In discussions with the GraphicsMagick developers, I believe we isolated
with imwrite not matching imread to Octave as all of the tests we performed
in Graphics Magick work correctly. Reading and writing images produced
identical results
The problem extends to PNG, BMP and Tiff formats atleast.
The comments from Bob Friesenhahn are as follows.
Properties which may influence the result are 'depth' and 'storage_class'
(classType in Magick++). If you read 8-bit data into an array then the
writer needs to know that the array data is 8-bit (and not 16-bit).
("your earlier postings of the problem description") causes me to suspect
that perhaps Octave has a bug in that it stores the low-order octet of a
16-bit gray value rather than dividing, or storing the high-order octet.
This would cause considerable confusion since the image would often look
almost correct (the low-order byte often matches the high-order byte), but
sometimes be terribly wrong.
-----Original Message-----
From: Benjamin Lindner [mailto:[email protected]]
Sent: Friday, April 16, 2010 12:29 PM
To: Søren Hauberg
Cc: Ian Journeaux; [email protected]
Subject: Re: [OctDev] Reading tiff's [was: Re: Latest Binaries]
Søren Hauberg wrote:
> fre, 16 04 2010 kl. 19:14 +0200, skrev Benjamin Lindner:
>>> The driver here was not that I wanted to upgrade my version of Octave,
it
>>> was changing the version of GraphicsMagick. There is a bug in imwrite
where
>>> a tiff file read in with imread is modified slightly when it is written
out
>>> by imwrite.
>> Am I understanding this correctly: there is a bug in graphicsmagick
>> (version 1.3.7) which is fixed in newer versions?
>> In other words, the bundled version of graphicsmagick should be updated?
>
> Honestly, I don't know (sorry). Perhaps others know?
>
> I do know that more recent versions of GraphicsMagick are causing
> problem by changing the locale. This screws up functions relating to
> date and time. So, I am not sure if you really want to upgrade the
> GraphicsMagick version.
And there have been reports on segfaults due to problems with proper
initialization. So I kept graphicsmagick at version 1.3.7...
benjamin
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Octave-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/octave-dev
