Leica M11: A Revolution – Review by Elmar S.

[SrMi]

Advertisement (gone after registration)

Just now, hdmesa said:

I haven't used the M11 yet, but its the same sensor as the GFX 100/S cut smaller. On the GFX 100S, I can overexpose the heck out of an image and easily recover the highlights. I'm wondering if those needing some negative exposure comp on the M11 are using the M10/R as a guide for what will be recoverable in post. It could be the exposure meter is giving a nearly-perfect ETTR image. One could examine the M11 images in Raw Digger to know for sure.

It may be the same type of sensor (also found in Sony a7rIV). But highlights recovery are not a property of the sensor, but the camera: metering and processing of data.

[hdmesa]

Just now, SrMi said:

It may be the same type of sensor (also found in Sony a7rIV). But highlights recovery are not a property of the sensor, but the camera: metering and processing of data.

I'm saying we need to be sure that images taken with the M11 that are thought to need negative exposure comp actually need it. Reviewing them in Raw Digger will reveal if they are truly overexposed or not.

[mike3996]

On 1/17/2022 at 4:30 AM, hdmesa said:

I'm saying we need to be sure that images taken with the M11 that are thought to need negative exposure comp actually need it. Reviewing them in Raw Digger will reveal if they are truly overexposed or not.

It's difficult to objectively estimate highlight recovery capabilities. It depends how you meter, what's your scene, how you like the scene to look like (high key, low key), do you go for ETTR, etc.

Somebody were to release a hypothetical user report about "how +1.33 stops overexposure is still recoverable", that tells absolutely nothing.

With these things you just have to spend a couple of hours with the camera and develop a feel to the raw files.

[hdmesa]

13 hours ago, mike3996 said:

It's difficult to objectively estimate highlight recovery capabilities. It depends how you meter, what's your scene, how you like the scene to look like (high key, low key), do you go for ETTR, etc.

Somebody were to release a hypothetical user report about "how +1.33 stops overexposure is still recoverable", that tells absolutely nothing.

With these things you just have to spend a couple of hours with the camera and develop a feel to the raw files.

All I was trying to say – don't assume you have optimal exposure unless you verify in Raw Digger. Especially don't assume based on how you exposed images with the M10 variants.

[jonoslack]

20 minutes ago, hdmesa said:

All I was trying to say – don't assume you have optimal exposure unless you verify in Raw Digger. Especially don't assume based on how you exposed images with the M10 variants.

Or maybe you just take pictures? Isn't it what they look like rather than what the numbers are? 

[Olaf_ZG]

9 minutes ago, jonoslack said:

Or maybe you just take pictures? Isn't it what they look like rather than what the numbers are? 

Huh? You want me to actually use the m11?

[hdmesa]

Advertisement (gone after registration)

2 minutes ago, jonoslack said:

Or maybe you just take pictures? Isn't it what they look like rather than what the numbers are? 

I didn't mean to suggest to pull every file through Raw Digger, that would be crazy I meant it's a good idea when getting a new sensor to check some files in Raw Digger to see how your ETTR is doing (if you are someone who does ETTR) until you get an idea of how far you can push the exposure to the right with the M11 and not blow the highlights. I only mentioned this since some people were posting that they were doing ETTR on the M11 based on the amount of + exposure they were doing for their M10.

It would be a lot simpler if Leica gave us a proper RGB histogram. Right now there is no way to trust the histogram or blinking highlight warnings because a runaway channel not shown in the average can still blow out.

But to your point, yes this new BSI sensor with high DR allows us to shoot with much fewer worries than before.

[jonoslack]

On 1/13/2022 at 5:44 PM, elmars said:

There are two meanings in German for this: Frivolous and superficial. Because I know how hard You worked on the M11 it must be the first one.

Hi Elmar

Indeed frivolous will do - but I've just read your report again, and it's both comprehensive and succinct!

I hope you're enjoying your new M11 (I've still got Leica's one hostage, but I'll have to pay the piper in the end!)

 

[jonoslack]

9 minutes ago, hdmesa said:

I didn't mean to suggest to pull every file through Raw Digger, that would be crazy I meant it's a good idea when getting a new sensor to check some files in Raw Digger to see how your ETTR is doing (if you are someone who does ETTR) until you get an idea of how far you can push the exposure to the right with the M11 and not blow the highlights. I only mentioned this since some people were posting that they were doing ETTR on the M11 based on the amount of + exposure they were doing for their M10.

It would be a lot simpler if Leica gave us a proper RGB histogram. Right now there is no way to trust the histogram or blinking highlight warnings because a runaway channel not shown in the average can still blow out.

