Csound Csound-dev Csound-tekno Search About

[Csnd] GEN16 strangeness

Date2012-01-16 02:15
From"Art Hunkins"
Subject[Csnd] GEN16 strangeness
I do not understand a difference I'm experiencing with GEN16 between Csound 
floats (5.06) and doubles (5.15). What works with floats does not work with 
doubles. Can someone explain?

Basically, Csound floats allows for a GEN16 length of either a multiple of 2 
*or* multiple of 2 + 1. Doubles allows only for multiple of two. Therefore, 
a tab value of 1 (scaled) is possible with floats (using multiple of 2 + 1), 
but not with doubles.

The only compatible arrangement seems to be to use a table which is a 
multiple of 2, and a tab (max) value of .999.

If this is correct, the manual should specify that with doubles, GEN16 
*cannot use a multiple of 2 + 1*.

This is a problem for me because, for some time I've been working with 
floats, with tab values through 1 (along with tables that are multiples of 2 
+ 1), and *assuming* this works with doubles (a reasonable assumption I'd 
think). Ugh.

Or am I missing something?

Art Hunkins





-odac --displays -b128 -B2048




sr = 44100
ksmps = 100
nchnls  =       2

        instr 1

;k1      linseg   0, 4, 1, 2, 1, 4, 0; works with floats only (but only with 
1st f1 below - which is reasonable), not doubles
k1      linseg   0, 4, .999, 2, .999, 4, 0; required with doubles (also 
works with floats)
kscale  tab      k1, 1, 1
k2      tab      k1, 3, 1
aout    poscil   10000 * k1, 400, 2
a1,a2,a3,a4 pan aout, kscale, 1, 4, 1, 0
        outs     a1, a2

        endin




;f1 0 8193 16 0 8193 9 1; works with floats, not with doubles ("segmentation 
violation")
f1 0 8192 16 0 8192 9 1; works for doubles; for floats, only works with 2nd 
linseg above (which is reasonable)
f2 0 16384 10 1
;f3 0 8193 16 0 8193 .7 1; works with floats, not doubles
f3 0 8192 16 0 8192 .7 1; works for doubles
f4 0 8193 9 .25 1 0
i1 0 10


 



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date2012-01-16 07:54
Fromjpff@cs.bath.ac.uk
SubjectRe: [Csnd] GEN16 strangeness
not multiple of 2, power of 3

>
> Basically, Csound floats allows for a GEN16 length of either a multiple of
> 2
> *or* multiple of 2 + 1. Doubles allows only for multiple of two.
> Therefore,
> a tab value of 1 (scaled) is possible with floats (using multiple of 2 +
> 1),
> but not with doubles.




Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date2012-01-16 08:06
Fromjpff@cs.bath.ac.uk
SubjectRe: [Csnd] GEN16 strangeness
I do not get a crash but i do see
PERF ERROR in instr 1: tab off end

Using doubles and code adjusted that way

==John ff




Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date2012-01-16 08:57
FromTito Latini
SubjectRe: [Csnd] GEN16 strangeness
AttachmentsNone