Sign up now! · Forgot password?
RSS/Atom feed identi.ca Twitter

save 32 bit float as exr file

This discussion is connected to the gegl-developer-list.gnome.org mailing list which is provided by the GIMP developers and not related to gimpusers.com.

This is a read-only list on gimpusers.com so this discussion thread is read-only, too.

4 of 4 messages available
Toggle history

Please log in to manage your subscriptions.

save 32 bit float as exr file Rasmus Hahn 18 Jul 17:13
  save 32 bit float as exr file Piotr Stanczyk 18 Jul 17:45
   save 32 bit float as exr file " 18 Jul 20:17
  save 32 bit float as exr file " 18 Jul 20:13
Rasmus Hahn
2011-07-18 17:13:34 UTC (over 2 years ago)

save 32 bit float as exr file

Hi,

i have written an operation for gegl that writes OpenEXR float images, since i noticed there does not seem to be one already and i could use it. This version is quite usable within it's limitations. Limitations currently are:

- writes only 32 bit linear floating point images. - only writes Y or RGB images with optional alpha. - does not make use of OpenEXR specialties like layers, mipmaps and multichannel.

It does support the tiled format; for simplicity, however, tiles are always square.

I would like to see it included in the distribution; for now i have attached a patch, which creates a new file gegl/operations/external/exr-save.cc and makes a modification to the Makefile.am to build it. The patch is against version gegl-0.1.6, but works with the git-version (as of 2011-07-13), too. Sorry, for not providing a URL, but i do not have one right now.

The built operator is named 'gegl:exr-save' and has two properties: path (string) - the filename to write to tile (int) - the width of a tile (used for both x and y size); the default is 0, which means a scanline file is written.

Any comments or suggestions for this?

Greetings - Rasmus

Piotr Stanczyk
2011-07-18 17:45:19 UTC (over 2 years ago)

save 32 bit float as exr file

Thanks for adding ... Any thoughts on adding 16-bit floating point support? You could then consider using the RGBA interface ...

On Mon, Jul 18, 2011 at 10:13 AM, Rasmus Hahn wrote:

Hi,

i have written an operation for gegl that writes OpenEXR float images, since i noticed there does not seem to be one already and i could use it. This version is quite usable within it's limitations. Limitations currently are:

 - writes only 32 bit linear floating point images.  - only writes Y or RGB images with optional alpha.  - does not make use of OpenEXR specialties like layers, mipmaps    and multichannel.

It does support the tiled format; for simplicity, however, tiles are always square.

I would like to see it included in the distribution; for now i have attached a patch, which creates a new file gegl/operations/external/exr-save.cc and makes a modification to the Makefile.am to build it. The patch is against version gegl-0.1.6, but works with the git-version (as of 2011-07-13), too. Sorry, for not providing a URL, but i do not have one right now.

The built operator is named 'gegl:exr-save' and has two properties:  path (string) - the filename to write to  tile (int) - the width of a tile (used for both x and y size);    the default is 0, which means a scanline file is written.

Any comments or suggestions for this?

Greetings - Rasmus

_______________________________________________ Gegl-developer mailing list
Gegl-developer@lists.XCF.Berkeley.EDU https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer

Gegl-developer mailing list
Gegl-developer@lists.XCF.Berkeley.EDU
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer
"
2011-07-18 20:13:45 UTC (over 2 years ago)

save 32 bit float as exr file

On Mon, Jul 18, 2011 at 6:13 PM, Rasmus Hahn wrote:

i have written an operation for gegl that writes OpenEXR float images, since i noticed there does not seem to be one already and i could use it. This version is quite usable within it's limitations. Limitations currently are:

Thank you :) I am not sure if the current way the format of the buffer is determined is the most robust, but it probably works well enough. I added the op to master and made the coding style more consistent with the rest of GEGL/GIMP.

I would like to see it included in the distribution; for now i have attached a patch, which creates a new file gegl/operations/external/exr-save.cc and makes a modification to the Makefile.am to build it. The patch is against version gegl-0.1.6, but works with the git-version (as of 2011-07-13), too. Sorry, for not providing a URL, but i do not have one right now.

---
Author: Øyvind Kolås
Date: Mon Jul 18 21:08:56 2011 +0100

AUTHORS: add Rasmus Hahn

commit 5470164d7a8677d9ea1ea67282650476b4d17aa4 Author: Øyvind Kolås
Date: Mon Jul 18 21:08:30 2011 +0100

exr-save: normalize coding style

commit d9512a262c845bb59f47ba103c1b7a32790ccf37 Author: Øyvind Kolås
Date: Mon Jul 18 20:56:50 2011 +0100

exr-save: add new op from Rasmus Hahn
----

It would have been slightly nicer if you used git format-patch and included a commit message for the patch yourself. Thank you :)

/Øyvind K.

«The future is already here. It's just not very evenly distributed»
                                                 -- William Gibson
http://pippin.gimp.org/                            http://ffii.org/
_______________________________________________
Gegl-developer mailing list
Gegl-developer@lists.XCF.Berkeley.EDU
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer
"
2011-07-18 20:17:10 UTC (over 2 years ago)

save 32 bit float as exr file

On Mon, Jul 18, 2011 at 6:45 PM, Piotr Stanczyk wrote:

Thanks for adding ...  Any thoughts on adding 16-bit floating point support?  You could then consider using the RGBA interface ...

The op already supported Y, YA, RGB and RGBA - but all only with 32 bit floating point support. If support for half floats were to be added this would have to be added to babl, either as an extension or to the base of babl (the latter probably makes most sense). Looking at the additional data types added for CIE Lab could act as inspiration.

I do however think having support for half-float exr is less important than 32bit float so this is a great start in itself :)

/Øyvind K.

«The future is already here. It's just not very evenly distributed»
                                                 -- William Gibson
http://pippin.gimp.org/                            http://ffii.org/
_______________________________________________
Gegl-developer mailing list
Gegl-developer@lists.XCF.Berkeley.EDU
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer