Hi all,

I've had in mind for a while now to build a high-quality Csound-based
sampler.  I was thinking originally that the sampler would be best to
implement the same features as some other formats to allow conversion
to from one format to the one the csound-based one could use.  The
reason for building a sampler in csound and not just communicating to
one via MIDI or through something like VST is that I want a format
that is cross-platform, completely open, and one which can be flexible
to use within the csound world (i.e. can take in frequency instead of
MIDI key nums so that it is easier to do glissandi as well as
microtonal work).

The sampler code I had planned to do all in Csound itself and not as a
separate binary opcode plugin.  This allows flexibility.

The sample spec that csound would need to read would have to be a
something one of the gen routines could read in; this should be fine
as the spec that csound will read is expected  to be generated by a
program.

Now, until today I wasn't sure on what kind of features to implement
nor what samplers to model after, but I came across a mention to the
sfz file format on one of the microtonal mailing lists:

http://www.cakewalk.com/DevXchange/sfz.asp

The program extreme sample convertor (http://www.extranslator.com/)
can translate to sfz.

This format is open but I don't know that much of the history of this
format and more importantly if it seems like it will stick around.
One thing that seems a little strange is not finding much commercial
support for this format which I assume is because it is extremely
open.

Anyone have thoughts on this format?  Perhaps it'll be worth using as
the model and ultimately other formats could be added to convert from.

steven