Leica M11 -purplish tint ???

[la1402]

Advertisement (gone after registration)

11 minutes ago, jaapv said:

 In digital photography the file out of the camera is nothing more than a half-product. 

Certainly true for the M11. With some other cameras, it can be closer to three quarters of a product. 

[jaapv]

2 minutes ago, la1402 said:

Certainly true for the M11. With some other cameras, it can be closer to three quarters of a product. 

Not in my book - I leave no image unedited. 

[Adam Bonn]

On 11/23/2023 at 3:26 AM, adan said:

Adobe Camera Raw offers the option to convert/save other formats as .DNGs.

Indeed, but you would need some profile there already (the adobe one probably) in order to open it in the first place. Adobe recommends their DNG converter to turn RAW into DNG if you’re profiling (or even working with) a format not supported by one’s incumbent version of ACR

On 11/23/2023 at 3:26 AM, adan said:

But it also reads them natively which is why it is constantly being updated for new cameras and their formats.

Those that refuse the subscription model would not agree 😉

On 11/23/2023 at 3:26 AM, adan said:

Simply slide the calibration-panel RGB-primaries hue/sat sliders around until the ColorChecker primaries in the image match the published RGB values they should have

I think someone wrote a PS plugin for that.. that’s exactly what those sliders are there for (I posted more about that above). Of course you understand this is tweaking a profile, not creating one from scratch. (Actually tweaking the XYZ D50 Transform. See the DNG spec for boring details and a lot of linear algebra 😅)

On 11/23/2023 at 3:26 AM, adan said:

One obvious hint being the "As Shot" WB metadata. The camera-written metadata includes the information that, for example, "This camera used a WB at 5500K + tint of +35 magenta, for this particular picture, for producing jpegs and previews."

It’s entirely possible that many RAW apps use this, or perhaps adobe does for non-DNG formats… dunno. In a native Leica DNG/adobe pipeline the WB is calculated via the ASN tag* and a weighted value derived between the inverse of the ColorMatrices this is certainly the route that Leica take (again see DNG spec). DNG also allows the camera maker to specify an xy value for RAW CCT, but Leica don’t use it. 

(*the white point of the non-white balanced RAW data basically)

On 11/23/2023 at 3:26 AM, adan said:

Of note, I shoot my ColorChecker sample images at: high noon ± 1 hour, with the CC in direct sunlight, and far away from anything reflecting any color that might skew the results (often in the middle of a wide street). That way I am pretty sure I am sampling true "solar spectral white," with as little Rayleigh Scattering and other color distortions as possible.

Sounds good to me. If the colorchecker app was my sole tool for this I’d just make a singe D50 matrix profile, and accept that WB tweaks in other light are a necessary part of workflow (which they are anyway)

The trouble with dual illuminant home brew profiles is that StdA (2800k more or less) doesn’t begin to cover the wide array of artificial light in the modern world and reliable D65 is quite hard to come by. 

Paid profile creation apps tend to have a few tricks here to give a helping hand (mine lets you shoot 2 shots of each target used, the regular one then one where the target is obscured by white/neutral paper/card/whatever. It then examines both shots to determine what’s colour from the target, what’s colour from the environment and also to correct for any glare/vignetting.

iirc.. that Hogan fella who’s blog I linked too before goes out into the mountains, measures the light with a spectrometer, shoots his target, then comes home and maps the RAW values of his camera against the CC chart, then writes the profile manually using maths.

That sounds like a fun project.. but wives can get pissed off y’know? 😅

Fundamentally the color matrix (or matrices) are created around set colour temps, the moment we use light that’s different (pretty much every shot away from a studio) there’s a degree of CCT editing needed in post… 

As is the backbone of the advice in this thread that I can only reiterate, a profile is the starting point. Leica’s one, adobe’s one, the one you bought from cobalt or whatever, your own one.. all a starting point, not the finished article. Between those options one can find the best foundation which won’t even be the same profile for every image, but then one still needs to build on top of it.

[lct]

1 hour ago, la1402 said:

Certainly true for the M11. With some other cameras, it can be closer to three quarters of a product. 

Been using several digital cameras since they exist (Leica, Sony, Canon, Nikon, Fuji, Sigma,Ricoh...) but i got this feeling only once: on my good old Canon 5D1. Perhaps also on a Foveon based Sigma. Now this does not prove anything objective of course, just that my subjective tastes match more those of Canon some 15 years ago.

Edited November 24, 2023 by lct

[Adam Bonn]

22 hours ago, jaapv said:

Oh, I would be surprised if Leica did not change their basic profile in due course. It is not a very difficult tweak and they like to keep the customers happy. 

Me neither, but they won't change their profile, they'll change their algorithm that produces the WB values to work with their existing profile. (Like they did in the M240 and M10)

1 hour ago, Edax said:

Strange thing is that in Lightroom Classic WB settings change depending on origin of selected profile;

A "random" M11-P AWB picture opens with Adobe Color profile, gets WB setting 5000/23. Change to other Adobe profile, WB stays at 5000/23.

Changing profile to profile M11 changes WB to 5200/22. Change to a Cobalt profile, and WB values are same as with profile M11, 5200/22.

Opening same picture in Iridient, WB values become 5189/22.

The Lightroom fixed WB "outside" settings (daylight, cloudy, shade) set the tint to +10, visibly a much better starting point.

That's expected behaviour... in adobe

In short WB = RAW ASN value x ColorMatrix

As Leica and adobe use different ColorMatrices then the same ASN multiplied by different numbers (CMs) equals a different answer.

Cobalt like to reuse the adobe ColorMatrices to provide consistent WB values when swapping between profiles. (ColorFidelity used to reuse the Leica CMs for the same reason)

If one's profile contains forward matrices and profile LUTs... then the ColorMatrices can pretty much be whatever you like (the FMs/LUTs are doing the heavy lifting)

23 hours ago, elmars said:

Over the last few days I have been making my annual family album. I mixed photos from the M11 with some from the iPhone. The magenta cast of the M11 photos is clearly noticeable and I can't bear it without correction. In the vast majority of cases, I adjusted the M11 photos because the magenta cast was distracting (tint from +20 to +10).

Even without comparison with other cameras, there are many subjects that do not tolerate magenta. This includes everything that has a lot of grey areas, but also sandstone. Here I find the magenta downright unbearable, whereas with grey it can even be interesting. See the example picture with sandstone below. First: tint as shot +20; second: tint +10.

 

 

 

 

Are you saying that the M11 always has a tint of +20 even on AWB (which would be wrong) or always has a +20 tint when you shoot every photo with the same static WB setting (which would be expected, -although perhaps at +20 for midday sun, undesirable- behaviour)

[Adam Bonn]

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!

Here's a throw away (although I kept it!) AWB snap from my M9.

Maybe the global photographic community just spent too many years extolling the great virtues of the M9 colour palette and Leica actually listened... 😂😂

Edited November 24, 2023 by Adam Bonn

[elmars]

Advertisement (gone after registration)

vor 2 Stunden schrieb jaapv:

Not in my book - I leave no image unedited. 

You are not the whole world. Others like more usable DNG - even from Leica.

[don daniel]

vor 1 Stunde schrieb Adam Bonn:

Are you saying that the M11 always has a tint of +20 even on AWB (which would be wrong) or always has a +20 tint when you shoot every photo with the same static WB setting (which would be expected, -although perhaps at +20 for midday sun, undesirable- behaviour)

M11 always hast a tint of +20 or even more, when you shoot at daytime in natural light on AWB. And it has such a magenta tint also when you shoot in the static settings for sun, cloud, shade. This is the problem we have been talking about the whole time! And this is also what some people do not see or they do not mind or they even like it. M11 only has a lower tint on AWB, when you shoot in artificial light. Then it can be precise.

The JPGs show the same magenta cast no matter which program or viewer you use. You even see it on the display of the camera itself. 

Edited November 24, 2023 by don daniel

[Adam Bonn]

9 minutes ago, don daniel said:

M11 always hast a tint of +20 or even more,

So the answer I was looking for is

no, when using AWB the M11 does not always produce a tint value of exactly +20

it's capable of selecting different values as it sees fit.

Thank you

Leica changed the WB in the M240 and the M10, keep on at Leica until they do so with the 11.

 

[don daniel]

vor 41 Minuten schrieb Adam Bonn:

no, when using AWB the M11 does not always produce a tint value of exactly +20

it's capable of selecting different values as it sees fit.

Sorry, yes, this is correct.

