Leica CL …..a discovery. In praise of not using the histogram.

[davidrc]

Advertisement (gone after registration)

Not using the histogram?….shock, horror! Surely a recipe for disaster?  Read on.

We have many things drummed into us about digital photography (yes, I know, are you not joining the club here?) that we forget that there can be other ways to things.

Having used the CL now for some time, and been amazed at the 13/14 stops of DR, I can add the accuracy of the metering modes. Since most things are buried within for the CL user, I will have to assume I have left the meter mode on averaging out. 🤔  
l will take a look……I digress, why no histogram?  Well, I am currently easily proving that with an accurate meter you may be quite able to just turn the h’gram off, with the caveat of being able to bring it back on for those of a nervous disposition.

The histogram appears in ratio to size of the EVF larger than on the LCD….it takes up more space.  Highly distracting, certainly for me.
Viewing without it is like a breathe of fresh air. So, do we need it?  Well, slavishly adhering to it and potentially not setting it right can land you in more exposure trouble than what I offer as a “natural alternative”.

Pre digital we never had a histogram, we had (if we were lucky) an exposure meter needle, or a hand held one. What you must remember is that the histogram is showing you what the jpeg will look like in terms of the scene’s contrast range.  It is not a form of meter reading. Hold that thought.
Where does that leave us?  Well, the exposure scale turns out to be doing the opposite, it’s output is from reading the scene as the old analogue meter would have done. Then we apply our skill.
Interesting. Which means, we can start using our experience (those of us who can remember using skills now with cobwebs on)

Using time served methods, and the meter averaging out the scene, set the exposure using the digital scale. Compensate with +/- exposure 
In whichever way serves you, dictated by the scene in front of you. Predominately bright scene (snow) or darks (your moggie in the cellar) then most of us know that a 1 stop compensation or a bit more (or less) will be generally needed, as all metering is for 18% grey result.

So, with practice and some simple scene tests, you will find your own individual meter and how it works and “sees” scenes. Any of the PSAM modes you may use, manual or otherwise, still leaves you at the mercy of auto, or correcting to shift the histogram anyway……AND YOU ARE DISTRACTED FROM TAKING THE SHOT.

Learn to “read” the scene yourself, if you don’t already do that.  
I am working with the P mode at the moment, something I have never really done before i’ll confess, suffering all the above with different cameras.  
Set the digital scale to 0, and if the scene has well balanced and a spread of H, M, S, then the meter has done it’s job correctly.  Trust the average scene and 0.
When in the photo below, the predominance of highlights are fooling the meter (the photo has been cropped) then apply + 2/3.  (As 1/3 wasn’t quite enough) and sure enough in LR it was a 30 sec edit to taste.  And perfect AWB by the way from the CL needing no adjustment.

Using a histogram correctly is not easy, it requires more complex theory (some of it misleading) to gauge the scene and judge if the camera has it right in the shown jpeg on screen, because that is what you see, a jpeg guess.

And enjoy a faster way of working, a clearer view and less hassle….and enjoy the process more.

Read the scene, and correct if needed, trust the digital scale and 0 on auto, leave the histogram in the dark. See what happens. 
Watch out for a scene which does not conform to average, and add a simple correction.  You will be surprised.  It’s how we all used to work.

 

 

 

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 February 6, 2024 by davidrc
Loads of typos!

[Advertisement]

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

[lct]

Never used an histogram in 20+ years of digital photography. WYSIWYG capabilities of modern mirrorless cameras, including CL and M11 with Visoflex 2, are enough for me. Now other people do like histograms actually so i would not expect Leica or any other camera maker to sell cameras without it.

[davidrc]

3 minutes ago, lct said:

Never used an histogram in 20+ years of digital photography. WYSIWYG capabilities of modern mirrorless cameras, including CL and M11 with Visoflex 2, are enough for me. Now other people do like histograms actually so i would not expect Leica or any other camera maker to sell cameras without it.

There you go, I knew there must be someone else! 😀.   It makes sense, does it not…..nice innit!
I wouldn’t necessarily advocate only going by a screen though, too many variables, auto screen brightness if turned on for one.

I know Lovegrove the Fuji ambassador, used to use a version of screen set up to simulate a flat raw file, but that’s a different discipline and not totally about exposure.
And film simulations…..there’s a gimmick and thorny subject in itself.  

Without a doubt, the CL is the first digital camera I have used where the exposure scale can be relied upon to be used as I use the Leica R6.  The metering principles in use are the same, and so is the accuracy.

[SrMi]

I do not use histograms except in image reviews. In live view, I use highlight clipping. The reason is that with very little effort, I can extract more out of the camera's potential (less noise/better tonality and less highlight loss) than when guestimating the exposure.

Digital photography has tools that are not available with film. Why not use it?

 

[davidrc]

2 minutes ago, SrMi said:

I do not use histograms except in image reviews. In live view, I use highlight clipping. The reason is that with very little effort, I can extract more out of the camera's potential (less noise/better tonality and less highlight loss) than when guestimating the exposure.

Digital photography has tools that are not available with film. Why not use it?

 

What if there are no extremes to clip?…..you still need to set the exposure somewhere.  In effect that is a sideways histogram.

If the meter is accurate, let it do it’s work.

Any ‘tool’ is a good one, if used right…..but this is only about what works more easily, and faster.  With the caveat of ‘each to his or her own’

[SrMi]

1 hour ago, davidrc said:

What if there are no extremes to clip?…..you still need to set the exposure somewhere.  In effect that is a sideways histogram.

If the meter is accurate, let it do it’s work.

Any ‘tool’ is a good one, if used right…..but this is only about what works more easily, and faster.  With the caveat of ‘each to his or her own’

Only highlights can clip in digital photography.

If there is no highlights to clip with camera’s automatic metering, and shooting raw, I add positive EC until the clipping starts (approximate ETTR). If I use manual metering, I adjust it so that it barely does not clip.

There are more “precise” ways to do that (e.g., UniWB), but they are too bothersome for me.

What does “meter is accurate” means? There are many ways to meter a scene, and they often deliver different results. 

I agree, that people should use what works for them. Highlight clipping warnings have become the most important metering tool for me.

There is also difference whether you are shooting JPEGs or raw. With JPEGs the initial brightness of the image matters.

[davidrc]

Advertisement (gone after registration)

6 hours ago, SrMi said:

Only highlights can clip in digital photography.

If there is no highlights to clip with camera’s automatic metering, and shooting raw, I add positive EC until the clipping starts (approximate ETTR). If I use manual metering, I adjust it so that it barely does not clip.

There are more “precise” ways to do that (e.g., UniWB), but they are too bothersome for me.

What does “meter is accurate” means? There are many ways to meter a scene, and they often deliver different results. 

I agree, that people should use what works for them. Highlight clipping warnings have become the most important metering tool for me.

There is also difference whether you are shooting JPEGs or raw. With JPEGs the initial brightness of the image matters.

Interesting thoughts there.

Yes, highlights clip but shadows can block or lose depth and show noise. I have demonstrated elsewhere the range that the CL has, one reason your methods may fair better than Fuji.
I wonder what your lower end of the DR is when you always push to clip?  In my experience that panacea is not a universal cure (for want of a better word) for every subject. If there are few areas where clipping is a danger then you risk increasing exposure in lower and mids to a point of losing contrast.

There is often a balance which should be struck, let a meter find that, then use experience to tweak either way depending on the contrast range of the scene and where you need the exposure to be. It applies to raw as much as jpeg.

Though I suspect you are a predominately jpeg shooter. And doing what you do i’d say is better than trying to juggle a histogram.
 

[lct]

Histograms for jpeg shooters? I did not think of that. Many (most?) raw shooters swear by what they call ETTR (expose to the right) and a histogram, i suppose, is necessary for that.

[davidrc]

16 minutes ago, lct said:

Histograms for jpeg shooters? I did not think of that. Many (most?) raw shooters swear by what they call ETTR (expose to the right) and a histogram, i suppose, is necessary for that.

It all rather depends on the type of metering, and the scene….this is why the ETTR fails for some and is a perennial hot debate.

When I do, I don’t….push it right to the point, there is no need. If it is used to gain near perfection it is too slow, and with digital in camera perfect exposure is a myth, well, universal 100%  perfect at any rate.

What is not often grasped, and fact, is that the histogram is merely a shown graph of what the jpeg being produced is.

If you set the scale to 0 and without moving the camera or light changing, change comp or aperture (if in M) you will see that the histogram really doesn’t change that much compared to what 2/3rd does (example).  The raw will change much more as a result.

Histograms can mislead trying to get a preferred outcome and raw.

[lct]

You should say this to good techies here. I'm not one (un)fortunately 😉 so i keep doing what i did in the film days with transparencies i.e. the opposite to what techies suggest i'm afraid. Shame on me 😄

Edited February 10, 2024 by lct

[jaapv]

3 hours ago, davidrc said:

It all rather depends on the type of metering, and the scene….this is why the ETTR fails for some and is a perennial hot debate.

