Tonal Curve of Leica M Monochrom

[Martin B]

Advertisement (gone after registration)

I have a quick question in regard to the registration and processing of light within the sensor in any of the Leica M Monochrom models: in theory any digital sensor strictly has a linear response to the tonal range of the image (linear curve when plotting brightness of the incoming light versus brightness of the revealed final image). This is different from traditional B&W film photography where darker and brighter areas follow a non-linear response (S-Kurve). How is this done with the Leica Monochrom - is the dynamic or tonal range improved by some in-camera post processing of the (linear) tonal range, or is there some physical change done in the sensor itself (if this is even possible?) to use a non-linear tonal range response? Does anybody know?

 

Thanks,

Martin

[xenar]

The DNGs record linear samples. They do not have an embedded tone curve for gamma correction; the "standard" ACR tone curve would be used for this, I imagine. You would use something like Lightroom or what you will to apply further tone curves to the result.

 

This is for the CCD Monochrom; I haven't examined the 246 DNGs at all.

[mmx_2]

I would say that the 246 does exactly the same. The jpegs created from the in camera software is more "S-like". In lightroom it's very easy to get any shape you'd like, if you want it to look like film that is easyli done and if you prefer any other look that is of course just as easy

[Martin B]

Thanks, guys, this is the answer I expected and assumed from what I was reading about this camera. Correct, you can of course apply an S curve in post processing - I do this most of the time in my digital RAW files. It just would have been very interesting if Leica had found a way to get such S tonal curve directly in the camera or maybe even from the sensor - the latter is likely physically not possible.

 

By applying an S curve in post processing you still don't get the same tonal response as you get it from B&W film. Digital is limited in the tonal range (or zones) compared to film. An S curve applied in post processing makes it more look alike, but you still suffer from the linear shadow and highlight cutoff by the linear digital sensor which registers the image in the first place.

 

I asked this question because I am currently using my Leica M6 for B&W film photography which I enjoy a lot. B&W film has still the advantage over digital with its S-curve shaped tonal range response - it is very visible to me in those photos. You can also see it later in the prints of those files. The Leica M Monochrom comes closest what I would desire if I would consider moving or adding a Leica digital camera. It is certainly faster and more convenient than dealing with film (even I admit that I have a lot of fun to develop my own film). But now I am not sure if I really want to invest money for such camera if I lose the S tonal curve advantage by having the digital sensor. This point is very important to me.

[AceVentura1986]

I apply a curve on import in Lightroom, along with several other corrections, such as setting highlights at -25 and clarity at +20. The tonal curve I set on import is shaped like a candy cane, not too extreme but with a bias toward highlights and no change at all in shadows or blacks. Obviously, this is subject to change after import. 

[AceVentura1986]

Also, regarding losing the S tone curve, just set what you want as a default setting in Lightroom and import. It's super simple. 

[jaapv]

Advertisement (gone after registration)

Thanks, guys, this is the answer I expected and assumed from what I was reading about this camera. Correct, you can of course apply an S curve in post processing - I do this most of the time in my digital RAW files. It just would have been very interesting if Leica had found a way to get such S tonal curve directly in the camera or maybe even from the sensor - the latter is likely physically not possible.

 

By applying an S curve in post processing you still don't get the same tonal response as you get it from B&W film. Digital is limited in the tonal range (or zones) compared to film. An S curve applied in post processing makes it more look alike, but you still suffer from the linear shadow and highlight cutoff by the linear digital sensor which registers the image in the first place.

 

I asked this question because I am currently using my Leica M6 for B&W film photography which I enjoy a lot. B&W film has still the advantage over digital with its S-curve shaped tonal range response - it is very visible to me in those photos. You can also see it later in the prints of those files. The Leica M Monochrom comes closest what I would desire if I would consider moving or adding a Leica digital camera. It is certainly faster and more convenient than dealing with film (even I admit that I have a lot of fun to develop my own film). But now I am not sure if I really want to invest money for such camera if I lose the S tonal curve advantage by having the digital sensor. This point is very important to me.

I don't think you have a abrupt shadow cutoff on a digital file, the details kind of fade into the noise, much like film highlights, but you do have it on negative film the moment you block the shadows. In that sense the two react in reverse, with digital behaving like slide film.

[Martin B]

I don't think you have a abrupt shadow cutoff on a digital file, the details kind of fade into the noise, much like film highlights, but you do have it on negative film the moment you block the shadows. In that sense the two react in reverse, with digital behaving like slide film.

