@gaaned92
Agreed. Maybe in a later stage. I have a bit trouble with the idea of the whole world wide web getting one of my nicest photo’s in weeks. Although I understand the pros of doing so.
Don’t you think that the buildings in the downtown area are leaning backward? A probable cause is that the lens is not exactly parallel to the ground. It might be better if you do some `perspective correction’ in addition to improving colour and contrast.
I followed the guidelines in “How to get Nicon ICM-profiles” and used one of the profiles, the “D7100 Camera Neutreal.dcp” as Custom Input Profile and checked the Tone curve and Look table under the DCP section. You also get a landscape, portrait and standard profile. Try and see what you like the best.
These profiles are created when you open a photo using Capture NX-D, Capture NX 2 or ViewNX as described in the Rawpedia documentation . You can download these programs at no cost from Nikon.
Ok this topic is getting a bit out of hand. It’s really cool how much activity this site has, very helpful and thanks everyone, but let’s go step by step.
First Sharpness/detail
As far as I know I tried everything I can think of. Profiles, all kinds of contrast settings, sharpness, noise reduction. But it’s still noisy, dirty, no smooth colors. The only way to get it clean is by using a lot of noise reduction which removes ‘the dirt’ but creates so much unsharpness that the main goal of getting more detail still isn’t reached.
Here is a different RAW/NEF from the same shoot. A bit different but the same things considering problem of missing detail apply. With the wide building above the train as reference.
Have you tried the Final Touchup panel in the Wavelet tab to increase apparent sharpness via local contrast?
Be careful not to do too much or halos may occur as I recently discovered (and almost got my head chopped off for it)
Ok I have got the neutral icc profiles from the Capture NX 2 software following the Wiki instructions.
One of the profiles gives the photo a red glow and extremely flat contrast, the other makes the photo really bright and without contrast. Adjusting seems very difficult and they just destroy the photo.
I gave it a try … applied a neutral input profile, did some highlight recovery … what is your problem? Is it possible that you just forgot to set a monitor ptofile?
Here is a combination of RT and GIMP. To get some three dimensionality I made the railway tracks a bit more brown so that the cool colours of the sky and buildings look distant. 5addaa20a006c519760a52b8ff59589676118625.jpg.out.pp3 (11.7 KB)
I am sorry. I forgot that I got the profiles from installing the Adobe DNG converter as described in Rawpedia under the paragraph “How to get LCP and DCP profiles”.
@obe
Np of course! I think I did that a while ago so I still got those profiles for my D3300.
I kept trying and finally I can say that I am closer to Aperture concerning the feeling/emotion/atmosphere the image gives.
The game changer was Curve=1;0;0;0.58461538461538409;0.54230769230769182;0.76382225751876831;0.68286071905722934;1;1;
At the same time I am thinking more and more that the Aperture plus Paint.NET combi is a bit of a lucky one. With the addition of the curve in RT and then doing levels in Paint.NET the result is much better. Manual leveling makes it even better. What would be the best tool in RawTherapee to control the colors in such way as the image below?
Open this one with the previous two from yesterday in browser tabs and switch between them to see the improvement. Technically it’s even richer than the Aperture version now.
As you can see, the new output still has a lot of noise compared to the Aperture output. No improvement at all.
It’s up to you to improve. I suggest to use “sharpen only edge” and then apply some noise reduction playing with luminance and luminance detail sliders (see rawpedia documentation)
So that normal you do have noise.
Activate it, put high level, curve, zoom 100 on a refferal zone, just adapt the curve to make dissapear the noise. Then raise detail recovery till it comes noise again
After that, I would advise to use a bit of Wavelet touch up on residual image or :