Interesting.  I pretty much always use ksmps = 1 so I should be getting the
best timing possible.  Rounding errors look less likely to be the reason for
the discontinuity with that information.

Here is a sample csd file that demonstrates the problem while removing as
much extranious code as possible.  You will need to look at this in a wave
editor to see the breakpoint graphed out.

<CsoundSynthesizer>
<CsVersion>
</CsVersion>
<CsOptions>
-W -o "test.wav"
</CsOptions>
<CsInstruments>
sr        =    44100
ksmps    =    1
nchnls    =    1

instr 2
kout    linseg    p4, p3, p5
        out        a(kout) * 2000
endin
</CsInstruments>
<CsScore>
t    0    120

i2    8.875    0.375    .95    1.05
i2    9.25    3.625    1.05    1.1
i2    12.875    0.375    1.1    1.2
i2    13.25    4.375    1.2    1.3
i2    17.625    0.375    1.3    1.45
i2    18    2    1.72    2.42
i2    20    0.75    2.42    2.47
i2    20.75    0.25    2.47    2.7
i2    21    0.87    3.1    3.2
i2    21.87    0.13    3.2    3.33
i2    22    0.2    3.33    3.39
i2    22.2    1.8    3.39    3.45
i2    24    0.1    3.45    3.49
i2    24.1    0.9    3.49    3.6
i2    25    0.2    3.6        3.64
i2    25.2    1.6    3.64    3.7
i2    26.8    0.2    3.7        3.85
i2    27    0.52    3.94    4.2
i2    27.52    0.88    4.2        4.25
i2    28.4    0.6    4.25    4.6
</CsScore>
</CsoundSynthesizer>

You can see that not every breakpoint has the dip to zero, but some of them
do.  If you move the starting positions, however, most of the dips
disappear.  The following csd demonstrates this.

<CsoundSynthesizer>
<CsVersion>
</CsVersion>
<CsOptions>
-W -o "test.wav"
</CsOptions>
<CsInstruments>
sr        =    44100
ksmps    =    1
nchnls    =    1

instr 2
kout    linseg    p4, p3, p5

        out        a(kout) * 2000
endin
</CsInstruments>
<CsScore>
t    0    120

i2    .875    0.375    .95    1.05
i2    1.25    3.625    1.05    1.1
i2    4.875    0.375    1.1    1.2
i2    5.25    4.375    1.2    1.3
i2    9.625    0.375    1.3    1.45
i2    10    2    1.72    2.42
i2    12    0.75    2.42    2.47
i2    12.75    0.25    2.47    2.7
i2    13    0.87    3.1    3.2
i2    13.87    0.13    3.2    3.33
i2    14    0.2    3.33    3.39
i2    14.2    1.8    3.39    3.45
i2    16    0.1    3.45    3.49
i2    16.1    0.9    3.49    3.6
i2    17    0.2    3.6        3.64
i2    17.2    1.6    3.64    3.7
i2    18.8    0.2    3.7        3.85
i2    19    0.52    3.94    4.2
i2    19.52    0.88    4.2        4.25
i2    20.4    0.6    4.25    4.6
</CsScore>
</CsoundSynthesizer>

Thanks for the help.

On Thu, Feb 21, 2008 at 12:25 PM, <jpff@cs.bath.ac.uk> wrote:

> Yes, the resolution of score events, notes starting and stopping is always
> based on kr
> ==John ff
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>



-- 
Electronically,
Jeff Taylor