[prev in list] [next in list] [prev in thread] [next in thread] 

List:       imagemagick-user
Subject:    Re: [magick-users] Fw:  -compose
From:       Anthony Thyssen <A.Thyssen () griffith ! edu ! au>
Date:       2008-12-29 2:13:43
Message-ID: 20081229121343.27bf453f () wraith
[Download RAW message or body]

On Sun, 28 Dec 2008 18:03:06 -0800
Glenn Linderman <perl@NevCal.com> wrote:

| On approximately 12/28/2008 5:33 PM, came the following characters from 
| the keyboard of Anthony Thyssen:
| > On Fri, 26 Dec 2008 23:11:32 -0800
| > magick-users-bounces@imagemagick.org wrote:
| > 
| > | So I was doing a book layout, putting small pages on a big page.
| > | 
| > | The idea is that page fragments, formatted to fit small page sizes 
| > | already, or easily cropped from something containing it, would be placed 
| > | in various spots on the big page.
| > | 
| > | Some pages are to be composed of multiple pages.  Note c004.tif and 
| > | c005.tif below... they are both contributing to the same small page, but 
| > | do not both contain black stuff at the same places.
| > | 
| > | All is bitonal black and white, so it seems like -compose multiply would 
| > | let one merge these various pages, and it seems to work in some smaller 
| > | cases I had done, where I build all the pages separately and merge them, 
| > | but then that took lots of memory, so I thought I'd -flatten as I go.... 
| > | but then c004.tif doesn't get included... apparently it gets overlaid by 
| > | c005.tif, even the white parts.
| > | 
| > | Does -compose only persist for one -flatten?  I thought as a setting, it 
| > | would persist until it was changed.
| > | 
| > The problem is that -compose is for some file formats a image meta-data
| > setting, as so reading an image overrides the existing setting.
| > 
| > I myself have had problems with this, and I think these problem will persist
| > until Image magick as a formal (internal) notion of a 'global' override
| > verses the image meta-data value.
| > 
| > My own thinking is that global values should override any and all meta-data
| > values unless left unset or the user purposefully unsets the global override
| > using +value.
| > 
| > Similarly if the user wants to specifically set the image meta data value the
| > -set operator should be used.
| > 
| > However IM has no true notion of a global override, verses image meta-data.
| > Compose in particular is only seen by the alpha composition operator simply as
| > image meta-data.
| > 
| > Until the library is changed to a data-structure that consists of
| > an 'image'  with pointers to 'image-meta-data' and 'global-overrides',
| > with library routines to get the appropriate settings. I doubt the confusion
| > will have any resolution anytime soon.
| 
| 
| Thanks.  I certainly hadn't reached that level of understanding of IM 
| operators and settings, and had never heard of meta-data setting.
| 
| So the safest thing for a na__ve user is to specify input file, then 
| settings, then operators, and not to assume that settings persist to the 
| next operator.  That sort of bloats the command line, but if it works 
| consistently... that is better than being surprised, I guess.
| 
It is a pain yes. and ideally if compose setting has been set then it
should remain set regardless of if images have been read in, or
parenthesis, image re-ordering or anything else is done.  But currently
that is not happening.

You could put in a BUG report in the IM Bugs Discussion Forum, so that if
-compose has been set it overrides any compose meta-data present in the
input image.  This is currently what is happening for -dispose and
-delay for animations, and would make it more user friendly.

Only a few file formats actually define a compose setting as part of
the images meta-data (settings associated with the image itself), TIFF
just happens to be one of them.

  Anthony Thyssen ( System Programmer )    <A.Thyssen@griffith.edu.au>
 -----------------------------------------------------------------------------
  ``Probable impossibilites are to be preferred to improbable possibilites''
                                                               -- Aristotle
 -----------------------------------------------------------------------------
     Anthony's Home is his Castle     http://www.cit.gu.edu.au/~anthony/
_______________________________________________
Magick-users mailing list
Magick-users@imagemagick.org
http://studio.imagemagick.org/mailman/listinfo/magick-users
[prev in list] [next in list] [prev in thread] [next in thread]