Raw decoding - non bayer

Hi, I have just discovered photoflow and it seems like a very interesting piece of software;

That said and before I take the plunge, There is one question I could not find a definitive answer to : does Photoflow handles non-bayer flles (namely Fujifilm RAF for Xtrans I, II and III)

Thanks in advance
Didier

Hi! X-trans format is not yet supported, but it is definitely in the TODO list. Would you be available for helping me with some testing once an initial X-trans support will be available? Which operating system are you using?

Thanks for looking!

I am running on linux mint and it would bé m’y pleasure to help

Then I would suggest you to install the photoflow-git package from Dariusz Duma PPA. It provides the current development version of the program, and X-trans will be available through this package as soon as it gets committed to github.

Meanwhile, you might want to start familiarising yourself with PhF, by editing tiff or jpeg files. If you are interested, I am collecting some useful tutorial here: https://plus.google.com/communities/108772952003062961576/stream/69429669-4a79-4834-b161-f55b3512e45c

@didiergm an update on the subject: I have just committed to the github repository a preliminary support for X-trans. For the moment the demosaicing is limited to a fast method, but it should be enough to have a first look and see if things come out as expected.

If you install photoflow-git from Dariusz Duma PPA, you should get the updated package in a couple of days.

I must here thank @heckflosse for his precious help, as he gave me all details I needed to get this started…

Full support should come rather soon, maybe in a week or so.

2 Likes

Whaouh! Call that a short reaction time!!

I know that XTrans demosaicing is not easy (Adobe does not have it right despite several attemps !!!)

I have quite a few friends who might be interested in helping, but I know what they’ll ask -
1- Is the new version available for Windows adn Mac ?
2- Is there a French version

For the second question I could add : would you be interested in a French translation, and if yes how should I proceed to have a go at it ?

Didier

Whaouh! Call that a short reaction time!!

I know that XTrans demosaicing is not easy (Adobe does not have it right
despite several attemps !!!)

I have quite a few friends who might be interested in helping, but I know
what they’ll ask -
1- Is the new version available for Windows adn Mac ?
2- Is there a French version

For the second question I could add : would you be interested in a French
translation, and if yes how should I proceed to have a go at it ?

Didier

Hi, I comiled photoflow-git stable branch with last commit and now I’m testing it with Fujifilm RAF files. It looks OK for now. Thank you. If I find any problems I’ll report them here.

Best regards.

1 Like

Well, this is actually a good example of the power of OpenSource… you can stand on the shoulders of The Giants and build some cool stuff with a reasonably limited effort.

In my case, the RAW decoding is taken from Darktable, and the demosaicing from RawTherapee… other great guys have already done the tough job for me, and I’m just putting together the pieces :smiley:

I can easily make test Win and OSX packages if needed. It’ll probably take a couple of days

There is an outdated French translation, which I only tested under Linux. The French translation is provided by one volunteering user, and I must admit that the current translation is outdated not because of his fault, but because of me putting the subject in very low priority. But I’m ready to get the translations better organised for the future.

Would it be fine for you if I tell the other translator to get in touch with you here on pixls.us?

Thanks!

Good to know! Thanks for checking. I’ll post here about any progress I am making with the X-trans support.

Best,
Andrea

  • Mac & Windows builds : if it take you time, may be wait a bit, I have placed a call to my fellow FUJI users; if as I hope some volunteer, I’ll let you know and then may be you can spend the time to make the builds

  • I am waiting for the ppa to be updated and I’ll test the Xtrans demosaicing

  • French translation no problems, you can ask the other user to contact me here; if we can arrange to work together that would be easier on both of us :wink:

1 Like

I just had a look, and the PPA seems to be up-to-date already (the package was built 9h ago)

Ok, I will do that. Thanks for the offer!

What crazy Frankensteins creature are you working on? :wink:

1 Like

You should have a look at the freaky lab in which the whole thing is being assembled :wink:

1 Like

Weird - I would have expected it to show on the update screen of mint (I am fairly new to Mint, having transferred recently from an RPM based distribution); I added the ppa to the list of repositories

I also had issues sometimes with the update being available but not visible in the update manager… just try to update your system, and then check the version of the photoflow-git package: it should be something like 1:2.6+git20160619-549

Maybe @Dariusz_Duma knows what could be wrong?

It was updated; apparently silentely!!! Ok, I’ll see what happens in the future; I’ll have a look tonight

Very interesting results, by ad large excellent for a first go!!!
Basic stuff : image size, showing that the file is properly read - is fine
Compared to in-body jpegs and Rawtherapee (Could not check against darktable yet)

The overall feeling is that images processed with photoflow are more refined ie I feel I have more control over the demosaicing, resulting in finer details, a bit less noise

Depending on the image, I had to dial in between 0.6 and 1.8 exposure compensation at the demosacing stage, not sure why. I could not get any visible results out of the AC correction. I have not tested yet and image with really fine and repetitive details (where LR falls short) High iso tree leaves are usually poorly handled, greens also tend to be difficult

I obviously don’t know your software well enough, (e.g. could not find any sort of sharpening tool, notr am i clear yet on the benefits of layer groups as opposed to individual layers) but I am sure that with a bit more time I will get better and from what I see it is definitely worth it!!!

This might be due to the different WB settings used by PhF and RT for the demosaicing: RT uses an auto-measured WB setting for the demosaicing, and then applies the final WB afterwards; PhF does the demosacing directly with the WB setting chosen by the user un the WB tab.

Do you mean you need exposure compensation to match RT and/or LR? RT has auto-exposure turned on by default, while PhF has no auto-exposure feature yet.

CA correction is presently disabled in the code, because it only works with Bayer patterns. I still have to see what RT and DT do in the X-trans case

The various tools are accessed from the buttons on the left of the layer list, with tooltips to briefly describe the tool associated with each button. The first button, with a red background, opens the more general tool chooser dialog, where all tools are listed and accessible. The tools are grouped in few tabs; the sharpening for example is in the “detail” tab.

Group layers are useful for example when you want to be able to enable/disable a bunch of editing steps, to quickly have a before/after toggle on the preview image.
Other uses involve for example the use of blend modes other than “normal”.

This video gives a basic overview of the program usage: Google Workspace Updates: New community features for Google Chat and an update on Currents

Concerning the use and need for group layers, you might have a look at this “Orton effect” tutorial: PhotoFlow Image Editor Blog: Tutorial: non-destructive Orton effect in PhotoFlow
The result can only be achieved with group layers… the tutorial uses a rather old version of the program, but the basic concepts are still valid.

So I prefer PhF way :slight_smile:

Not a problem at all, as this way I could get more details on the bright red ship’s hull (mainly water reflexions) than with RT

Understood - just wanting to mention that

Thanks for the explanation on the intention behind the layers group. :slight_smile: