[teiki arii]

Hello everyone,
I am sorry if I missed something but are K-1 II files compatible with RT and PS as well?
Best regards.

[heckflosse]

Hello everyone,
I am sorry if I missed something but are K-1 II files compatible with RT and PS as well?
Best regards.

RT 5.4 opens K-1 II files. PS also works fine, but dynamic PS is not supported. In RT 5.4 dynamic PS files don't have the small 'PS' icon on the thumb and pixelshift does not work for these files.
In RT development version dynamic PS files have the small 'PS' icon on the thumb and you also can select PS for these files. Though because the result is not good, I tend to remove the PS symbol from dynamic PS files thumbs...

[teiki arii]

Hi Ingo,
thank you very much for your prompt answer. I am very glad that You answered me. I hope everything is all right for you.
I am to upgrade the K-1 in a month time... Let's wait a little bit and see...

[pentaxfall]

The the Multiple Editor Tab Mode is made for you. (see first paragraph).

Thanks I will try. I have plenty of RAM.

[heckflosse]

Hi Ingo,
thank you very much for your prompt answer. I am very glad that You answered me. I hope everything is all right for you.
I am to upgrade the K-1 in a month time... Let's wait a little bit and see...

Hi Teiki,

everything all right here. Currently I'm trying to implement kind of selective sharpening in RT. Rationale: Even PS images need sharpening. In fact, PS images allow stronger sharpening then demosaiced images. But sharpening the whole image often leads to sharpening artifacts in regions with low contrast (sky, blurry background and so on). That I try to solve currently for RT 5.5. In case of interest (I just started work on this)

Ingo

[teiki arii]

Hi Ingo,
Well, I noticed something different: I prefer sharpening after demosaicing Pixel Shift RAW files. I had better results, but I am opened to other opinions. Perhaps I made it wrong after all. I'll send the K-1 to update next week...

[heckflosse]

I prefer sharpening after demosaicing Pixel Shift RAW files

Sharpening is always after demosaic or pixelshift (at least in RT)

[beholder3]

kudos to heckflosse for all the excellent work and time investments made!

[The Squirrel Mafia]

UPDATE!

I have uploaded a new set of PP3 files. There are some noise reduction tweaks. The previous set smeared the fine detail a bit too much for my taste. There is a tiny bit more grain, but more detail is present in the pictures. You can compare this new set to the previous one to see if you like the noise reduction settings.


I've been compiling my own version of RawTherapee lately & it definitely runs a bit faster. I don't add the lensfun database 'cause that seems to slow down RawTherapee quite a bit. There are some new features that I'm experimenting with for the upcoming RawTherapee 5.5. Hopefully I'll include them in the next batch of PP3 files.


Have fun!

[The Squirrel Mafia]

UPDATE!

There are some exciting new features coming to RawTherapee. There are some additional threshold sliders in the Sharpening & Microcontrast sections & a new AMaZE+VNG4 demosaicing algorithm with a threshold slider that gives more fine detail & less noise as the ISO climbs.

I've been compiling the latest builds for my machine. It's fairly easy to when you follow these fairly simple instructions. You just copy & paste lines into MSYS2. Remember to ignore the "$" sign. That just tells you that the command line starts there. You end up getting the fastest build for your machine possible. I personally don't add the lensfun database. I don't really use it & RawTherapee runs a lot faster without it. Once you're done adding the files, you can run RawTherapee right from the folder or make your own installation.

If you do end up making an installation file for yourself, you don't really need to do anything else outside of installing Inno Setup. No need to change lines in the WindowsInnoSetup.iss file. You don't even really need to get the gdb.exe file either. It's just for yourself. You just right click on the WindowsInnoSetup.iss file & compile. The installation.exe will be in the build folder.

You can also go HERE to get the latest builds if you don't want to compile RawTherapee yourself. You'll have to click on the links shown to get the latest builds.

I have been playing with some of those new features & have added them into the set of experimental PP3 files that I have attached at the end of this post. I can basically use RL Deconvolution sharpening with all the PP3 files to get some sharpening & not get so much artifacts. It's pretty amazing. I've also added a few more ISO settings for cameras that can shoot up to ISO 819,200. I don't really know why anyone in their right mind would shoot such a high ISO, but at least they are there. You will need to have the latest development build for these PP3 files. They will work on the 5.4 stable release, but you'll notice that the images will look green, since 5.4 doesn't have the AMaZE+VNG4 algorithm. You'll have to change the democaising algorithm, but the images won't look the way they are supposed to look in the new development builds.

I'm not sure if all the features that they are working on will make it to the next release, but I guess we will see once it gets released. I'm gonna keep playing with the newer features in the meanwhile. Well, if you feel brave enough to compile a build for yourself or download a development build to try out, then download these PP3 files & give them a whirl. See what they taste like. Hahaha!

[Amoon]

UPDATE!

There are some exciting new features coming to RawTherapee. There are some additional threshold sliders in the Sharpening & Microcontrast sections & a new AMaZE+VNG4 demosaicing algorithm with a threshold slider that gives more fine detail & less noise as the ISO climbs.

Thank you for keep us posting informed about this software The Squirrel Mafia!

[heckflosse]

I'm not sure if all the features that they are working on will make it to the next release

At least Amaze+VNG4 and the new threshold adjusters you mentioned will be in next release. Also the new sharpening contrast mask preview.

[jcdoss]

Yesterday, I fired up two images in Lightroom that were the exact same, except one was pixel shift and other was not. To my eye, and to all other eyes in the room, the two images LR conjured up were essentially exactly the same. My conclusion is that LR must be using only one of the embedded RAW files to do the editing, and who knows how it was being cooked on its way to output. I do recall I had done another test a while back where I could find virtually zero difference between two images taken similarly, processed by LR (one PS, one not).

So I've been doubting LR and looking around for alternatives. RT showed up on my radar, so I ran the same two images thru last night and the difference was stark, to say the least. I'm not going to show examples because I'm such a neophyte that the results will have too many variables to consider. However, just running the non-PS image thru AMAZE and the PS image thru PIXEL SHIFT demosiac engines, respectively, produced such dramatically different images I was nearly blown away. So I guess I'm a convert.

Thanks for this thread, which I just discovered late last night and have yet to read more than half the first post. I am greatly looking forward to learning from you guys over the next few weeks.

Final question... do you guys still use LR in your workflow for data management purposes? Do you use separate DAM software for archiving, keywording, and cataloging? I would like to wean myself off Adobe if this new path works out.

Three cheers for all of you!

[heckflosse]

However, just running the non-PS image thru AMAZE and the PS image thru PIXEL SHIFT demosiac engines, respectively, produced such dramatically different images I was nearly blown away.

Compare the same PS image in RT using AMAZE and PS to see the difference (Open the PS image in RT and process is using AMaZE and PS)! Comparing two different images can have too many outliers.

[Rondec]

Compare the same PS image in RT using AMAZE and PS to see the difference (Open the PS image in RT and process is using AMaZE and PS)! Comparing two different images can have too many outliers.

What are the odds of RT supporting dynamic pixel shift (hand held pixel shift RAW) in the future?

(I could try to get you some files if you want to look at them)...