Digital Lens Corrections

[Bgood]

Advertisement (gone after registration)

I now own a Leica Q3 and am considering buying a new Leica M (either the M11P or the new M rumored to come out). I would like to buy the 35mm APO f2.0. I tried the M11 and the 35mm APO in my local Leica store. When I brought the images into Lightroom I noticed that the Lightroom profile correction noticeably crops and corrects distortion even on a high end fixed focal length lens. The images are beautiful and the colors are excellent. But honestly, I was expecting less software correction from a lens of this caliber. I have seen lenses with far more software distortion correction but I also have seen a lot less and would welcome comments. I am by no means an expert on lenses.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[lct]

Only way to approach perfection in a compact package i suppose. Solution? Embrace bulkiness or imperfection.

[adan]

Your Q3 lens - natively - has massive fisheye-like distortion. And even vignettes in the corners of its sensor.

https://diglloyd.com/prem/s/LEICA/LeicaM/LeicaQ-distortion-Dolls.html?dglyPT=true

The difference is that the Q cameras themselves are designed, ahead of time, to do their own digital cropping of the image, and correcting of that distortion, from their dedicated, non-interchangeable lens, right in the camera.

In fact, before you even see it in the electronic viewfinder, so a question of microseconds.

Not much left to be corrected by software after that. Except for some sophisticated processing software that can "disregard" the camera's corrections, so that the lens's actual distortion performance can be seen.

.....................

M lenses such as the 35mm APO-Summicron-M-ASPH (or any other) cannot depend on such dedicated in-camera, or even post-processing software correction - because they have to ALSO work in a completely non-digital work flow.

I.E. they have to work well on any Leica M back to the 1954 M3, up through the current MP, M-A and revived M6, on film, that then may be printed in a chemical darkroom. With no digital stages at all to do the corrections the Q does.

So the M lens is more expensive (because - counting the lenses alone, it DISTORTS LESS than the Q's 28mm f/1.7).

Edited March 29 by adan

[jaapv]

I think that you will find very few lenses on the market nowadays, especially wideangle ones, that do not resort to digital corrections.

[Bgood]

5 hours ago, adan said:

Your Q3 lens - natively - has massive fisheye-like distortion. And even vignettes in the corners of its sensor.

https://diglloyd.com/prem/s/LEICA/LeicaM/LeicaQ-distortion-Dolls.html?dglyPT=true

The difference is that the Q cameras themselves are designed, ahead of time, to do their own digital cropping of the image, and correcting of that distortion, from their dedicated, non-interchangeable lens, right in the camera.

In fact, before you even see it in the electronic viewfinder, so a question of microseconds.

Not much left to be corrected by software after that. Except for some sophisticated processing software that can "disregard" the camera's corrections, so that the lens's actual distortion performance can be seen.

.....................

M lenses such as the 35mm APO-Summicron-M-ASPH (or any other) cannot depend on such dedicated in-camera, or even post-processing software correction - because they have to ALSO work in a completely non-digital work flow.

I.E. they have to work well on any Leica M back to the 1954 M3, up through the current MP, M-A and revived M6, on film, that then may be printed in a chemical darkroom. With no digital stages at all to do the corrections the Q does.

So the M lens is more expensive (because - counting the lenses alone, it DISTORTS LESS than the Q's 28mm f/1.7).

Thank you very much for this very helpful reply. I started with photography in the 1970s, long before the availability of digital corrections. My notions are probably out of date. I think I should just focus on the end result. Cameras and lenses in the 1970s weren’t perfect and today’s cameras and lenses are far better. 
 

I will say from my test images the 35mm APO colors just pop to my eye. They are very pleasing. 
 

But I also love what the Q3 produces and did not know it relied heavily on digital correction. 
 

I guess I’m just looking for unattainable perfection. 
 

I also shoot Canon and Canon has many lenses that rely on digital correction but they still are pleasing to me. 
 

I keep reading that there is always a price for digital lens correction but if the result is better with it why worry about since it reduces size and weight?

Again, thank you. 

[LocalHero1953]

19 minutes ago, Bgood said:

I guess I’m just looking for unattainable perfection. 

One has to ask: what is the real-world benefit of such perfection?

[lct]

Advertisement (gone after registration)

People wanted sharpness but HCB said it is a bourgeois concept, so now they want perfection. Bourgeoisie is not what it used to be anymore 😁 Nothing personal 😎

[pgk]

3 hours ago, LocalHero1953 said:

One has to ask: what is the real-world benefit of such perfection?

Its in the mind.

[charlesphoto99]

3 hours ago, LocalHero1953 said:

One has to ask: what is the real-world benefit of such perfection?

Absolutely none except chasing the gear dragon. 

