[Play Raw] Processing a very high-contrast raw

@chroma_ghost Like this one more. The second image was definitely one of my ROIs.

Cheers @afre 
 and welcome to the forums :stuck_out_tongue: :hugs:

 
LR with daylight WB, and all auto pilot + RNI camera profile

 
BTW @Jacal I put a hat so that I can take it off
 then run, respect the stack bro, you’ve created a zilla :facepunch:

1 Like

Fun picture. Had a quick play in darktable:

Flat base curve + Global Tone Map to get all of the details into a quite flat image. Then local laplacian pyramids + to get the details back and velvia for some quick saturation.

1 Like

And another attempt using shadows/highlights + local contrast, with a slightly warmer white balance.

.

4 Likes

Just foolin’ around with some modules 
 Was fun, thanks for the image.

DSCF3284
DSCF3284.RAF.xmp (7.8 KB)

Ok, last one from me, I promise :stuck_out_tongue_winking_eye:
Down to less than 30 seconds, 3 mouse clicks and one slider adjustment:

https://filebin.net/cpwzvzwymqbjfpzw/fattal-anchor-demo.mp4

No AI power though
 :sob:

3 Likes

A Filmulated attempt:
DSCF3284-output

3 Likes

Wow, glad I watched that video, I didn’t notice the auto match tone curve in the UI.

I’ve been compiling latest dev branch for ages, I really need to pay more attention to things I think. :slight_smile:

@agriggio power, applying math like a :robot:.

RawTherapee 5.1.5.1


DSCF3284.jpg.out.pp3 (13.6 KB)

1 Like

Some curves, a lot of saturation and white balance
DSCF3284
DSCF3284.RAF.pp3 (11.9 KB)

1 Like

Here is my take. Required lots of tuning and patience. Doing this from a sad screen so I don’t know how good or bad it looks. In my estimation, it is average looking compared to the other attempts. However, I hope it recovered enough of the sky and greenery to satisfy @isagalaev.

  1. PhotoFlow
    a) Linear Rec.2020 with negative clipping.

  2. gmic
    a) Curve to remove linearity.
    b) Subdue details (guided smoothing) in prep for #3.

  3. CLAHE (wish gmic had it).

  4. gmic cont.
    a) Expose shadows (luminosity mask).
    b) Curve to increase contrast + brighten shadows, stretch chroma, recover details (guided sharpening).
    c) Prep for sharing (contrast, resize, fp to jpg).

Enjoy!

[Edit: uploaded wrong version. This one is slightly sharper.]

RAF
DSCF3284.RAF.tif.out.pp3 (10,7 KB)

some quick RT basics, gmic popshadows + simplelocalcontrast, color-efex vignette + warm-up curve
image

thx for this nice play raw

Here is my first take with PhotoFlow. @isagalaev - thanks for sharing this image, it’s an excellent example for testing the various shadow adjustment methods!

My edit is mostly based on exposure compensation + ad-hoc masks, plus some fine tuning with RGB curves:

DSCF3284-PhF

And the corresponding PFI file: DSCF3284.pfi (84.5 KB)

Hi Morgan,

I have downloaded your .PP3 file and tried to apply it to the RAF file of this playraw, to see how you achieved your excellent result in practice. However, after opening the RAF file in RT and applying your PP3 file with the “Load a profile from file”, this is how the processed image looks like:

Here is the information about my RT version, running under OSX:

Version: 5.3
Branch: 5.3
Commit: ec0f7936
Commit date: 2017-09-30
Compiler: clang-mp-3 3.9.1
Processor: generic x86
System: Apple
Bit depth: 64 bits
Gtkmm: V3.22.1
Lensfun: V0.3.2.0
Build type: release
Build flags: -arch x86_64 -O3 -mmacosx-version-min=10.9 -march=x86-64 -std=c++11 -mtune=generic -Werror=unused-label -mmacosx-version-min=10.9 -flto -fopenmp=libomp -Werror=unknown-pragmas -Wall -Wno-unused-result -Wno-deprecated-declarations -O3 -DNDEBUG
Link flags: -arch x86_64 -mmacosx-version-min=10.9 -march=x86-64 -mtune=generic -headerpad_max_install_names -flto
OpenMP support: ON
MMAP support: ON

Any idea why the output looks so different with respect to your edit?

Thanks!

The version @Morgan_Hardwood used has some new tools. For example:

[FattalToneMapping]
Enabled=true

Thanks Ingo! Is there a place where one can download bleeding-edge OSX builds? If not, would there be some interest in such a possibility?

@Carmelo_DrRaw

2 Likes

Stupid me! I completely forgot about that thread
 thanks!

1 Like

Also, if the image still doesn’t match, you might want to checkout a commit from dev from the day I posted that image, as Fattal’s behavior changed over time and I’m not sure whether its behavior from the day I posted hadn’t changed till today. Backwards compatibility is absolutely not supported for development builds, but we do try to support it for stable releases.

1 Like