Amazing color difference 90 cron to 35 Lux

[flyalf]

Advertisement (gone after registration)

I have been testing my new SH 90 Cron APO in difficult winter light up here in North of Norway. I was simply stunned about the difference in color between these shots. Yeah I knew that the winter light when sun is well below horizon is strange, but would never have guessed how different the lenses saw this. Same camera, identical LR settings (color temperature, a bit saturation enhancemnet, no adjustment in color channels):

 

Edit: 90mm with Leica UV filter.

 

90 Cron APO fully open:

L1004569 by fotoalf, on Flickr

 

35 Lux ASPH FLE fully open:

L1004567 by fotoalf, on Flickr

Edited December 14, 2014 by flyalf
Added UV filter info.

[lct]

Auto WB?

[Advertisement]

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

[richfx]

Which is truer to life?

[gpwhite]

There are a number of variables between the two shots, so I am not greatly surprised.

 

The SX 35 FLE is a warmish lens (although not as warm and pleasing to me as the SX 35), whereas the AA 90 is surely a cool lens. Puts once published a subjective scale for warmth/coolness of Leica M lenses, and, as I remember, the SX 35 and AA 90 were at opposite ends of the range.

 

I find the AWB on my M240 (and the M9 before it) occasionally loses its bearings in cold light. This is why the white/gray picker tool is so valuable in ACR

[CheshireCat]

Same camera, identical LR settings (color temperature, a bit saturation enhancemnet, no adjustment in color channels):

 

Edit: 90mm with Leica UV filter.

 

Same color temperature AND tint ?

In any case, some UV filters may impact color balance.

[billo101]

My 35 FLE dosen't make this vignetting like this!?

 

b

[younes]

Advertisement (gone after registration)

What camera was used ?

 

The red tint and vignetting with the 35/1.4 looks like what you get with many lenses, especially wides, used wide open on non Leica sensors, for example Sony A7. Differences with longer focal lenses tend to disappear when using smaller apertures.

[rscheffler]

While I'm highly impressed with the Leica lenses I have, one thing they are not great at is overall color consistency. I would classify the 21 Lux, 21 SEM and 50 Lux ASPH as 'neutral' whereas the 28 Cron has a noticeable magenta tint, the 90 Summarit a somewhat more subtle magenta tint and the 90AA a cyan shift. It makes equalizing all of them a bit of extra work in post.

[lct]

I don't see any significant colour inconsistency between 35/1.4 FLE and 90/2 apo personally but i may be wrong. I would redo the comparo in manual WB mode and w/o filter, or with the same filter on both lenses to be sure.

[flyalf]

Thanks to all for answers!

 

Auto WB?

Yes, but chosen same color temperature and tint in LR, so shouldnt make that much difference.

 

Which is truer to life?

None, but 35 Lux is closer to what I want, and for me a tad easier to fix in LR

 

Same color temperature AND tint ?

In any case, some UV filters may impact color balance.

 

Yes. Synced and checked afterwards.

 

 

My 35 FLE dosen't make this vignetting like this!?

 

b

Hmm, it probably does, but you will not normally not see the difference except for white-out conditions like this.

 

What camera was used ?

 

The red tint and vignetting with the 35/1.4 looks like what you get with many lenses, especially wides, used wide open on non Leica sensors, for example Sony A7. Differences with longer focal lenses tend to disappear when using smaller apertures.

Sorry, I forgot. This is a M (240), and the profiles were chosen in-camera.

[luigi bertolotti]

....

 

Auto WB ?

 

Yes, but chosen same color temperature and tint in LR, so shouldnt make that much difference.

....

 

.

 

I think THAT is the issue : the picture taken with the 90 I think is more influenced by the great icy/snowy foreground , in comparision with the picture taken with the 35 for which the Auto WB takes more in account the grey sky (its area is clearly larger, in proportion : I suppose that you attached the whole frames, of course): setting in LR developer the same temperature applies to the DNG file the same correction, resulting in a file with a different balancing (I often observed this fact in pictures taken in interiors with artifical light) ; as said above, a real test on the difference from the two lenses (which can anyway exist) ought to be carried on setting a certain temperature on the camera (and keeping the same in developing)

The UV filter can also be a cause, in a picture that is full of light reflected by a perfectly White surface.

Edited December 17, 2014 by luigi bertolotti

[LocalHero1953]

I think THAT is the issue : the picture taken with the 90 I think is more influenced by the great icy/snowy foreground , in comparision with the picture taken with the 35 for which the Auto WB takes more in account the grey sky (its area is clearly larger, in proportion : I suppose that you attached the whole frames, of course): setting in LR developer the same temperature applies to the DNG file the same correction, resulting in a file with a different balancing (I often observed this fact in pictures taken in interiors with artifical light) ; as said above, a real test on the difference from the two lenses (which can anyway exist) ought to be carried on setting a certain temperature on the camera (and keeping the same in developing)

The UV filter can also be a cause, in a picture that is full of light reflected by a perfectly White surface.