But to your point, yes this new BSI sensor with high DR allows us to shoot with much fewer worries than before.

I think I feel that all modern cameras (well, all good ones) have magnificent DR, and as a photographer you just need to make sure you don't blow the highlights (and everything else follows). I thought ETTR was out of date ten years ago, but now, with 14 stops of DR it seems absolutely pointless - there is infinite detail in the shadows in modern sensors, and my feeling is that the only trap to fall into is to blow the highlights! Why do you need to push the exposure to the right? 

I just don't think you need a better histogram - If you just make sure nothing is drastically over-exposed everything else is fine!

All the best

[SrMi]

33 minutes ago, jonoslack said:

Or maybe you just take pictures? Isn't it what they look like rather than what the numbers are? 

In other words, not every picture needs to be optimally exposed.

[hdmesa]

1 minute ago, jonoslack said:

I think I feel that all modern cameras (well, all good ones) have magnificent DR, and as a photographer you just need to make sure you don't blow the highlights (and everything else follows). I thought ETTR was out of date ten years ago, but now, with 14 stops of DR it seems absolutely pointless - there is infinite detail in the shadows in modern sensors, and my feeling is that the only trap to fall into is to blow the highlights! Why do you need to push the exposure to the right? 

I just don't think you need a better histogram - If you just make sure nothing is drastically over-exposed everything else is fine!

All the best

As I mentioned, someone was talking about using exposure comp based on what they did with their M10, which I said wasn't a good assumption since this is a new sensor.

As for ETTR, every shot has an optimal exposure – that may require moving to the right of the meter, it may require moving to the left. You can still make a good image if you're one stop below optimal, but you could make a better image with an optimal exposure. Modern sensor are indeed better, but at high ISO, noise in pulled shadows is still an issue that can be reduced by getting the best exposure possible. ETTR requires zero overhead/thought process once you know how the meter in the camera works compared to how the histogram looks in Raw Digger. Why not look at a few RAWs from high contrast scenes and see if you could be letting in more light without blowing the highlights? 

[hdmesa]

2 minutes ago, SrMi said:

In other words, not every picture needs to be optimally exposed.

Agreed. I try to think more about optimal exposure when I think it will make for a significantly better image (high ISO where I know I will pull the shadows, or a landscape shot I think might be printed 40-60" wide, etc.). For all other shots, I take more of a casual approach – and the sensor in the M11 allows that approach with fewer negative consequences.

[SrMi]

10 minutes ago, jonoslack said:

I think I feel that all modern cameras (well, all good ones) have magnificent DR, and as a photographer you just need to make sure you don't blow the highlights (and everything else follows). I thought ETTR was out of date ten years ago, but now, with 14 stops of DR it seems absolutely pointless - there is infinite detail in the shadows in modern sensors, and my feeling is that the only trap to fall into is to blow the highlights! Why do you need to push the exposure to the right? 

I just don't think you need a better histogram - If you just make sure nothing is drastically over-exposed everything else is fine!

All the best

A better histogram can tell us when the highlights are blown. I am OK if the implementation is conservative (warns too early). There is plenty of DR around. It is a problem if it does not show clipping when there is such. AFAIK, the averaged histogram, as implemented in M11, will not warn of clipping if only one channel clips. An RGB one would likely be better, and a raw-based one would be optimal.
Experience with the camera and its metering can often compensate for any inadequacies of metering and histogram. Ergo, reading forums does not replace in-the-field experience. So go out and photograph :), as Jono does!

 

[hdmesa]

5 minutes ago, SrMi said:

A better histogram can tell us when the highlights are blown. I am OK if the implementation is conservative (warns too early). There is plenty of DR around. It is a problem if it does not show clipping when there is such. AFAIK, the averaged histogram, as implemented in M11, will not warn of clipping if only one channel clips. An RGB one would likely be better, and a raw-based one would be optimal.
Experience with the camera and its metering can often compensate for any inadequacies of metering and histogram. Ergo, reading forums does not replace in-the-field experience. So go out and photograph :), as Jono does!

 

A live RGB histogram based on RAW data is the holy grail of histograms, but I'd just be happy with one based off the live view image like everyone else does right now.

Edited January 18, 2022 by hdmesa

[SrMi]

3 minutes ago, hdmesa said:

Agreed. I try to think more about optimal exposure when I think it will make for a significantly better image (high ISO where I know I will pull the shadows, or a landscape shot I think might be printed 40-60" wide, etc.). For all other shots, I take more of a casual approach – and the sensor in the M11 allows that approach with fewer negative consequences.

I enjoy technical part of photography as well. Therefore, if I have time, I like to thinker with the exposure.

On the other hand, in some cameras I use even highlight weighted metering. It produces less than optimal exposed images, but most of the times preserves the highlights well. Leica has it in SL2 cameras, it could be added to M11 as well.

[jonoslack]

11 minutes ago, hdmesa said:

As I mentioned, someone was talking about using exposure comp based on what they did with their M10, which I said wasn't a good assumption since this is a new sensor.

As for ETTR, every shot has an optimal exposure – that may require moving to the right of the meter, it may require moving to the left. You can still make a good image if you're one stop below optimal, but you could make a better image with an optimal exposure. Modern sensor are indeed better, but at high ISO, noise in pulled shadows is still an issue that can be reduced by getting the best exposure possible. ETTR requires zero overhead/thought process once you know how the meter in the camera works compared to how the histogram looks in Raw Digger. Why not look at a few RAWs from high contrast scenes and see if you could be letting in more light without blowing the highlights? 

I think this is where we depart company. I do acknowledge that there might be an 'optimal exposure' for any given image, but my feeling is that with modern cameras (and certainly the M11) that as long as you don't overexpose then you have about 3 stops of leeway without spoiling your perfect picture. I don't need to know if I could be letting in more light without blowing the highlights because I'm really not worried with a shot which is underexposed by a couple of stops. With 15 stops /14 stops of DR I only need to be certain that I don't blow the highlights . . . ETTR seems to me to be really unnecessary!

Best 

[SrMi]

26 minutes ago, jonoslack said:

I think I feel that all modern cameras (well, all good ones) have magnificent DR, and as a photographer you just need to make sure you don't blow the highlights (and everything else follows). I thought ETTR was out of date ten years ago, but now, with 14 stops of DR it seems absolutely pointless - there is infinite detail in the shadows in modern sensors, and my feeling is that the only trap to fall into is to blow the highlights! Why do you need to push the exposure to the right? 

I just don't think you need a better histogram - If you just make sure nothing is drastically over-exposed everything else is fine!

All the best

Most of my work on exposure is focused on not blowing the highlights. Once you lower the exposure so much that you do not blow the highlights, you have used the ETTR approach :).

[hdmesa]

Just now, SrMi said:

I enjoy technical part of photography as well. Therefore, if I have time, I like to thinker with the exposure.

On the other hand, in some cameras I use even highlight weighted metering. It produces less than optimal exposed images, but most of the times preserves the highlights well. Leica has it in SL2 cameras, it could be added to M11 as well.

I'll admit that more often than not, for landscape I will simply take the shot as the meter says (usually underexposed or close to ok) then take a few more shots at +1/3 stop each time then sort it out in post as what was best. My R5 doesn't have an option for highlight-based metering, but in all metering modes it seems to shoot like that by default – usually a bit underexposed for most shots. But the R5 also lets me put a small RGB histogram on the screen overlay, so that helps. The GFX 100S RGB histogram is huge and takes up almost the entire screen, and I can't leave it turned on – have to set a function button to turn it off and on.

[hdmesa]

12 minutes ago, jonoslack said:

I think this is where we depart company. I do acknowledge that there might be an 'optimal exposure' for any given image, but my feeling is that with modern cameras (and certainly the M11) that as long as you don't overexpose then you have about 3 stops of leeway without spoiling your perfect picture. I don't need to know if I could be letting in more light without blowing the highlights because I'm really not worried with a shot which is underexposed by a couple of stops. With 15 stops /14 stops of DR I only need to be certain that I don't blow the highlights . . . ETTR seems to me to be really unnecessary!

Best 

 

I had the exact same thought process until I was challenged by Jim Kasson on the DPR medium format forum. After accepting the challenge to review a few of my images I thought were well exposed in Raw Digger, I discovered I had about a stop of extra room to the right. And really that's the end of it, and it wasn't complicated. I simply discovered I could push my images a little further to the right than what the GFX meter was telling me without blowing the highlights. It didn't mean all the images I took before that were bad, it just meant I had a better understanding of what the meter was doing versus what the RAW file could handle moving forward.

 

Edited January 18, 2022 by hdmesa

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Rollin]

@elmars Great write up - thank you!

I have been using your Lightroom Calibration settings when working up my M10 files every since I stumbled across them in the M10 vs. M240 - Farben und so ... thread (thank goodness for Google translate), and they have helped give my images a natural feel I desire, and should have said thank you long ago. Have you developed any such LrC settings for the M11 yet?