On Tue, 11 Dec 2012, Thorsten Wilms wrote:
> It has been too long since I last cared about Specimen, but aren't
Ok, thanks... I should actually probably just post what I'm trying to
I previously handled most of my drum sampling needs from an Akai MPC2000
sampler/sequencer. This actually a great machine, however since it is a
hardware device that uses its own proprietary filesystem, and because
Akai has now decided they're going in an entirely Windows/Mac-based
virtual instrument direction instead of supporting their existing
hardware, it is clear that these machines are going to be more and more
of a pain to maintain projects on. I have many vintage instruments, and
I'm ok with most of them because their only job is to create sounds, but
I can't be relying on an unsupported piece of hardware to save and
manage actual project data. So I'm moving all my sampling and
sequencing needs to Linux open source apps, which I probably should have
done a long time ago. There's a lot of really great software out there.
For drum sampling, the first place I looked was Hydrogen (of course).
Without a doubt, hydrogen drumkits are the file format Linux studios
seem to exchange sounds in even if they don't like Hydrogen. I've found
it has some limitations though...
* The sequencer leaves a lot to be desired: I realize it's built to
emulate old Roland x0x style drum machines. For many styles that's
great, but sometimes I like to program patterns that are similar to what
a drummer would actually play in a non-electronic track, with guitar and
bass. Hydrogen's "everything is a looping pattern" paradigm is a bit
awkward for that sort of stuff. Also, unless I'm missing something,
there is no undo, and no way to clear all the notes from a pattern at
once without deleting the pattern. It seems to have no notion of
oddtime unless you fool it by creating bars made out of arbitrary
groups of "beats." The quantizer leaves a lot to be desired -- it's
very seldom that I ever sequence anything on it and have it guess the
note subdivisions I meant it to record...and I'm a drummer, so it's not
that I'm playing the patterns too poorly.
* While it's really nice that Hydrogen lets you send each instrument out
a separate Jack port (perhaps its nicest feature, I think), it'd be nice
if the ports didn't all have to be stereo. Something I'm doing with the
last 6 analog outs of my 8-by-8 RME interface is having samples leave
the interface to be processed on my analog mixer, and then fed back into
the digital mix. It'd be really much more convenient if I could have
some of these outputs be mono, and handle panning on my actual mixer
desk. That's not crucial, but it'd be nice.
* Speaking of stereo, even within Hydrogen, there's something screwy
about the stereo imaging: It sounds like left and right are being mixed
together somehow, like there isn't a full separation of the channels.
I've confirmed this by listening to the same samples in Petri-Foo. It's
almost as though there's something off about Hydrogen's mixer.
So...because of these reasons, I planned on using Muse or Rosegarden for
drum sequencing, and at most just using Hydrogen as a sample playback
engine for drumkits (if I use it at all). I'm already much more
impressed with the sequencing in both those programs. However, since
I'm talking about just needing an engine for drumkit sample playback, I
thought I'd look into other possibilities besides Hydrogen.
That's how the investigation of Petri-foo got started. It doesn't seem
to have multiple Jack outputs unless you want to run multiple sessions
of Petri-foo, plus from what you've described, it sounds like it can't
even do something as simple as a hihat choke within a single patch. So
much for Petri-foo (too bad, nice program otherwise)...
My next direction is to try to create a SoundFont in Swami. It looks
like it should be possible, and then perhaps I could get what I want
from LinuxSampler? Can I setup a virtual drumkit without multiple
outputs and hihat cutoff within the patch using SoundFonts?
+ Brent A. Busby + "We've all heard that a million monkeys
+ Sr. UNIX Systems Admin + banging on a million typewriters will
+ University of Chicago + eventually reproduce the entire works of
+ James Franck Institute + Shakespeare. Now, thanks to the Internet,
+ Materials Research Ctr + we know this is not true." -Robert Wilensky
Linux-audio-user mailing list