[Digikam-users] How does digiKam do raw conversions for the slide shows ?

[Linuxguy123]

On Wed, 2009-12-23 at 18:46 +0100, Gilles Caulier wrote:
> 2009/12/23 Linuxguy123 <linuxguy123 at gmail.com>:
> > On Wed, 2009-12-23 at 18:14 +0100, Gilles Caulier wrote:
> >> 2009/12/23 Mark Greenwood <fatgerman at ntlworld.com>:
> >> > On Wednesday 23 Dec 2009 16:59:45 Linuxguy123 wrote:
> >> >>
> >> >> How does digiKam generate the images for the raw files it displays in
> >> >> the slide shows ?   How could I easily generate jpegs that look as good
> >> >> as the images in the slideshow ?
> >> >>
> >> >
> >> > I would assume that for the slideshow Digikam is using the embedded
> >> JPEG data from the RAW file and not the RAW data. You can extract the
> >> embedded JPEG using exiftool (though I forget how, try googling it)
> >
> > Thanks for the tip.
> >
> >>  but that really defeats the object of using RAW. When you use RAW you
> >> are expected to do all the white balance etc adjustments yourself. The
> >> embedded JPEG has the camera's built-in adjustments already applied.
> >> If you're going to use the embedded JPEG you may as well just shoot
> >> JPEG in the first place.
> >
> > Thanks for the explanation about raws.  I never would have known that
> > from shooting over 20,000 of them and converting them to jpegs.
> >
> >> Yes, it use JPEG preview. Fast and usefull...
> >
> > Are you serious/ sure this is the case ?  I always thought the preview
> > jpegs were thumbnails.  I displayed the slideshow images on a 52" LCD TV
> > at 1920x1080.  There was no way they were thumbnails.
> >
> 
> YES. I'm serious !
> 
> RAW are containers. camera record a reduced version (and somethime the
> full version) of JPEG image processed at the same time than RAW file.
> Why ? to reder quicly the image on screen device and TV of course.
> 
> This is why preview in digiKam sound always better (in the first
> approach) than the decoded RAW : all the algorithm from camera are
> processed there...
> 
> This is not the case of RAW. You need to reproduce the camera job. And
> as each camera maker and model use different algorithm, it's a worse !

Very interesting discussion.  I've learned a lot this morning. Thanks to
everyone who contributed.