From: Andres Cabrera ()
Date: Apr 29, 2005
Subject:
Re: contanenating tables
I've been thinking those concatenation functions and other table
manipulation possibilities could go into the the vectorial opcodes by
adding a few optional arguments (to maintain compatibilty, and not have
to add more opcodes), so you can copy a section of a table into any part
of another table, much more powerful than plain concatenation.
Any thoughts on this?
Andres
On Wed, 2005-04-27 at 22:34, David Akbari wrote:
> Wow, "concat" as an opcode is a great idea! I don't know if it would be
> sufficient to have it as *just* an i-rate though. I say this because
> the tables we are dealing with are mostly getting indexed at k-rate.
>
> If you were say, using the i-rate concat to alter the k-rate force
> vectors of scanu it could sound a little coarse if you start
> interpolating. This may also be true of dynamically altering an oscil's
> waveform at perf time when ksmps != 1.The i-rate version would probably
> sound spectacular though if it was integrated with something like
> scantable.
>
> It might even be worth implementing something like "kconcat" or
> "iconcat" or even an optional i-arg to a generic "concat" to specify
> which rate to output at. To formulate the latter would make an
> excellent Csound 4 plugin tutorial since there isn't the added
> difficulty of dealing with complex datatypes... AND it sounds fun!
>
> Another possible solution to this concatenation idea could be to use
> the ftgen and then also arrange an array you wish to append to this
> existing f-table using fprintks (to output the whole mess, the f-table
> AND what you wish to concat) and then reading in the result with
> something like GEN 2 / 23 / 43. In this way you could have conditional
> statements in the .orc that would dynamically change the resultant
> tables. You could also output separate temp files for each table and
> then use GEN 23 / tablemix for something wild.
>
> Also, I think it would be neat to have the function of tablegpw as an
> optional k-arg to tablemix.
>
> ^^ $0.02 ... well spent!
>
>
> -David
>
> On Apr 27, 2005, at 9:52 PM, John Harrison wrote:
>
> > I do concatenation with tablemix, after David Akbari told me about
> > that opcode. It could be you could also use scanhammer, but the
> > documentation for scanhammer seems ambiguous to me (specifically which
> > table ipos refers to) and I have never used the opcode. So...for
> > concatenation you could make your destination table with an ftgen
> > function. Tablemix ifn1 into ifn, then tablemix ifn2 into ifn. Of
> > course, this isn't as elegent as a concat opcode would be, and you
> > can't do multiple tables on one line this way either.
> >
> > -John
> >
> > James Hearon wrote:
> >
> >>
> >>
> >> I was looking for a way to concatenate tables. It'd be neat to have
> >> something like:
> >> ifn concat size, ifn1, ifn2, etc.
> >>
> >>
> > -- [EMAIL REMOVED] > > (or to http://www.cs.bath.ac.uk/cgi-bin/csound ) [EMAIL REMOVED] > >
-- [EMAIL REMOVED] (or to http://www.cs.bath.ac.uk/cgi-bin/csound ) [EMAIL REMOVED]