Does "Lens detection" on Leica M240 impact DNGs?

shorelineae

Finder of ranges
Local time
9:46 PM
Joined
Jan 17, 2019
Messages
170
Is the lens correction applied to DNGs? I have shot a lot of photos at the wrong lens detection (21mm) when it should have been set to Undetected.

I shot in DNG-only mode. The photos don't look distorted as such to my eye, but it makes me wonder if the modifications that the camera does based on the lens settings affects the DNGs?
 
My understanding is that it does effect DNG data. But I think the correction is more about shading and vignetting, not distortion correction.

Shawn
 
If you are using one of the RAW conversion programs that allows you to turn off corrections, you can probably test it.

Also, if you like how the images came out, and they look good to you, it doesn't really matter.
 
As far as I recall, it may affect vignetting control. I think may have done the same thing you did at one point on one of my digital Leicas and that’s what I noticed.
 
My DSLR manual has four settings that can be turned on or off: distortion correction, peripheral illumination correction, lateral chromatic correction, and diffraction correction. It also states that correction information is saved as a RAW file parameter and you can select whether to apply the correction to the images when developing them. So it appears at least on my camera the correction is not embedded but can be chosen to be used later if desired.
 
I really do want to have images that are not 'manipulated' by algorithms but to reflect the way my lenses render. So I never use detection. On rare occasions I do some vignetting etc adjustment but I try not to. Most of my lenses are older Leitz lenses, each with their own character. Recently I bought a 1.5/75 7A lens to try and that also I use as it is. I think with my approach I am in control. Why would I want to change the optical behavior of a lens anyway and create a fake image? I also would not replace the sky or change objects etc in an image.



There is enough fake in the world.
 
It is to try and deal with how a digital sensor responds to off axis light compared to film. Film doesn't care about light angle, digital does. Additionally there is more glass in the optical path with digital (vs. film) that effect this too. This is why some lenses that are great on film can show color shifts, vignetting and smearing on digital. The lens coding is to try and reduce the color shifts and vignetting.

Shawn
 
On the topic. I use a lot of non-Leica lenses. If I use lens detection with these one the "closest match" will it be better than not using detection? Lenses are:

ZM 25/2.8
ZM 35/2.8
ZM 50/ 2
VC 34/1.4
 
I think when trying for the closest match, the specification of focal length and aperture is a good starting point; but I believe the distance of the rear element from the film (sensor) also matters. If not the rear element, then the rear nodal point (but who knows where that is). So if possible that should also be considered.

Trying out the available options and checking the results is likely the best approach.
 
I really do want to have images that are not 'manipulated' by algorithms but to reflect the way my lenses render. So I never use detection. On rare occasions I do some vignetting etc adjustment but I try not to. Most of my lenses are older Leitz lenses, each with their own character. Recently I bought a 1.5/75 7A lens to try and that also I use as it is. I think with my approach I am in control. Why would I want to change the optical behavior of a lens anyway and create a fake image? I also would not replace the sky or change objects etc in an image.



There is enough fake in the world.

I view this as a humble but unnecessarily puritan point of view. if end result (image) matters, and not by how it was made, then why not add adjustments that technology provides? only if content has some added value by pointing out that "these images have authentic Leitz 1930's look and have 0 post processing in them", guess then it would make some sense to me ;)

come to think of it, thats good idea for dedicated photo blog :)
 
On the topic. I use a lot of non-Leica lenses. If I use lens detection with these one the "closest match" will it be better than not using detection? Lenses are:

ZM 25/2.8
ZM 35/2.8
ZM 50/ 2
VC 34/1.4

good question, am having also 35/1.4 and have coded it as Summilux. to my eyes, in-camera correction does relatively little work on image. when I import it to Lightroom and set lens detection on, transformation is bigger and been wondering how accurate it is.
 
when I import it to Lightroom and set lens detection on, transformation is bigger is.

This is what I ended up doing. I applied Lens detection in Lightroom for the lens I _had_ been using (Voigtlander 50mm f1.5) and there was a significant distortion correction.....

I will also test it with the photos that did not have the wrong lens detection (I switched to "Unset" at some point during my trip when I realized the issue).
 
This is what I ended up doing. I applied Lens detection in Lightroom for the lens I _had_ been using (Voigtlander 50mm f1.5) and there was a significant distortion correction.....

I will also test it with the photos that did not have the wrong lens detection (I switched to "Unset" at some point during my trip when I realized the issue).

Those LR profiles can be very inaccurate from what I have seen.

Why didn't you just turn it off in camera to see the difference?
 
Those LR profiles can be very inaccurate from what I have seen.

Why didn't you just turn it off in camera to see the difference?

My M240's LCD does not work (long story). It's easier for me to test on the shots I already took. Will update shortly.
 
Easy to compare with DNG Cleaner. My feeling is 6-bit coding does nothing but adjusting color shifts when using WA lenses but i may be wrong. I mean on the M240. Digital CL is another story.
 
Leica's lens profiles, for *all* of their interchangeable lens cameras—the ones capable of being fitted with M-mount and R-mount Leica lenses via a Leica adapter or with their native mount—are designed to preserve the optical rendering of these lenses on whatever sensor they're being fitted to. They're not designed to 'correct' the lenses' behavior, just make them image the same as they did for their original design intent and capture medium. This is an important thing if you want all your Leica M and R lenses to give you the same rendering qualities as they do/did on film and between digital bodies.

This is a different strategy compared to the SL and TL mount lenses which are designed from the outset to be used with image processing corrections to 'finish' their rendering intent.

The Leica lens profiles also help the various bodies you use them on to evaluate exposure, and they fill out the camera-embedded EXIF metadata for post exposure image management purposes.

Leica of course does not provide lens profiles for any other than Leica M and R lenses. You can choose to use them manually by looking them up on the several tables compiled by enthusiasts for whatever lenses you want to use them with and see how they do. The effects are most significant/noticeable with short focal lengths (under 50mm). Sometimes you find an excellent match (for instance, I find the lens profile for the Summaron-M 28mm f/5.6 is a very good match to the Color Skopar 28mm f/3.5) and sometimes it's better to use an aftermarket lens correction (like the ones provided by Adobe with Lightroom).

Leica's M and R mount lens profiles are applied in-camera in the translation process from capture data to the raw (DNG) file and cannot be disabled after the fact, like the lens profiles for SL and TL (and Micro-FourThirds, etc) lenses. So you test what profile does what by shooting test scenes with them off and then on again, and in the context of whatever your preferred raw converter might be. Be prepared to have to look carefully to see what they do for most lenses, as the corrections are often very subtle, for most lenses.

G
 
Back
Top Bottom