[adan]

There is a - rumor or myth - that the Q was originally developed,in the early 20-teens, strictly as an in-house Leica test-bed for digital optics corrections in-camera. Because Leica saw the writing on the wall, for competing in the digital-only lens world (with the soon-to-be SL line, and perhaps an EVF/mirrorless version of the larger-format S line).

And was possibly the reason that the late Erwin Puts - https://photo.imx.nl - a connoisseur and technical expert of Leica's "pure optical" designs across 100 years, announced his break with Leica. He thought digital correction of distortions, planned-into-the-camera/optics from the beginning, was a form of "cheating."

Make of that what you will.

It is true, as Digilloyd complains, that stretching and resampling images digitally to remove distortion, can lead to some image degradation. And noisier high-ISO Q images, especially the Monochroms, can show some strange "waffle" patterns at times. But less noticeably as megapixels have climbed above 24.

In any event, the Q worked so well that it was released as its own very successful commercial product line, just before the SL.

[Jeff S]

5 hours ago, Bgood said:

I guess I’m just looking for unattainable perfection. 
 

 

How do you view your final output?  I can’t think of a single well known picture/fine print that didn’t require at least some degree of post-process editing, film or digital.  And even if the pic was perfect out of camera, different display lighting conditions alone would require editing adjustment.

The photographer always decides, for better or worse, from shot to edit to display.  No such thing as ready-made perfection in this craft.

[Bgood]

15 hours ago, adan said:

Your Q3 lens - natively - has massive fisheye-like distortion. And even vignettes in the corners of its sensor.

https://diglloyd.com/prem/s/LEICA/LeicaM/LeicaQ-distortion-Dolls.html?dglyPT=true

The difference is that the Q cameras themselves are designed, ahead of time, to do their own digital cropping of the image, and correcting of that distortion, from their dedicated, non-interchangeable lens, right in the camera.

In fact, before you even see it in the electronic viewfinder, so a question of microseconds.

Not much left to be corrected by software after that. Except for some sophisticated processing software that can "disregard" the camera's corrections, so that the lens's actual distortion performance can be seen.

.....................

M lenses such as the 35mm APO-Summicron-M-ASPH (or any other) cannot depend on such dedicated in-camera, or even post-processing software correction - because they have to ALSO work in a completely non-digital work flow.

I.E. they have to work well on any Leica M back to the 1954 M3, up through the current MP, M-A and revived M6, on film, that then may be printed in a chemical darkroom. With no digital stages at all to do the corrections the Q does.

So the M lens is more expensive (because - counting the lenses alone, it DISTORTS LESS than the Q's 28mm f/1.7).

You are really educating me. Thank you. I own a license to DxO PureRaw. I have attached a before and after and the amount of distortion astounds me. Is this what you are referring to?

 

https://photo.imx.nl/blog/files/75caecf5a21140ab6c9b46f825ef253c-143.html

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited March 29 by Bgood
Add link

[adan]

- Regarding the Erwin Puts blog post - yes, that is the one.

- Regarding your images - that looks about right for how barrel/fisheye distortion from a lens (e.g. Q 28mm, as demonstrated without correction in the DigiLLoyd link)    

(   )

would be corrected mathematically by a computer (which can be the camera's CPU) with a counter-acting pincushion distortion

)  (

to "square up" things like the lines in the Smithsonian building.

But distorting the shape of the total captured image rectangle from a rectangular sensor,  which leads to the black "backgound arcs." ("no data available here").

Which are then cropped off, leaving a "normal-looking 28mm" final picture.

It's been suggested that uncorrected, the Q "28" actually produces about a 24-26mm fisheye field of view at the corners, Which, cropped away, leave behind a "legitimate, rectilinear 28mm" field of view.

...............

Although it should be noted that focal lengths and their fields of view - even for Leicas - always have some - leeway - in labelling.

I have a 1974 90mm M-lens from Leitz/Canada that is sub-marked on the focusing ring, "05". Which means that when that specific lens was assembled and tested, it had an actual focal length of 90.5mm. Leica then used a focus cam tweaked for such "90.5mm" lenses. That marking is (or always has been) used for longer M lenses - so that techs can make accurate repairs 30-50 years later.

The "official" focal length of the 90mm Summarit, as listed in Leica's own data sheet, is "91mm," while the 90mm APO-Summicron-ASPH is listed as "90.9mm." The "official" 28 Summicron ASPH and 28 Elmarit ASPH focal lengths are listed as 28.5mm and 28.4mm, respectively. My 135mm APO-Telyt is nominally 136mm per the spec sheet, but is marked "57," so mine has an actual focal length of 135.7mm.