[elmars]

vor 4 Stunden schrieb Adam Bonn:

Are you saying that the M11 always has a tint of +20 even on AWB (which would be wrong) or always has a +20 tint when you shoot every photo with the same static WB setting (which would be expected, -although perhaps at +20 for midday sun, undesirable- behaviour)

Even if Don Daniel has already answered: 

With AWB the values vary (a little), with fixed white balance they do not. JPGs from the camera also have the magenta cast, so it's not an Adobe problem. Thanks for your assessment that +20 is too much for midday sun.

May I also finally say that I am always impressed by your knowledge, even if I probably don't understand everything. No one in the forum is as knowledgeable as you in this area.

[jaapv]

JPG or DNG does not make a difference if both camera and adobe use the same base to convert the data out of the sensor. As Adam explained, the origin of those data lies in an earlier stage of the processing pipeline. Basically whether DNG conversion or OOC JPG show the same deviation tells us very little.

[Jeff S]

6 hours ago, Adam Bonn said:

As is the backbone of the advice in this thread that I can only reiterate, a profile is the starting point. Leica’s one, adobe’s one, the one you bought from cobalt or whatever, your own one.. all a starting point, not the finished article. Between those options one can find the best foundation which won’t even be the same profile for every image, but then one still needs to build on top of it.

This.  And it’s very easy to change, if desired. 
 

Jeff

[Adam Bonn]

43 minutes ago, elmars said:

Even if Don Daniel has already answered: 

With AWB the values vary (a little), with fixed white balance they do not. JPGs from the camera also have the magenta cast, so it's not an Adobe problem. Thanks for your assessment that +20 is too much for midday sun.

Leica will need to tweak their algorithm that makes the ASN tag (like they have before with other cameras)

Midday sun on a white paper, midday sun in a green forrest maybe not so much.

44 minutes ago, elmars said:

May I also finally say that I am always impressed by your knowledge, even if I probably don't understand everything. No one in the forum is as knowledgeable as you in this area.

Thank you. Learning this was a relationship damaging Covid lockdown project. To be honest my knowledge is pretty basic, helicopter view on the theory, I can't do most of the maths (I can make FMs though), but I have a sense of how it all works, what all the nut and bolts do... I think others here will know more, a lot more, but are smart enough to be quiet about it.

 

[Adam Bonn]

Speaking of profiles... anyone know if the Cobalt one fixes it/provides more pleasing results (delete as applicable as to whether you perceive it to be a problem or not)? I get the sense from their marketing that it should... 

[Adam Bonn]

25 minutes ago, jaapv said:

JPG or DNG does not make a difference if both camera and adobe use the same base to convert the data out of the sensor. As Adam explained, the origin of those data lies in an earlier stage of the processing pipeline. Basically whether DNG conversion or OOC JPG show the same deviation tells us very little.

In a highly woolly analogy... a camera profile / the DNG pipe line is a little like a relay race...

The first runner sprints to the second, hands over the baton, who in turn sprints to the 3rd...

The first runner of DNG is the ASN (or whatever other DNG spec compliant tag is used, but for Leica it's ASN)

This rather screws the race for everyone else in the team if the first runner is lousy...

Luckily (unlike a real relay race) this thread not only has me explaining DNG nuts and bolts, but a LOT of really practical advice on editing DNG for WB.

Read those posts before mine folks!

[bcaslis]

1 hour ago, Adam Bonn said:

Speaking of profiles... anyone know if the Cobalt one fixes it/provides more pleasing results (delete as applicable as to whether you perceive it to be a problem or not)? I get the sense from their marketing that it should... 

First image RAW in AWB with M11 profile exposure -.35 EV

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!

 

Same image, same exposure with profile set to Cobalt Standard.

 

[Adam Bonn]

2 hours ago, bcaslis said:

First image RAW in AWB with M11 profile exposure -.35 EV

 

 

Same image, same exposure with profile set to Cobalt Standard.

 

 

I'm on my MacBook.. so not ideal... but I'm filing that under improvement to Cobalt, I don't think the sky is magenta free, but globally the image is cooler

[don daniel]

Both images full of magenta! Just take the tint slider 10 points to the left and it will be fine. The first image at least shows vivid colours, the second one with the Cobalt profile seems rather dull too me. Cobalt profiles are not the solution of the problem. Only adjusting white balance in post will help.

Edited November 24, 2023 by don daniel

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[bcaslis]

Just suppling the data. My final image is neither of these but the image looks like when it was taken. November heading towards winter with the sun starting to slant down. You may not like it, but it doesn't look incorrect to me.