| The best solution is to drop the float build and keep the doubles build. Performance lost in this way should be more than regained by completing work on multi-threading Csound, and other optimizations. On Windows, at any rate, the performance gap is not very large. I will get back to the list with accurate comparisons.
Don't worry, this is not something I will (or can) do unilaterally. It should be based on research.
Regards,
Mike
-----Original Message-----
>From: Victor Lazzarini
>Sent: Oct 20, 2008 1:12 PM
>To: Developer discussions
>Subject: Re: [Cs-dev] csound 64-bit API
>
>I agree this perhaps needs a better solution. As for the OSX framework,
>I have not yet decided which way I am going to package the doubles
>build.
>
>Regards
>
>Victor
>
>At 15:50 20/10/2008, you wrote:
>>Oh, I didn't realize that I must set USE_DOUBLE also on my own sources.
>>I assumed that MYFLT would be defined as double if csound was compiled
>>with useDouble=1. Why else export MYFLT to the public API?
>>
>>The scons build script could easily write a simple "#define MYFLT
>>double" to myflt.h and let csound.h include it. Then apps using the
>>csound API could just relax and use MYFLT and it would just work. Now I
>>need to check if csound was built with double precision by checking if
>>the library name is "csound" or "csound64" which is kind of ugly. And
>>does this work on MacOSX framework? Is there any other way to check if
>>csound is float or double?
>>
>>Victor Lazzarini wrote:
>> > That is because MYFLT is set to be float not double. You will need
>> > -DUSE_DOUBLE to set MYFLT to double. The insert_score_event() is
>> > basically getting garbage as the pointer sizes don't match.
>> >
>> > Victor
>> >
>> >> But now I got another problem, I use code like this to send score events
>> >>from AlgoScore to Csound:
>> >> MYFLT *parms;
>> >> int i;
>> >> naRef vec = VECARG(2);
>> >> int sz = naVec_size(vec);
>> >> parms = malloc(sizeof(MYFLT)*sz);
>> >> for(i=0;i> >> parms[i]=naNumValue(naVec_get(vec,i)).num;
>> >> csoundScoreEvent(cs, 'i', parms, sz);
>> >>
>> >> But I get "insert_score_event(): invalid instrument number or name" from
>> >> csound. This only happens with the double version and works fine with
>> >> the float version of csound!
>> >>
>> >> --
>> >> /Jonatan [ http://kymatica.com ]
>>
>>--
>>/Jonatan [ http://kymatica.com ]
>>
>>-------------------------------------------------------------------------
>>This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
>>Build the coolest Linux based applications with Moblin SDK & win great prizes
>>Grand prize is a trip for two to an Open Source event anywhere in the world
>>http://moblin-contest.org/redirect.php?banner_id=100&url=/
>>_______________________________________________
>>Csound-devel mailing list
>>Csound-devel@lists.sourceforge.net
>>https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>Victor Lazzarini
>Music Technology Laboratory
>Music Department
>National University of Ireland, Maynooth
>
>
>-------------------------------------------------------------------------
>This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
>Build the coolest Linux based applications with Moblin SDK & win great prizes
>Grand prize is a trip for two to an Open Source event anywhere in the world
>http://moblin-contest.org/redirect.php?banner_id=100&url=/
>_______________________________________________
>Csound-devel mailing list
>Csound-devel@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/csound-devel
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net |