And all the "75mm" lenses I have used from Voigtlander are not marked, but include about 1% more in the picture than a Leica 75 does - thus are really about 73mm lenses.

And so it goes.

[Robert Blanko]

vor 8 Stunden schrieb Bgood:

You are really educating me. Thank you. I own a license to DxO PureRaw. I have attached a before and after and the amount of distortion astounds me. Is this what you are referring to?

 

https://photo.imx.nl/blog/files/75caecf5a21140ab6c9b46f825ef253c-143.html

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

vor 8 Stunden schrieb Bgood:

You are really educating me. Thank you. I own a license to DxO PureRaw. I have attached a before and after and the amount of distortion astounds me. Is this what you are referring to?

 

https://photo.imx.nl/blog/files/75caecf5a21140ab6c9b46f825ef253c-143.html

This photo was taken with the 35 APO? I don’t remember this huge amount of distortion in the photos taken with my 35 APO and this does not seem to be in line with the spec sheet ( https://leica-camera.com/sites/default/files/2021-09/DE_APO-Summicron_35_2.pdf ), but I would have to check my photos.

BTW, the correction of the Q2 is far more extreme, as the lens is rather a ~24mm lens, as one may notice in Capture One..

[Bgood]

This photo was taken with the Q3. I should have specified. 

[UliWer]

vor 5 Stunden schrieb Bgood:

This photo was taken with the Q3. I should have specified. 

 

If you have the chance to show an example taken with the Apo Summicron this would be very interesting. All I know from M-mount lenses - be it from Leica or other brands - is that they certainly do not depend on digital correction. Otherwise they wouldn‘t be usable with film, which they certainly are.

Lens design may allow a certain amount of  distortion to achieve better resolution. The old Sonnars for the Zeiss Ikon Contax designed by Ludwig Bertele were (in)famous for this. For their lenses to be used with film Leitz never went this way and accepted less resolution in favor of less distortion. The pre-asph. 35mm Summilux is a good example: low resolution wide open but almost no distortion. Leica still publishes the graphs which show the amount and character of distortion for their M-lenses, but never shows them for the Q- and SL-lenses. The reason is evident.

 

[Bgood]

48 minutes ago, UliWer said:

If you have the chance to show an example taken with the Apo Summicron this would be very interesting. All I know from M-mount lenses - be it from Leica or other brands - is that they certainly do not depend on digital correction. Otherwise they wouldn‘t be usable with film, which they certainly are.

Lens design may allow a certain amount of  distortion to achieve better resolution. The old Sonnars for the Zeiss Ikon Contax designed by Ludwig Bertele were (in)famous for this. For their lenses to be used with film Leitz never went this way and accepted less resolution in favor of less distortion. The pre-asph. 35mm Summilux is a good example: low resolution wide open but almost no distortion. Leica still publishes the graphs which show the amount and character of distortion for their M-lenses, but never shows them for the Q- and SL-lenses. The reason is evident.

 

I am attaching two sample images made with the 35APO and the M11P. The LR correction to my eye crops in a bit. The first image is DxO and the second is LR with the profile correction.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited March 30 by Bgood
Identify images

[lct]

1 hour ago, UliWer said:

All I know from M-mount lenses - be it from Leica or other brands - is that they certainly do not depend on digital correction.

Sure but with the same M lenses, Leica makes intensive use of digital correction for perspective control. Aside from electronic cropping, pictures don't seem to suffer from that.

[Advertisement]

Hallo,

Look here: Leica M

[3D-Kraft.com]

vor 2 Stunden schrieb UliWer:

All I know from M-mount lenses - be it from Leica or other brands - is that they certainly do not depend on digital correction.

Not really...
BastianK has shown, that the M10 applies digital corrections when using a Summilux-M 35 ASPH FLE - even to the raw files and even when “lens recognition” was turned off, see here: https://phillipreeve.net/blog/review-leica-35mm-1-4-summilux-m-asph-fle-mk-ii/ and search for "forced vignetting and color shift correction to the raw files".

Edited March 30 by 3D-Kraft.com

[Bgood]

20 minutes ago, 3D-Kraft.com said:

Not really...
BastianK has shown, that the M10 applies digital corrections when using a Summilux-M 35 ASPH FLE - even to the raw files and even when “lens recognition” was turned off, see here: https://phillipreeve.net/blog/review-leica-35mm-1-4-summilux-m-asph-fle-mk-ii/ and search for "forced vignetting and color shift correction to the raw files".

Thank you for sharing this. I reckon software  corrections are part of the landscape today. Some lenses have more and some have very little. 
 

I suppose I need to be more accepting of software corrections if I like the end result. After all in the 1960s software corrections weren’t an option.