When I do, I don’t….push it right to the point, there is no need. If it is used to gain near perfection it is too slow, and with digital in camera perfect exposure is a myth, well, universal 100%  perfect at any rate.

What is not often grasped, and fact, is that the histogram is merely a shown graph of what the jpeg being produced is.

If you set the scale to 0 and without moving the camera or light changing, change comp or aperture (if in M) you will see that the histogram really doesn’t change that much compared to what 2/3rd does (example).  The raw will change much more as a result.

Histograms can mislead trying to get a preferred outcome and raw.

Depends on the camera. The histogram on the M9M is raw and calibrated to the zone system. I agree that one must learn to understand and use a histogram. But that goes for conventional metering as well. In my experience the histogram is still the best tool to use the full DR of the camera. 

[SrMi]

7 hours ago, davidrc said:

Yes, highlights clip but shadows can block or lose depth and show noise. I have demonstrated elsewhere the range that the CL has, one reason your methods may fair better than Fuji.

Yes, shadows get overwhelmed by noise, but they do not clip like highlights. Clipping causes a sudden loss of data, i.e., structure and color. I do not understand why "my "methods may fare better than Fuji."

7 hours ago, davidrc said:

I wonder what your lower end of the DR is when you always push to clip?

I do not understand that. 

7 hours ago, davidrc said:

If there are few areas where clipping is a danger then you risk increasing exposure in lower and mids to a point of losing contrast.

I do not understand what you mean by "risk increasing exposure in lower and mids to the point of losing contrast." The contrast does not change in "my approach," and the brightness in the lower and mids can be restored in the post as desired.

We can assume that all sensors work only in their linear range (required non-linearity is added by the post-processor's profiles).

7 hours ago, davidrc said:

There is often a balance which should be struck, let a meter find that, then use experience to tweak either way depending on the contrast range of the scene and where you need the exposure to be. It applies to raw as much as jpeg.

The meter does not understand the scene; you need experience adjusting the initial metering result. That certainly works, but you are probably wasting your camera's potential. 

7 hours ago, davidrc said:

Though I suspect you are a predominately jpeg shooter. And doing what you do i’d say is better than trying to juggle a histogram.

Except with smartphones, I only shoot raw. "My approach" is less suitable for JPEGs.

[SrMi]

6 hours ago, davidrc said:

It all rather depends on the type of metering, and the scene….this is why the ETTR fails for some and is a perennial hot debate.

When I do, I don’t….push it right to the point, there is no need. If it is used to gain near perfection it is too slow, and with digital in camera perfect exposure is a myth, well, universal 100%  perfect at any rate.

What is not often grasped, and fact, is that the histogram is merely a shown graph of what the jpeg being produced is.

The main reason that ETTR fails for some seems to be that they clip the highlights, which is the ultimate ETTR sin.

I believe that most people who use histograms and highlight clipping know that it is, unfortunately, based on JPEGs in most cameras.

6 hours ago, davidrc said:

If you set the scale to 0 and without moving the camera or light changing, change comp or aperture (if in M) you will see that the histogram really doesn’t change that much compared to what 2/3rd does (example).  The raw will change much more as a result.

I do not understand what you wrote.

[davidrc]

4 hours ago, jaapv said:

Depends on the camera. The histogram on the M9M is raw and calibrated to the zone system. I agree that one must learn to understand and use a histogram. But that goes for conventional metering as well. In my experience the histogram is still the best tool to use the full DR of the camera. 

But we are talking about how good the meter is on the CL. For the benefit of the CL users. 
And if the M9M actually shows you the raw file…..it’d be the only camera that does to my knowledge.
You only have to understand a histogram if you want to use it. 😉
”best tool to use the full DR of the camera”.  Since you can’t render in camera a full DR (as a jpeg doesn’t have it) then the shown histogram (from the jpeg) is not an absolute of the raw file.
So the histogram is only of use to show you are not approximately clipping the highlights of the jpeg….if jpegs are the only thing you want to use.  

It is well documented that a histogram is not a rendition, interpretation, of the raw file. It is a rendering of the jpeg which is being produced in camera. And a histogram shows about a DR of about 9, (from the jpeg) and the CL raw has about 13.
Quoting:
“The other issue that makes it incorrect to refer to a histogram as "a graphical representation of the pixels exposed in your image" is that a histogram changes significantly with changes in the camera settings such as contrast, picture style, brightness, white balance, etc. Here is why. Those camera settings change the in-camera JPEG rendering, that includes the JPEG embedded as a preview into the RAW file. The histogram is derived from that embedded preview, and as that preview changes, so does the histogram”.

You can’t see in camera what the raw will look like.  A histogram with either end not overshot keeps you safe, but for putting the exposure range where you need it for your subject, it’s not the be all and end all.

The CL, via the exposure scale has a great meter, which you can use without a histogram more quickly and effectively.

[davidrc]

2 hours ago, SrMi said:

The main reason that ETTR fails for some seems to be that they clip the highlights, which is the ultimate ETTR sin.

I believe that most people who use histograms and highlight clipping know that it is, unfortunately, based on JPEGs in most cameras.

I do not understand what you wrote.

I was explaining that that the histogram is not really showing you a raw file, a way to show that.

[jaapv]

1 hour ago, davidrc said:

But we are talking about how good the meter is on the CL. For the benefit of the CL users. 
And if the M9M actually shows you the raw file…..it’d be the only camera that does to my knowledge.
You only have to understand a histogram if you want to use it. 😉
”best tool to use the full DR of the camera”.  Since you can’t render in camera a full DR (as a jpeg doesn’t have it) then the shown histogram (from the jpeg) is not an absolute of the raw file.
So the histogram is only of use to show you are not approximately clipping the highlights of the jpeg….if jpegs are the only thing you want to use.  

It is well documented that a histogram is not a rendition, interpretation, of the raw file. It is a rendering of the jpeg which is being produced in camera. And a histogram shows about a DR of about 9, (from the jpeg) and the CL raw has about 13.
Quoting:
“The other issue that makes it incorrect to refer to a histogram as "a graphical representation of the pixels exposed in your image" is that a histogram changes significantly with changes in the camera settings such as contrast, picture style, brightness, white balance, etc. Here is why. Those camera settings change the in-camera JPEG rendering, that includes the JPEG embedded as a preview into the RAW file. The histogram is derived from that embedded preview, and as that preview changes, so does the histogram”.

You can’t see in camera what the raw will look like.  A histogram with either end not overshot keeps you safe, but for putting the exposure range where you need it for your subject, it’s not the be all and end all.

The CL, via the exposure scale has a great meter, which you can use without a histogram more quickly and effectively.

But the exposure meter does not show you the clipping highlights, the "blinkies" do that and those are equally based on the jpg. Thus the histogram will allow you to interpret the clipping better as you see it moving.  Of course you must be able to interpret the histogram as it is only a tool with its limitations, but so is the exposure meter. The most precise way is to measure the scene with a handheld spot meter and throw in an incident light metering too - but who takes the time to do that? 

[SrMi]

58 minutes ago, davidrc said:

I was explaining that that the histogram is not really showing you a raw file, a way to show that.

Yes, that is true. The consequence is that, in most situations, you are wasting a bit of IQ (noise).

However, in many cases, histogram and/or highlight clipping will allow a "better" exposure (less noise) than when not using them.

 

[davidrc]

2 hours ago, jaapv said:

But the exposure meter does not show you the clipping highlights, the "blinkies" do that and those are equally based on the jpg. Thus the histogram will allow you to interpret the clipping better as you see it moving.  Of course you must be able to interpret the histogram as it is only a tool with its limitations, but so is the exposure meter. The most precise way is to measure the scene with a handheld spot meter and throw in an incident light metering too - but who takes the time to do that? 

We agree on one thing then…..the histogram has limitations. If you have the time to faff around with that trying to interpret it, you have the time to use a HH meter. 
We get to the bottom line….the histogram is only really any good for trying to preserve clipped highlights, and if they are well  outside the DR, you’ll just under expose the rest.

 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[jaapv]

Speak for yourself. I can faff well within a second. 

[SrMi]

9 minutes ago, davidrc said:

We agree on one thing then…..the histogram has limitations. If you have the time to faff around with that trying to interpret it, you have the time to use a HH meter. 
We get to the bottom line….the histogram is only really any good for trying to preserve clipped highlights, and if they are well  outside the DR, you’ll just under expose the rest.

 

Histogram and highlight clipping warnings are useful to save highlights from clipping or reduce noise in the image. Using histograms or highlight clipping warnings is as simple and fast as adjusting exposure compensation while framing the image. In my book, that is much faster than using a handheld meter to determine exposure.