Call for example raw files

Sometimes I use an old lens which in some situations has low contrast caused by internal reflections and probably old coating. For this reason I’m currently experimenting to improve the results using a very simple method on raw data.
If someone has raw files shot using low contrast lenses like for example the one from first screenshot, I would be glad to get some raw files to test.

Here are first results. Left without, right with the new very simple method.
First screenshot was made using the raw from this playraw ([PlayRaw] Extreme telephoto landscape)
Other screenshots were made using raws shot with my old lens.


Very cool. Is this similar to DCP?

There you go, if you manage to make this look good without stacking it’s going to be something.
It is a Beroflex 500mm f8 (aka wundertütte) with an 2x teleconverter.

DSCF2573.RAF (14.9 MB)

A landscape shot with this lens would be fine. And please not 4 stops underexposed :wink:

@heckflosse Arent most of your examples atmospheric haze rather than low contrast lens?

1 Like

That can be more complicated at this time in year, as it is very windy where I am. But I will see what I can do.

This should be doable :slight_smile: The picture is indeed quite underexposed as I was testing if I can use the lens for an ISS pass over the moon, so I need a quite fast shutter speed. Since the camera is almost ISO invariant, it should be fine to just raise the exposure after the capture (at least this is how I understood the theory :smiley: )

Of course there is also atomspheric haze in some of the examples, but this lens is really low contrast compared to other lenses I own and use, but yes, it also works in “correcting” atmospheric haze a tiny bit

A pair of pictures: with an MTO 500 mm (mirror lens, for this one also i attach a jpg) and one with a Petri 28mm F1:2.8 lens full wide (btw there is also a profile for the incredible ca aberration in lensfun).

Wetransfer link here:

btw, what is your lens?

1 Like

Also curious. I doubt it is. There are simpler methods.

No atmospheric haze in this shots I would say :wink:
Left RT neutral profile, right with neutral + my experimental code…


:heart: :dog:


Nikkor Ai 200mm/4.0

And for the dog shots I used a more modern lens (Sigma APO 100-300mm/4.0 EX HF ISM), which has quite low contrast for the upper focal lengths.

Another one using the 200mm lens
Left RT neutral profile, right with neutral + my experimental code… which tries to keep the data linear refered to the scene

1 Like

I am very happy to see my play raw getting a second use.

Just a word of caution, in the case of my extreme telephoto landscape photo, the low contrast was caused primarily by atmospheric haze, rather than optical imperfections, so given your stated intent to develop a poor optical contrast fixing algorithm, this may not be the best test image. Same goes for some of the some of your shots in the original post that seem like the dominant limiting factor in contrast is aerial perspective/atmospheric haze, rather than the lens.

I think better test subjects would be studio setups, with the primary light source hitting the lens, though your test shots do seem to be demonstrating good dehazing.

Primarely, though not only caused by haze. Anyway, doesn’t matter a lot, because the source of the issues is similar :wink:

Does your algorithm have any local adaptations? Because often times, internal reflections don’t produce uniform illumination.

DSC00974.ARW (46.8 MB)
Here is a shot through crappy window glass, with a really close subject. I will also do a studio lighting deliberate haze test right now.

Not at the moment. I’m just experimenting. But try to set raw black levels for your file to
279/703/703/441 and have a look what happens

For you play raw file I mean. I not only increase contrast, but also calculates a much more accurate auto radius value for Capture Sharpening…

LOL, I thought you meant my pigeon photo I just uploaded, leading to some hilariously hard clipped blacks and blue splotches when I applied the recommended raw black points.