> > apply ladpsa plugins to the sampler for example if i
> > want to use some of the ladpsa filters... how is the
> > audio routed from specimen back into the midi host
> > mixer (it displays audio in a channel called rec... )
> > but it has no slots for plug ins...
>
> This is, as Chris mentioned, done with a mixer application, like Jack-rack
> (pending that specimen can output to jack, but I think it does). I think
> you could use Rosegarden as the mixer application also, I use MusE like
> this, add an input and connect the jack-app to it, this limits the number
> of apps I need to keep track of and most settings are stored in one place.
>
> > anyone knows how to do this??? could this be done? or
> > do i have to record the audio and then process it in a
> > standard audio track ?
> >
> > thanx
> > ivan.
>
> /Robert