22 Mar
2010
22 Mar
'10
10:44 p.m.
On Mon, March 22, 2010 18:50, Ralf Mardorf wrote:
Louigi Verona wrote:
I'm with Louigi on this - I'd like to set up a sequence in whatever
sequencer I'm using and draw graphs to control cut-offs modulations etc
etc in some softsynth BEFORE the sound is output/recorded.
But because I can't do this, I have to settle with recording the audio of
the synth into ardour, and then use automation to control the params of
various LV2/LADSPA plugins. It has provided some interesting (to me)
results :-)
http://jwm-art.net/art/audio/qtest_fallibility.mp3
but I doubt the engineers (or real musicians) amongst you will approve.
James.
Both are still limited in the sense that you have
to define that data explicitly.
Ciao,
--
FA
I am sorry, I do not understand what kind of limitation you are
speaking about.
I may not be a programmer and perhaps I lack knowledge and
understanding of some concepts like this
defining data explicitly, I do know that currently on Linux I cannot
do a whole class of things. So if I wanna
do an ambient tune and use Zyn as my sound source, I would not be able
to automate, either in real time
or non-real time mode in a graph some of its parameters. And I do need
that functionality. It can be "limited",
but this is all I need, since just playing notes is no good. I need to
work with the sound itself. This is what
automation is for me, it is not about mixing, it is about manipulating
sound itself.
Again, maybe I am not getting something.
Louigi.
Because of movements for knobs in the timeline I can't see any
limitations, without sample-accuracy, just the ticks should be good
enough. Because for the values of the data 128 steps I guess for most
sounds should be enough, but indeed because of the Hammond example,
14bit or 16bit would be better.
I don't agree because of the limitation.
Ralf