G'MIC-8bf - A Photoshop-compatible filter interface for G'MIC-Qt

Thanks for all the updates @PDN_GMIC , hopefully the rythm of releases will slow down this year :grin:

Release version 3.2.5, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.2.5.

The Exchange version has been submitted to Adobe for review.

1 Like

Here are some download statistics for the Adobe Exchange version:

It has been downloaded a total of 3693 times across all versions.
This breaks down to 1097 downloads of version 3.2.5, and a combined 2596 downloads for versions 3.1.6 to 3.2.4.

For comparison with the GitHub release of version 3.2.5, the x64 version has 426 downloads, and the combined count for the ARM64, x64 and x86 versions is 823 downloads.

The downside to this increased popularity on the Exchange is that people report issues by leaving a one-star review with a vague description of their problem, instead reading and following the instructions in the support section of the plugin description.
I left a reply asking for more information, but I am not optimistic that I will get any.

1 Like

Thanks for sharing the stats!

For the saving issue of this user, could it be related to the fact that the output may be in 16bits/channels rather than 8bits/channels ?
If the output image is in 16bits, it cannot be saved as a .jpg file without dividing the values by 256 (or 257, depending on your taste :slight_smile: ).

I don’t think so. The 8bf plugin uses a custom image format when communicating with G’MIC-Qt, which the 8bf plugin will convert convert to PNG for 8-bit and 16-bit images or EXR for 32-bit images when required. I tested the code with an 8-bit export and it was working.

The actual review is the following:

tries to export a preview image when applying the filter then saves a random image of lines

I did notice when testing the Polaroid filter that clicking Apply to execute it a second time would update the G’MIC-Qt preview image, but when you click OK the output only has the first invocation of the filter. The GIMP version of G’MIC-Qt 3.2.5 has the same behavior.

I am not sure if it is considered a bug, but I was surprised by the final output not matching the G’MIC-Qt preview.
Perhaps this behavior is what the above bug report is trying to convey, but there is not enough detail for me to definitively say what that user’s issue is.

OK, I thought that when the user said ‘Exporting’, he was actually talking about the file export done by the 8bf plug-in because of a different output size.

Release version 3.2.6, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.2.6.

The Exchange version has been submitted to Adobe for review.

1 Like

I received another report from a user of the Exchange version related to the previously mentioned apply bug. For reference, here is the original bug report:

tries to export a preview image when applying the filter then saves a random image of lines

The new report adds a few more details, but it is still not enough for me to have any idea what is going on:

I have the same issue with several filters - it doesn’t happen always and I haven’t figured out why yet, but once the bug is there in a file it seems to persist. Couldn’t get it to work in one Photoshop document, in another it did again, in yet another it doesn’t. For me it happened, amongst others, with noise filters, chromatic abberation, huffman glitches, breaks, etc. etc.

I asked that user to file a GitHub issue with steps to reproduce the issue and one of the files that causes the bug.

In other news, I recently attempted to cross-compile a native Windows ARM64 build of G’MIC-Qt but I was unable to get it to use the correct include and library paths. That may have been due to a problem with my cmake toolchain file, or it could be an issue with the MinGW clang cross compiler support on Windows.

1 Like

Release version 3.3.0, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.0.

The Exchange version has been submitted to Adobe for review.

3 Likes

Release version 3.3.1, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.1.

The Exchange version has been submitted to Adobe for review.

1 Like

Release version 3.3.2, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.2.

The Exchange version has been submitted to Adobe for review.

2 Likes

Release version 3.3.3, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.3.

The Exchange version has been submitted to Adobe for review.

1 Like

Release version 3.3.4, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.4.

The Exchange version has been submitted to Adobe for review.

2 Likes

Thanks Nicholas. Your work is much appreciated! :beers:

FYI - plugin v.3.3.4 is working fine on Affinity Photo 2.4.

My firewall (Comodo) throws some warnings about the plugin accessing certain processes, but once OK’d, all works well.

Thanks for the hard work.

1 Like

…this setting works also on Windows 11

This said, it’s pretty slow. I didn’t do extensive testing, but an image processed with “simple local contrast” (I think it’s called) took like 15 Seconds or so.

Hardware: Notebook, i7-1355U, 32Gb, DDR5, Iris XE

Is that normal or is there any tuneable setting? Additionally, does eGPU brings any benefit?

G’MIC can’t do GPU processing. I wish it can though!

2 Likes

Good to know. Saves me lots of money :wink:

Release version 3.3.5, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.5.

The Exchange version has been submitted to Adobe for review.

3 Likes

Release version 3.3.6, Releases · 0xC0000054/gmic-8bf · GitHub
Changes:

Updated G’MIC to version 3.3.6.

The Exchange version has been submitted to Adobe for review.

3 Likes