 

Nops, this is not correct. B&W film gives you a tonal response in the zones 1-10. In the shadow and highlight ending, the response it non linear (ending of the S curve). Only in the grey zones the response is linear similar to the digital sensor. Any digital sensor only displays zones from 3-8 (that's why often bracketing is needed). The response is strictly linear - that's why you can get easily clipped highlights or pitch black unstructured shadows. Or in other words: The number of photons registered inside the photon wells of a digital sensor is always linear to the amount of light/brightness displayed in the image. This is not the case with film where the chemical process in the darker and brighter areas is non linear. There are ways in digital to minimize this of course - like applying a curve into the linear tonal response setting for example. This still doesn't change your limited zone response compared to film.

[Martin B]

Also, regarding losing the S tone curve, just set what you want as a default setting in Lightroom and import. It's super simple. 

 

No, not the same at all. See my response above which explains why. You explain correctly what can be done to change the linear digital response through post processing. But you still are limited in the first place by getting a zone cutoff in the dark and bright areas. That's why film is not the same as digital - physically it can't be. Only if the sensor would somehow register photons/incoming light in a non linear response in the bright and dark areas, yes, then it would be the same.

[xenar]

Nops, this is not correct. B&W film gives you a tonal response in the zones 1-10. In the shadow and highlight ending, the response it non linear (ending of the S curve). Only in the grey zones the response is linear similar to the digital sensor. Any digital sensor only displays zones from 3-8 (that's why often bracketing is needed). The response is strictly linear - that's why you can get easily clipped highlights or pitch black unstructured shadows. Or in other words: The number of photons registered inside the photon wells of a digital sensor is always linear to the amount of light/brightness displayed in the image. This is not the case with film where the chemical process in the darker and brighter areas is non linear. There are ways in digital to minimize this of course - like applying a curve into the linear tonal response setting for example. This still doesn't change your limited zone response compared to film.

This would be the case if digital sensors were somehow recording less light than film, and I'm not convinced this is the case.

 

Linearity of sensors vs. nonlinearity of film doesn't necessarily demonstrate that sensors capture less light. In fact, I'm pretty sure to argue they are commensurate is by looking at equivalent exposures and ISO settings.

[zlatkob]

Nops, this is not correct. B&W film gives you a tonal response in the zones 1-10. In the shadow and highlight ending, the response it non linear (ending of the S curve). Only in the grey zones the response is linear similar to the digital sensor. Any digital sensor only displays zones from 3-8 (that's why often bracketing is needed). The response is strictly linear - that's why you can get easily clipped highlights or pitch black unstructured shadows. Or in other words: The number of photons registered inside the photon wells of a digital sensor is always linear to the amount of light/brightness displayed in the image. This is not the case with film where the chemical process in the darker and brighter areas is non linear. There are ways in digital to minimize this of course - like applying a curve into the linear tonal response setting for example. This still doesn't change your limited zone response compared to film.

 

 

Any digital sensor only displays zones from 3-8?  Where do you get this info?  Digital has amazing detail in the darker tones.  That's one of the advantages of digital over film.  Digital is linear, but digital records a lot, especially the better sensors.

[Giulio Zanni]

In Silver Effex you can replicate with a click most of the BW films totals curves 

[muri]

 

More info

[Martin B]

Any digital sensor only displays zones from 3-8?  Where do you get this info?  Digital has amazing detail in the darker tones.  That's one of the advantages of digital over film.  Digital is linear, but digital records a lot, especially the better sensors.

 

Film registers more than any digital sensor can do. We need to distinguish between popping up information in an existing digital file by post processing and what film captures all at once without need of post processing. The limited zone range from 3-8 in digital is well described in the literature, for example here:

 

In the paragraph of dynamic range - comparison between film and digital:

http://photography.tutsplus.com/articles/light-photography-exposure-and-tonal-range-considerations--photo-5685

 

There is a quite good graph referring to what I expressed above with the limited DR of digital vs film under "Non-linear Film Photography Processes":

http://thucydides.sjca.edu/~stars/ccd.html

 

The handbook of photography from Marchesi also explains this difference.

[Martin B]

In Silver Effex you can replicate with a click most of the BW films totals curves 

 

Comes back to the same point mentioned above: this is post processing of a registered digital image which has a limited DR in the first place. You can pull more information out of the shadows and highlights by applying an S curve onto the linear tonal curve in digital, but you still suffer the cutoff in the S-curve endings in dark and bright compared to film.

