[Clam-devel] SDIFIn question.

Xavier Amatriain xavier at create.ucsb.edu
Fri May 4 16:27:33 PDT 2007


I did work on this classes but that was a long time ago. I can't give
an answer without actually diving far, far away into my memory ;)

>From reading Giulio's email and having a quick look at the code I would
say that he basically caught a bug: the expected behavior should be the
same regardless of what you have loaded before. I am all for applying
the patch if it does not break any test.

XA

On Fri, 2007-05-04 at 12:08 +0200, Pau Arumi wrote:
> Hi Giulio,
> 
> sorry for letting this question unanswered. From my part I don't
> have experience with the use of SDIF in Clam. Maybe Xavier have
> an answer to this?
> 
> However, since one of Google of Code projects I'm mentoring deal
> with sms synthesizers, it's likely we're gonna dive into it in
> the next few months and probably clarify and improve the SDIF
> support.
> 
> Best,
> Pau
> 
> 
> En/na Giulio Paci ha escrit:
> > Hi to all,
> >     some day ago I had modified some code that loads some SDIF files
> > using the SDIFIn reader.
> >     Our program was using the same reader, with the same (default)
> > options to load all the SDIF files. I decided to change this behavior
> > and to use a new SDIFIn reader for each SDIF file.
> > 
> >     I was surprised that the output of my program (wav file, synthesized
> > from loaded segments) changed. So I investigated the SDIFIn reader
> > and I found that LoadSDIFDataIntoSegment() (called from Do()) changes
> > mLastCenterTime and mPrevIndexArray.
> > 
> > Stop() doesn't clean these fields. I was wondering if this is the desired
> > behavior, 'cause I don't know what this implies (And why we may want
> > to load segment in different ways, depending on what we have already
> > read).
> > 
> > I attached a patch that make Stop() cleaning these two fields too. If the
> > actual behavior is not the intended one.
> > 
> > Any way I would be very happy to know how SDIF reading should work
> > and why.
> > 
> > Cheers,
> >     Giulio.
> > 
> > 
> > ------------------------------------------------------------------------
> > 
> > Index: src/Processing/SDIFIO/SDIFIn.cxx
> > ===================================================================
> > --- src/Processing/SDIFIO/SDIFIn.cxx	(revisione 10063)
> > +++ src/Processing/SDIFIO/SDIFIn.cxx	(copia locale)
> > @@ -108,6 +108,8 @@
> >  bool SDIFIn::ConcreteStop()
> >  {
> >  	mpFile->Close();
> > +	mLastCenterTime=-1;
> > +	mPrevIndexArray.Resize(0);
> >  
> >  	return true;
> >  }
> > 
> > 
> > ------------------------------------------------------------------------
> > 
> > _______________________________________________
> > 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
-- 
/*********************************
 *       Xavier Amatriain        *
 *  Associate Director - MATi    *
 *  Research Director - CREATE   *
 *    UCSB, Santa Barbara CA     *
 *      1-(805)- 893 83 52       *
 ********************************/





More information about the clam-devel mailing list