ScoreEvent() or InputMessage() are the alternatives. They do not do any score preprocessing, which ReadScore() does.
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 2 Jun 2014, at 21:27, David Worrall  wrote:

> hi all,
> 
> I preprocess very large chunks of data to score events and the pass the string through to a cs.PerfThread()
> (cs = csnd6.Csound()
> 
> The RT performance eventually stutters/splutters into action and, after the initial startup phase, functions flawlessly (no dropouts etc)
> 
> It sounds to me like by the time cs gets the events to process them, there are multiple events already 'behind schedule'.
> This is confirmed by the effect disappearing on adding 15 secs to the start time of all events.
> I'm interested in how to minimize this time. I've tried playing with various buffer sizes to no effect.
> ('bleeding in' n events in with a time.sleep() between, to no effect, which is consistent with the above)
> 
> I tried using cs.Event() (also checked for cs.event) but I get a swig "not found" error (OSX, cs 6.3.02)
> 
> Where does csPerf.Play() get its timing from? I tried executing it at the last moment (and using csPerf.Pause()) both to no effect.
> Does csPerf.Play() start counting from when it is first executed or when it receives its first score event?
> 
> So many questions, sorry, and if I have to live with a 15 sec delay on execution I can in this instance, but it would be good to hear people's strategies for circumstances like these.
> 
> thanks,
> 
> David
> 
> ______________________________________
> Prof. Dr. David Worrall
> Emerging Audio Research (EAR)
> Audio Department
> International Audio Laboratories Erlangen
> Fraunhofer-Institut für Integrierte Schaltungen IIS
> Am Wolfsmantel 33
> 91058 Erlangen
> E-Mail: david.worrall@iis.fraunhofer.de
> Internet: www.iis.fraunhofer.de 
> ---
> Adjunct Senior Research Fellow
> School of Music, Australian National University
> david.worrall@anu.edu.au
> 
> 
> 
> 

Thanks Victor. These might solve the problem if it was one of time spent in sort routines. But I am not sure that is the issue, which is the time i need between turning the Perf. thread on and sending the first record.

Where is the best up-to-date on-line reference to these (python/API) functions these days? Google didn't help much.
[The Csound journal http://www.csounds.com/journal/issue14/realtimeCsoundPython.html? is useful but contains rough html and not complete]

D

On 02.06.2014, at 23:00, Victor Lazzarini  wrote:

> ScoreEvent() or InputMessage() are the alternatives. They do not do any score preprocessing, which ReadScore() does.
> ========================
> Dr Victor Lazzarini
> Senior Lecturer
> NUI Maynooth, Ireland
> victor dot lazzarini at nuim dot ie
> 
> 
> 
> 
> On 2 Jun 2014, at 21:27, David Worrall  wrote:
> 
>> hi all,
>> 
>> I preprocess very large chunks of data to score events and the pass the string through to a cs.PerfThread()
>> (cs = csnd6.Csound()
>> 
>> The RT performance eventually stutters/splutters into action and, after the initial startup phase, functions flawlessly (no dropouts etc)
>> 
>> It sounds to me like by the time cs gets the events to process them, there are multiple events already 'behind schedule'.
>> This is confirmed by the effect disappearing on adding 15 secs to the start time of all events.
>> I'm interested in how to minimize this time. I've tried playing with various buffer sizes to no effect.
>> ('bleeding in' n events in with a time.sleep() between, to no effect, which is consistent with the above)
>> 
>> I tried using cs.Event() (also checked for cs.event) but I get a swig "not found" error (OSX, cs 6.3.02)
>> 
>> Where does csPerf.Play() get its timing from? I tried executing it at the last moment (and using csPerf.Pause()) both to no effect.
>> Does csPerf.Play() start counting from when it is first executed or when it receives its first score event?
>> 
>> So many questions, sorry, and if I have to live with a 15 sec delay on execution I can in this instance, but it would be good to hear people's strategies for circumstances like these.
>> 
>> thanks,
>> 
>> David
>> 
>> ______________________________________
>> Prof. Dr. David Worrall
>> Emerging Audio Research (EAR)
>> Audio Department
>> International Audio Laboratories Erlangen
>> Fraunhofer-Institut für Integrierte Schaltungen IIS
>> Am Wolfsmantel 33
>> 91058 Erlangen
>> E-Mail: david.worrall@iis.fraunhofer.de
>> Internet: www.iis.fraunhofer.de 
>> ---
>> Adjunct Senior Research Fellow
>> School of Music, Australian National University
>> david.worrall@anu.edu.au

You can send the score lines even before you turn on the perf thread. They will be executed as soon as it starts (provided p2 is 0 of course).
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 2 Jun 2014, at 23:09, David Worrall  wrote:

> Thanks Victor. These might solve the problem if it was one of time spent in sort routines. But I am not sure that is the issue, which is the time i need between turning the Perf. thread on and sending the first record.
> 
> Where is the best up-to-date on-line reference to these (python/API) functions these days? Google didn't help much.
> [The Csound journal http://www.csounds.com/journal/issue14/realtimeCsoundPython.html? is useful but contains rough html and not complete]
> 
> D
> 
> On 02.06.2014, at 23:00, Victor Lazzarini  wrote:
> 
>> ScoreEvent() or InputMessage() are the alternatives. They do not do any score preprocessing, which ReadScore() does.
>> ========================
>> Dr Victor Lazzarini
>> Senior Lecturer
>> NUI Maynooth, Ireland
>> victor dot lazzarini at nuim dot ie
>> 
>> 
>> 
>> 
>> On 2 Jun 2014, at 21:27, David Worrall  wrote:
>> 
>>> hi all,
>>> 
>>> I preprocess very large chunks of data to score events and the pass the string through to a cs.PerfThread()
>>> (cs = csnd6.Csound()
>>> 
>>> The RT performance eventually stutters/splutters into action and, after the initial startup phase, functions flawlessly (no dropouts etc)
>>> 
>>> It sounds to me like by the time cs gets the events to process them, there are multiple events already 'behind schedule'.
>>> This is confirmed by the effect disappearing on adding 15 secs to the start time of all events.
>>> I'm interested in how to minimize this time. I've tried playing with various buffer sizes to no effect.
>>> ('bleeding in' n events in with a time.sleep() between, to no effect, which is consistent with the above)
>>> 
>>> I tried using cs.Event() (also checked for cs.event) but I get a swig "not found" error (OSX, cs 6.3.02)
>>> 
>>> Where does csPerf.Play() get its timing from? I tried executing it at the last moment (and using csPerf.Pause()) both to no effect.
>>> Does csPerf.Play() start counting from when it is first executed or when it receives its first score event?
>>> 
>>> So many questions, sorry, and if I have to live with a 15 sec delay on execution I can in this instance, but it would be good to hear people's strategies for circumstances like these.
>>> 
>>> thanks,
>>> 
>>> David
>>> 
>>> ______________________________________
>>> Prof. Dr. David Worrall
>>> Emerging Audio Research (EAR)
>>> Audio Department
>>> International Audio Laboratories Erlangen
>>> Fraunhofer-Institut für Integrierte Schaltungen IIS
>>> Am Wolfsmantel 33
>>> 91058 Erlangen
>>> E-Mail: david.worrall@iis.fraunhofer.de
>>> Internet: www.iis.fraunhofer.de 
>>> ---
>>> Adjunct Senior Research Fellow
>>> School of Music, Australian National University
>>> david.worrall@anu.edu.au
> 
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
>