Thanks Oeyvind.
I had thought of this, but I didn't realize the loop points could
change at k-rate, and I wasn't sure why a start point would be
necessary.  It all makes sense now.
These k-rate or a-rate changes, would that be through table writing
opcodes?  In that case, any of the values could change, no?  Although
it would seem far more efficient to have the values all loaded and
only change the start and end values.

-Chuckk

On Thu, Apr 3, 2008 at 12:45 PM, Oeyvind Brandtsegg <obrandts@gmail.com> wrote:
> Ah, Good question,
>  I did not anticipate the manual coud be understood in this manner.
>
>  Yes, for 19 sets of values (19 mix "points"), you'd need 97 values in the table,
>  and you would leave the rest of the ftable values empty (zero, or
>  whatever, does not matter, they will not be read). As you correctly
>  assume, we would still use power-of-two size tables, so some empty
>  space will be allocated.
>
>  Regarding the loop points:
>  I see that you transfer the knowledge about loop points from audio
>  sample playback into thinking about partikkel ftables reading. Maybe I
>  should have used aother terms to make the poeration more clear:
>  The loop start point is actually the "read start point" nothing before
>  this point (index) is read, and the loop end point could be termed the
>  "read end point", nothing beyond this point will be read.
>  When a note starts, the first value read is the one at the loop start
>  point, reading continues on a per grain basis until the loop end point
>  and then wraps around to the loop start point. Any release segment of
>  a not event does not change this behaviour, the looping continues as
>  for a held note.
>  The loop start and end points can be modified at k (or even a-) rate,
>  with the changes being effected at the next grain. The loop period
>  will play to the (possibly new) end point before wrapping to the
>  (possibly new) start point. When changing values, if the read pointer
>  is outside of the (newly changed) loop points, it will read the next
>  value from the loop start point.
>
>  best
>  Oeyvind
>
>  2008/4/3, Chuckk Hubbard <badmuthahubbard@gmail.com>:
>
>
> > Hi.
>  > iwaveamptab reads ftables in clumps of 5 values, the first two places
>  > of the table being reserved for marking loop start and end points.
>  > So what happens if I want to have, say, 19 sets of values?  I'd need
>  > 97 places in a table; what do I do with the other 31?
>  > I wonder the same for the other tables that are read the same way, if
>  > I want, say, 24 values, and I have to use a size 32 table?
>  >
>  > I take it that the loop executes in the middle of the note; that
>  > anything before the loop plays exactly once at the start, and anything
>  > after the loop exactly once at the end?  So is there a value to use
>  > that indicates to ignore this part of the table?
>  >
>  > Thanks.
>  > -Chuckk
>  >
>  > --
>  > http://www.badmuthahubbard.com
>  >
>  >
>  > Send bugs reports to this list.
>  > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>  >
>
>
>  Send bugs reports to this list.
>  To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>



-- 
http://www.badmuthahubbard.com