32-bit float format for gimp 2.6

[DesertBroncoFan]

I have some raw image data (not the RAW format) that is composed 32-bit floats that I need to find the differences in.  I'm thinking a simple subtraction would be sufficient and I'm able to do that by simply subtracting the images pixel by pixel but I would like to be able to display the resulting image for management. 

Is there a simple format for 32-bit floats that I can transform my data into that GIMP 2.6 can read?  Is there some way that I can tell GIMP where the data is and the dimensions of the images?

[rich2005]

Linux and Gimp 2.6? could be Debian 6 or a Fedora from about 2011.

Honestly do not know, I would ask the question on the ImageMagick forum https://www.imagemagick.org/discourse-server/ Users section.

Give details of your OS and the available version of IM for your distro. AFAIK you will need a version of IM that uses HDRI.

...but.. that is just my opinion, keep looking here for a sensible answer.

(edit: wouldn't be a sound file by any chance)

[DesertBroncoFan]

(12-03-2018, 03:48 PM)rich2005 Wrote: Linux and Gimp 2.6? could be Debian 6 or a Fedora from about 2011.

Honestly do not know, I would ask the question on the ImageMagick forum https://www.imagemagick.org/discourse-server/ Users section.

Give details of your OS and the available version of IM for your distro. AFAIK you will need a version of IM that uses HDRI.

...but.. that is just my opinion, keep looking here for a sensible answer.

(edit: wouldn't be a sound file by any chance)

I'm actually using RHEL 6.9 and Gimp 2.6.9.  Unfortunately where I work it's like pulling teeth to get upgrades to the software on our systems.  I've used ImageMagick in the past and would prefer to use it here but it's not on the company's acceptable list.
And no, they are not sound files.

[Ofnuts]

AFAIK there is no such format (in any Gimp version).

If you can subtract you can round to integer, and even round to gamma corrected integer.

There is a "RAW" Gimp format in the load selector where you provide plain data and tell how to interpret the data:

   

An alternative is the TGA format which is mostly plain bitmap data with a simple header giving width/height.

[DesertBroncoFan]

(12-03-2018, 09:47 PM)Ofnuts Wrote: AFAIK there is no such format (in any Gimp version).

If you can subtract you can round to integer, and even round to gamma corrected integer.

There is a "RAW" Gimp format in the load selector where you provide plain data and tell how to interpret the data:



An alternative is the TGA format which is mostly plain bitmap data with a simple header giving width/height.

I can't find the Load Selector in my version of Gimp (2.6.9).

Also, I guess I should have mentioned that my images are 2048x2048 so TGA isn't an option since it can only handle 512x482, otherwise that looked like a good option.

[Ofnuts]

It must exist in some other form... AFAIK it has always been there. What do you see when you click on items 10 or 11:

(image lifted from https://docs.gimp.org/2.6/en/gimp-file-open.html)

[DesertBroncoFan]

(12-04-2018, 10:55 AM)Ofnuts Wrote: It must exist in some other form... AFAIK it has always been there. What do you see when you click on items 10 or 11:

(image lifted from https://docs.gimp.org/2.6/en/gimp-file-open.html)

[img]file:///z:/Desktop/Image Selection.png[/img]I get a list of the acceptable image formats in both sections.  I don't see anything about custom formats.

[Kevin]

The "Select File Type" has a "Raw image data" option:
   

BUT as GIMP 2.6 cannot handle anything above 8-bit data it's a bit of a lost-cause.

[Ofnuts]

(12-04-2018, 03:20 PM)Kevin Wrote: BUT as GIMP 2.6 cannot handle anything above 8-bit data it's a bit of a lost-cause.

For display you don't need more that 8-bit channel (unless you have a rather expensive monitor)

[DesertBroncoFan]

(12-04-2018, 03:20 PM)Kevin Wrote: The "Select File Type" has a "Raw image data" option:


BUT as GIMP 2.6 cannot handle anything above 8-bit data it's a bit of a lost-cause.

I see it now.  I guess I was in a hurry and didn't look at all of the options.

Thanks for pointing that out to me.

As for the 32-bit vs. 8-bit, I think the 8-bit will work fine as you say.  I'll only be using it to demonstrate the differences between images.

Thanks for your help!