 

I'm puzzled by this, but I'm prepared to be educated. My understanding was that you could use whatever WB setting you wished in-camera (including Auto WB), but the setting in LR overrode the in-camera setting. Indeed, I understand that you are not applying a correction in LR, but making a absolute setting to be applied in the raw conversion. Your argument implies that if you changed WB from tungsten to fluorescent between two shots of the same scenes, you would have to use two different WB settings in LR to achieve the same visible colour balance on-screen, which is not my experience.

 

I almost always shoot AutoWB (except with flash), but only to get the initial look right for quick appraisal. I almost always override WB temp and tint in LR. I have never seen this approach cause the colour anomaly shown in the OP's shots, though I do not often shoot such colourless scenes.

 

I agree that fixing the WB in-camera would be a useful test, though.

Edited December 17, 2014 by LocalHero1953

[CheshireCat]

In all normal cameras, Auto WB just writes two values in the file: temperature and tint, without touching any data from the sensor.

The raw developer then applies white balance based on these two values, and if the user overrides these values, then similar lenses should produce similar colors.

 

I guess the problem here is that the M raw files are not exactly raw. They are preprocessed based on the lens model detected or set. At least, the camera tries to remove color vignetting, and may possibly perform other kinds of unknown corrections which may be dependent on the WB values.

 

I would try to disable the lens detection, which should hopefully disable any kind of lens-model based processing, and set the same WB in camera. Then repeat the experiment.

[Erik Gunst Lund]

It would be so easy to see what is going on if you uploaded the two raw files, then we could have a look and help explain.

Edit: I'll have a look at the jpg files, and agree that it's probably due to the difference in focal length and the UV filter

Edited December 18, 2014 by Erik Gunst Lund

[Erik Gunst Lund]

I cropped the 35mm shot to the same FOV as the 90mm shot, so the house, trees and mountain range in the BG are similar, then individually setting the WB for the snow on the roof, then two images are almost idential slight differences in CA gives some different colour issues but minor...

 

IMHO it has nothing to do with the glass of the lens.

 

Thanks for posting these! Very educational observation !

Edited December 18, 2014 by Erik Gunst Lund

[luigi bertolotti]

It would be so easy to see what is going on if you uploaded the two raw files, then we could have a look and help explain.

Edit: I'll have a look at the jpg files, and agree that it's probably due to the difference in focal length and the UV filter

 

This is also a factor, maybe... I mean (and is an intriguing matter...) : let's think a little... a picture taken with a tele (90 mm is not so long... but we are comparing to 35mm) within an atmosphere that probably is someway foggy.. with many microdrops of humidity in suspension... I think that the "color" of the air is different (with strong teles is a well known effect) ; well... maybe this consideration is vague and unsupported by scientific facts... at the end, the only trivial truth is that if you want to evaluate the difference in color rendering from two lenses... it must be a lab test with controlled environment, proper targets, strict workflow...

[luigi bertolotti]

Always on the matter of color difference from one lens to another... it came to my mind an old test I made at the times of my M8 and my then new 21 2,8 asph, compared to the Super Angulon 21 3,4... both at f8, artificial home-light, same WB setting, same development in LR :

 

Elmarit

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!

Super Angulon

 

is really different : the asph was coded, and lens recognition ON, but I doubt this is the reason : I'm more prone to think of two facts, considering that any manufacturer probably tends to give a certain consistent color rendering to its lenses' lineup :

 

- There are more than 40 years between the two designs... times and tastes change... (this clearly isn't the case for the OP lenses, which are both from today)

- The Super Angulon was entirely a Schneider design, and they could pursue a slightly different standard rendering from Leitz

 

Btw, it was rather easy to turn the two images in identical color.

Edited December 18, 2014 by luigi bertolotti

[otto.f]

Oh that's just the M240's white balance, nothing special about the lenses. The 90 selects a different part of the landscape to draw his auto WB from. I bet you won't see this difference between the lenses on an M9. Plus the UV-filter on the 90 and not on the 35

Edited December 18, 2014 by otto.f

[luigi bertolotti]

Oh that's just the M240's white balance, nothing special about the lenses. The 90 selects a different part of the landscape to draw his auto WB from. I bet you won't see this difference between the lenses on an M9. Plus the UV-filter on the 90 and not on the 35

 

But... is this related to lens' recognition ? I mean , if Chesire Cat assertion is correct (WB data are simply stored - no influence on the raw file - and you can set them at the same value in raw developer) you ought to achieve anyway the same balancing... unless, as speculated in his post, it is lens recognition that "casts" a different balancing directly on the raw file.

[Advertisement]

Hallo,

Look here: Leica M

[Erik Gunst Lund]

I can recommend to download the two files an try to make them look alike, similar WB. You will be surprised how different the two images are

 

I had a hard time doing it, give it go and post...