AW: AW: [Clam-devel] [PATCH] DEBUG-related, PortAudio, SinTracking, AudioFileFormats, SConstruct

[Rettenbacher, Bernhard]

You have already created something, that needs double: The FreewheelingNetworkPlayer.

Cheers,

Bernhard

> -----Ursprüngliche Nachricht-----
> Von: clam-devel-bounces at llistes.projectes.lafarga.org 
> [mailto:clam-devel-bounces at llistes.projectes.lafarga.org] Im 
> Auftrag von Pau Arumi
> Gesendet: Donnerstag, 14. Juni 2007 17:37
> An: clam-devel at llistes.projectes.lafarga.org
> Betreff: Re: AW: [Clam-devel] [PATCH] DEBUG-related, 
> PortAudio, SinTracking,AudioFileFormats, SConstruct
> 
> En/na Rettenbacher, Bernhard ha escrit:
> > Hi Pau,
> > 
> > AudioSource/Sink patch:
> > There is indeed an error in audiosource.cxx. I used 
> mDoubleBuffer for float and double TData. So I'll send you an 
> updated patch.
> 
> Your last audiosource.patch no been commited (actually it was 
> but immediately reverted). Reason below.
> 
> > Q: Why do you use different members for double and float 
> versions and not a TData? (I didn't go into the code very 
> deep, so maybe this is a trivial question).
> 
> Because we need to have buffers for the callbacks. For example in
> CLAM/src/Flow/Networks/JACKNetworkPlayer.cxx:172:
> it->source->SetExternalBuffer( jackInBuffer, nframes );
> 
> But since it should work with different back-end architectures we
> defined two overloaded functions
> 
> void SetExternalBuffer(float* buf, unsigned nframes );
> void SetExternalBuffer(double* buf, unsigned nframes );
> 
> Then the intended behaviour of AudioSource::Do() is:
> find which one of the two buffers is used and use it as a source
> for the copy to the audio-out-port.
> 
> So, choosing the float or double buffer is not related to Clam's
> DOUBLE define and TData type.
> 
> However, what AudioSource::Do() currently does is asserting that
> the buffer is float. As said, it should conditionally choose one or
> another. I think I prefer to leave is this way till some callback
> architecture that needs double is used.
> 
> Cheers
> pau
> 
> 
> 
> > AudioFileFormats patch:
> > I really do not know the exact cause for changing this (it 
> was 2 months ago, I did it ).
> > One cause I remember is the different number of entries in 
> the extensionMap struct (19) and the number of entries in 
> EAudioFileFormat (16). There was an error when accessing 
> extensionMap elements > 15.
> > 
> > Bye,
> > 
> > Bernhard
> > 
> >> -----Ursprüngliche Nachricht-----
> >> Von: clam-devel-bounces at llistes.projectes.lafarga.org 
> >> [mailto:clam-devel-bounces at llistes.projectes.lafarga.org] Im 
> >> Auftrag von Pau Arumi
> >> Gesendet: Mittwoch, 13. Juni 2007 18:48
> >> An: clam-devel at llistes.projectes.lafarga.org
> >> Betreff: Re: [Clam-devel] [PATCH] DEBUG-related, PortAudio, 
> >> SinTracking,AudioFileFormats, SConstruct
> >>
> >> En/na Rettenbacher, Bernhard ha escrit:
> >>> Dear CLAM-Team,
> >>>
> >>> Tuesday is "Patch Day"! 
> >>>
> >>> I send you some patches for the following:
> >>> * Some pieces of CLAM code seems to be tested only in 
> RELEASE mode 
> >>> =>Array-Patch =>AudioFileFormats-Patch
> >> These (including audiosink and audiosource patches) have NOT 
> >> been commited yet. audiosource/sink patch causes segfault 
> >> here (linux ubuntu feisty, default configuration). I'll look 
> >> the details tomorrow.
> >>
> >> And don't understand the reason behind AudioFileFormats 
> >> patch. Can you explain?
> >>
> >>
> >>> * You only partially migrated to PortAudio V19 => Pa*-Patches, 
> >>> PortAudioUtils-Patch
> >> Commited
> >>
> >>> *SinTracking gives an error, when no peak is found (eg. 
> for digital 
> >>> zero streams ).
> >>> =>SinTracking-Patch
> >> Commited
> >>
> >>> *Building CLAM for Windows results does not work, if you do not 
> >>> specify an audio architecture for rtaudio.
> >>> => Sconstruct-Patch (Workaround by specifying an 
> architecture-flag)
> >> Commited
> >>
> >> * Revision 10226
> >>
> >> Thanks a lot! I hope the patches will keep coming :)
> >>
> >> pau
> >>
> >> _______________________________________________
> >> Clam-devel mailing list
> >> Clam-devel at llistes.projectes.lafarga.org
> >> https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo
> > /clam-devel
> >>
> >> 
> --------------------------------------------------------------
> ----------
> >>
> >> _______________________________________________
> >> Clam-devel mailing list
> >> Clam-devel at llistes.projectes.lafarga.org
> >> 
> https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo
/clam-devel
> 
> 
> _______________________________________________
> Clam-devel mailing list
> Clam-devel at llistes.projectes.lafarga.org
> https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo
/clam-devel
>