AWARE [SYSTEMS] Imaging expertise for the Delphi developer
AWare Systems, Imaging expertise for the Delphi developer, Home TIFF and LibTiff Mailing List Archive

LibTiff Mailing List

TIFF and LibTiff Mailing List Archive
October 2000

Previous Thread
Next Thread

Previous by Thread
Next by Thread

Previous by Date
Next by Date

Contact

The TIFF Mailing List Homepage
This list is run by Frank Warmerdam
Archive maintained by AWare Systems



Valid HTML 4.01!



Thread

2000.10.03 20:44 "Varying BitsPerSample", by Rick Lamont
2000.10.05 04:37 "Re: Varying BitsPerSample", by Frank Warmerdam
2000.10.05 04:58 "Re: Varying BitsPerSample", by Sam Leffler
2000.10.05 05:26 "Re: Varying BitsPerSample", by Sam Leffler

2000.10.05 04:37 "Re: Varying BitsPerSample", by Frank Warmerdam

Rick LaMont wrote:
> 
> There was a discussion on a newsgroup recently about having TIFF files
> with different data formats per channel.  For example, 8-bit RGB with
> a 32-bit Z channel.  Some thought that the TIFF format does not allow
> this but then Derek Kusiak wrote:
> 
> > Unless I'm misunderstanding what the TIFF 6.0 spec says, that's not
> > true:
> >
> > "Note that this field [BitsPerSample] allows a different number of bits
> > per component for each component corresponding to a pixel. For example,
> > RGB color data could use a different number of bits per component for
> > each of the three color planes." (p.29)
> 
> Is there a limitation in libtiff that prevents exploitation of this
> section of the spec?
> 
> I'm interested in interpretations of the spec, pros/cons of this feature,
> limitations and workarounds.

Rick, 

I didn't trace this through completely, but my quick read of the situation
is that the TIFFDirectory structure only has a single element for BITSPERSAMPLE,
SAMPLEFORMAT, MINSAMPLEVALUE and MAXSAMPLEVALUE, all of which should be
per-sample, not singular.  The writing interface seems to duplicate this
number out to the number of samples, and the reading interface seems to work
whether there is one or more values, but just captures the first into the
TIFFDirectory. 

I would really like to see some of these (in particular the MINSAMPLEVALUE
and MAXSAMPLEVALUE) handled on a per-sample in libtiff.  It would also
be nice if the types of each sample could be controlled, though I think this
would result in datasets that work properly with very few packages.  It 
might be quite difficult to even get the libtiff tools programs to support
this fully. 

In short, I am willing to work with you to make libtiff handle this better,
but I think it unwise to generate files like this if you actually want them
to work with any other software. 

I too would be interested in more educated opinions.

Best regards,

---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerda@home.com
light and sound - activate the windows | http://members.home.com/warmerda
and watch the world go round - Rush    | Geospatial Programmer for Rent