I snapped and stitched a panorama with a very high dynamic range, which you can see here, in TIFF.
I have tried my best to map the dynamic range in such a way that exposes more of the detail - and made heavy use of tone mapping and even selectively made the blues (sky) less bright and the browns (cliffs) more bright via the HSV equalizer. You can find the .pp3 file here. I ended up with what I thought was a rather passable image.
Out of the blue, then, a colleague tells me there’s this NIK plug-in (called pro contrast - it seems I am not allowed to put more than 4 links in a post…) which can deal with dynamic range issues - apparently my efforts didn’t really look like they dealt with the issue. And with a few rapid clicks, this is the image he got.
Now, imho this is a considerably better result. Any ideas how I could get something similar via RawTherapee?
I’m on my phone right now, so I can’t see your original TIFF nor your final edited image, but the one your friend made with Nik seems to have badly posterized whites in the clouds.
I’d recommend trying the higher-radius effects with tone mapping, wavelets, etc, along with compression. Nik Pro Contrast in the Color Efex tool gets you a nice effect for very little effort: there’s so few sliders to pull! I think these things could be possible in RT but requires some creativity in finding the right type of contrast to adjust. Also noticed you might want to restitch your pano, as there is a big error in the bridge’s arch.
Can you say something more about how you took the shot(s)? Have you bracketed several exposures? Which software have you used for stitching?
For example, Hughin gives you the possibility to save the individual stitched exposures into separate TIFF files, which you can then manually combine in GIMP or PhotoFlow to compress the dynamic range and increase the local contrast, and then eventually further process in RT… I wrote some guidelines on how to save the individual exposures with Hugin here.
I have no bracketed shots (this was all done without a tripod, and on “auto” - so my three shots have different exposures each).
I used the following command to convert my three shots to tiff: ufraw-batch --out-type=tif --out-depth=16
and then I stitched them up with Hugin. I opted to have Hugin generate just one image, as I don’t understand how I can postprocess three resulting shots in a coherent manner. Do you have any ideas? I don’t have Photoflow, nor, it seems, any easy way to get it.
Maybe you could share the three RAW files? I have never stitched shots with auto-exposure, so I would need to do some experimenting to see how to properly combine the Hugin outputs.
Which system are you using?
Here is what I could get quickly out of your image, using curves, luminosity masks, some local contrast enhancement and saturation boost:
Gentoo Linux. I might add that I found Photoflow on github, and it seems I could easily write a package, but I feel my workflow is too convoluted anyway, and I would rather not use an additional program at this point. I would like to see what the best I can get out of RawTherapee is.
Did you generate the Image you posted with RawTherapee? If so, could you share the pp3?
The first file I posted was obtained with PhF, but one could get to the same result with GIMP as well…
Here is one more attempt with RT, using shadows/highlights, tone mapping and HSV equaliser for selective saturation increase (predominantly in the reds):
You can add more contrast and saturation if you like that.
I’m not particularly happy with the color of the stones, but I didn’t find an easy way to fix that.
Also the higher radius wavelet parameters including the Residual Image adjustments. In my version I tried everything and ended up raising the image lightness/chroma in in the Ciecam 2002 area, and used Residual Image in wavelets to give it some “dynamic” contrast.
Thank you! Your pp3 is very educational.
RT can perform a lot of wonderful things, but it does not help just to read about it, one has to see real, live results as well. Like pp3s.
After all, who would know that we could use Residual Image in wavelets to give it some “dynamic” contrast?
What I wrote a few minutes ago is just as valid for your example:
Who would know that we could use default tonemapping, Lab*, and the experimental “Retinex in Wavelet”?