Basic Photoshop settings for RAW JPEG conversion.. EX2F?

Petak wrote:

ttbek wrote:

Petak wrote:

ttbek wrote:

Petak wrote:

ttbek wrote:

Petak wrote:

I actually don't recommend Lightroom, I don't really use it, so someone else could probably be more helpful on that. RawTherapee (I think on OSX?, Linux, and Windows) is probably what I would recommend for absolute quality

I have tested both RTh and Lr5 with my NX2000 files and Lr has both more straightforward interface

Subjective, I don't find the LR5 interface that great either.

and visibly higher quality conversion (better detail

I'm not seeing that, see below.

, smoother and more accurate colour profile

They use the same color profiles, unless you're using that as a layman's term.

).

Did you only use defaults? I would like to see some evidence.

Because you did present evidence to support your claim of RT's "absolute quality"?

I wasn't asked to. Maybe they've made improvements in this from LR4 to LR5, but:

RawTherapee 400% crop CA correction, everything else turned off

LR4 400% crop CA correction (can you turn other things off?, can you use different demosaicing?)

There are notably more "jaggies" in the LR4 image. This could be due to some default sharpening I don't know how to turn off. The default tone curve in LR4 is more contrasty and pleasing, but it's just a default. Also the default jpeg export quality setting in LR4 is 60%.... I did them a favour and bumped it to 100% for this comparison. The jaggies are not due to the CA correction, I also looked without and that was the same.

You said you already tested them both for your NX2000, so it should have been easy for you to show the results. I had to go out of my way to do these just now, requiring me to reboot in Windows etc...

Now, it may be that some other things in your process LR does better than RT, a certain kind of noise reduction, etc... but at the very basic conversion stage (demosaicing) RT offers the user more options and the potential for better results. Unless you (or someone else) can show otherwise I think that this is where the evidence is currently pointing.

I believe what you call "jaggies" is in fact the inevitable result of up-sizing to 400%

I'm not referring to general pixelation, but rather to the aliasing evident in the tickmarks of the chart as you look to the ones more to the up and right in the LR4 image. It was enlarged to that level without any resampling (it wasn't done in a resampling viewer, it was done in a program that doesn't screw with your pixels) and was done so to make the difference obvious.

- something rarely anyone does or needs. Below are two 100% crops of the same raw file developed in both LR5 and RT 4.1 with the default settings.

See, that's what I was asking you about. Default is not best.

To me the one produced by LR is clearly better - cleaner, smoother, with better detail and less magenta and sharpening halo

See, you didn't turn off the default sharpening in RT and probably didn't turn on the CA correction either. You probably didn't use the same demosaicing I did either. I in good faith tried to get the best result I could from LR4, so do enlighten me regarding anyplace I can turn off some more default adjustments or try a different demosaicing method. I'm basically seeing that you like the LR5 defaults better rather than any kind of quality comparison of both program's best case scenario's. I didn't have issues with magenta or sharpening halo in the RT result I posted.

You should also note that what I meant by "absolute quality," was not quite as literal as you're taking it, but you're certainly not showing any compelling evidence. Or does bias pass as evidence these days?

around the text and with better colour separation.

The image I used is the 100 ISO SRW sample available here

I know where to find it ^_^.

- feel free to do whatever you want with it in RT to get the best out of it and show us the result (using the above crop of the image) in 100%.

I personally feel like some of your metrics were subjective rather than objective and that 100% doesn't do anyone favors when they're trying to pick out the very small artifacts/flaws in the renderings.  I don't think my posting another image will wind up being constructive but rather degenerate into subjective measures of image quality.  I can do it later if you insist though.

And BTW LR and RT are NOT using the same camera profiles,

I'm well aware of that.

which is clearly evident once you open the same file in both. RT is also not correcting lens distortion based on the info available in the file itself.

And of this as well.

All in all RT is very good if you like spending time playing with the software itself instead of with the images but if photography is your priority, LR is the better option IMO.

I don't share that opinion, but there's a wide variety of opinions on this topic.  My personal overall preference is Darktable.