new version of GIMP is out

[Joseph O'Neil]

http://www.gimp.org/downloads/
http://www.gimp.org/features/

Just down loaded and installed it, tried out a couple of things. Don't worry, I'm not read to abandon my copy of Photoshop Elements, :) but from what little I've used it, It seems to be an improvement over older versions.

I find it is faster and the interface is a bit easier to use. For some of you, might be worth a look.

joe

[Ash]

In fact, I had the older version of Gimpshop, and I reluctantly started using Elements (as it came free with the V700) but if it wasn't for GS not being compatible with the scanner/printer I have, I'd not have bothered to install elements. Thanks for the HU on new Gimp.

[Ken Lee]

It seems to support ICC profiles. :p

Image > Mode > Assign Color Profile

Image > Mode > Convert to Color Profile

Now if it only let you work with 16-bit images, I would really consider using it, and take the time to explore it more deeply.

[David Luttmann]

It seems to support ICC profiles. :p

Image > Mode > Assign Color Profile

Image > Mode > Convert to Color Profile

Now if it only let you work with 16-bit images, I would really consider using it, and take the time to explore it more deeply.

Agreed. Until it can support 16 bit files to work with, it's not ready for any serious work.

[Nathan Potter]

I took a look - makes me feel grumpy and miserable - but in time may be OK.

Nate Potter, Austin TX.

[Ken Lee]

According to the GIMP 2.6 Release Notes (http://gimp.org/release-notes/gimp-2.6.html), there appears to be (optional) support for 32-bit floating point operations.

According to the section entitled GEGL:

"Important progress towards high bit-depth and non-destructive editing in GIMP has been made. Most color operations in GIMP are now ported to the powerful graph based image processing framework GEGL, meaning that the interal processing is being done in 32bit floating point linear light RGBA. By default the legacy 8bit code paths are still used, but a curious user can turn on the use of GEGL for the color operations with Colors / Use GEGL."

"In addition to porting color operations to GEGL, an experimental GEGL Operation tool has been added, found in the Tools menu. It enables applying GEGL operations to an image and it gives on-canvas previews of the results. "

I downloaded a copy, and when creating a new image, I did not see where one can specify anything other than Monochrome/RGB, and the size of the image. There is no mention of bit-depth per se... have I missed something ?

[Ash]

gimp's raw processing works with my Canon 40D files, and it is as accurate as DPP (Canon Bundled software) albeit slower to use. Thanks again for posting the link.

[Ken Lee]

The latest version of GIMP is not available for Mac OS X yet. According to http://gimp.org/downloads/, only the 2.4.7 version is available.

In addition, you need to run GIMP, even the latest available version, under the X11 windowing system. It does not run as a native (Cocoa ?) app. It runs on a Mac, and is free. This much, you can't deny.

[Ash]

Sorry I thought 2.4.7 was the most up to date.

[Brian_A]

Ken,

That's talking about 32 bit processors not image bit depth.

-Brian


According to the GIMP 2.6 Release Notes (http://gimp.org/release-notes/gimp-2.6.html), there appears to be (optional) support for 32-bit floating point operations.

According to the section entitled GEGL:

"Important progress towards high bit-depth and non-destructive editing in GIMP has been made. Most color operations in GIMP are now ported to the powerful graph based image processing framework GEGL, meaning that the interal processing is being done in 32bit floating point linear light RGBA. By default the legacy 8bit code paths are still used, but a curious user can turn on the use of GEGL for the color operations with Colors / Use GEGL."

"In addition to porting color operations to GEGL, an experimental GEGL Operation tool has been added, found in the Tools menu. It enables applying GEGL operations to an image and it gives on-canvas previews of the results. "

I downloaded a copy, and when creating a new image, I did not see where one can specify anything other than Monochrome/RGB, and the size of the image. There is no mention of bit-depth per se... have I missed something ?

[Ken Lee]

"That's talking about 32 bit processors not image bit depth."

The old 8-bit code has already been running on 32-bit machines for a long time, and on 16-bit machines before them.

I presume all this stuff is written in C or C++ and then compiled on different platforms, the Windows version getting the most support.

If we want to perform calculations with precision up to 32 bits, we do need a 32-bit processor, but most importantly, we need to re-write the code to use 32-bit floating point variables, rather than 8-bit or 16-bit. This amounts to big changes in the code base, and a lot of re-testing to go along. Higher precision also slows down performance accordingly, I presume. That's a big consideration.

That's why GIMP - and Adobe too for that matter - have been slow to adopt a higher standard of precision: It requires a lot of programming work ($), and probably slows performance - but is not perceived by marketers as the next "must-have" feature.

Adobe's business model is based on rolling out features each release cycle, to drive new sales and upgrades, by making life easier for the legions of digital professionals and amateurs who need faster processing, not better processing. Or rather, for those to whom faster is better.

Large Format emphasizes image quality over speed and other considerations. We shouldn't be too surprised, if we find ourselves at odds with the rest of the world at times.