 

I am heavily using Silver Efex for my digital files - it helps a lot, but it can't fully replicate B&W film for reasons mentioned above.

[Martin B]

This would be the case if digital sensors were somehow recording less light than film, and I'm not convinced this is the case.

 

Linearity of sensors vs. nonlinearity of film doesn't necessarily demonstrate that sensors capture less light. In fact, I'm pretty sure to argue they are commensurate is by looking at equivalent exposures and ISO settings.

 

It depends on the situation and composition if the difference is seen between B&W film and monochrome digital. In some situations they are hard to distinguish, in others I was able to capture a scene in one single shot on film where I clearly would have needed to use bracketing or HDR to get to something similar in digital. I am not saying that film is better than digital, I am saying that film has a DR advantage because of its non linear response to light.

[Martin B]

 

More info

 

Not sure how this is related. You are plotting wave length vs. response to the given energy/wavelength. I am talking about the tonal latitude of film versus digital sensor, not about UV/visible/IR light.

[Giulio Zanni]

Comes back to the same point mentioned above: this is post processing of a registered digital image which has a limited DR in the first place. You can pull more information out of the shadows and highlights by applying an S curve onto the linear tonal curve in digital, but you still suffer the cutoff in the S-curve endings in dark and bright compared to film.

 

I am heavily using Silver Efex for my digital files - it helps a lot, but it can't fully replicate B&W film for reasons mentioned above.

 

I am not tech guy, but I am coming from film. With good postprocessing skills I can get more from the monochrom files than I what I was getting on film....

[Martin B]

I am not tech guy, but I am coming from film. With good postprocessing skills I can get more from the monochrom files than I what I was getting on film....

 

This might be true and could potentially be related sort of to my initial question. I definitely can't get a better output from my B&W digital files compared to B&W film - I tried it several times. But I also admit that I don't have a Leica Monochrom - best I could compare with is the Sony A7R which has a pretty good FF sensor. Its monochrome JPG mode is already very good. But maybe there is something in the Leica Monochrome that the RAW files are automatically in camera "pushed" to deliver better results similar to film. Possible that Leica found a way to simulate in camera already sort of the film effect - this is exactly what I am interested in to know. We already can exclude that Leica would use a special sensor - from earlier responses it is clear that the CCD sensor has a linear output as other digital sensors have. This only leaves the option of some in-camera processing of the RAW files - potentially, I don't know if this is the case or not. Whatever you can do in post processing afterwards is not of interest to me - I am very familiar with those things, but so far I could never overcome just with post processing tools the range I can achieve with film. This might be no issue for some, but it is certainly for me - or saying it in other more direct words: why should I spend $$$$ for a Leica M if I get a similar monochrome effect on my A7R and still not the same tonal range as from real B&W film (assuming that film of course has other debits compared to digital, but this is not the question here).

[Advertisement]

Hallo,

Look here: Leica M

[otto.f]

Film registers more than any digital sensor can do. We need to distinguish between popping up information in an existing digital file by post processing and what film captures all at once without need of post processing. The limited zone range from 3-8 in digital is well described in the literature, for example here:

 

In the paragraph of dynamic range - comparison between film and digital:

http://photography.tutsplus.com/articles/light-photography-exposure-and-tonal-range-considerations--photo-5685

 

There is a quite good graph referring to what I expressed above with the limited DR of digital vs film under "Non-linear Film Photography Processes":

http://thucydides.sjca.edu/~stars/ccd.html

 

The handbook of photography from Marchesi also explains this difference.

 

You could say that in an ideal world of very exact exposure of the film given the dynamic range of the scene in reality and the ideal development given that exposure to the film in that situation. Moreover, this ideal situation is very hard to reach on 35mm film, you would at least need 4*5 inch.

Also, I am not so sure whether you are comparing here with sensors with or without the Bayer filter

 

I am not tech guy, but I am coming from film. With good postprocessing skills I can get more from the monochrom files than I what I was getting on film....

 

Me either and this is my experience too, which I cannot theoretically combine with MartinB's statement. And I wonder if MartinB has practical experience with the MM1. 

One has to consider the end-product too, is it paper or transparencies, like screen or slide. A transparency or good screen has one or two more stops to encompass the DR of reality.

In my poor technical wisdom the S-curve is just a way of forcing the dynamic range of reality into the dynamic range of the medium that is going to represent this reality in a reduced manner. 

So, stating that an S-curve is better than linear,... I don't know