Exposing for Highlights

[Michael Naylor]

Advertisement (gone after registration)

27 minutes ago, designdog said:

Part of the reason for my move was my stagnation with the Nikon - or any other DSLR. Another part was the enjoyment I had during a brief foray back into film two years ago. Got me slowing down and appreciating taking pictures again.

I'll be 73 this December and feel exactly the same.  My best pictures were about 30 years ago when I was using a Bronica SQA.  Slowing down certainly works.

[pico]

19 hours ago, Chaemono said:

I love small computers. The only issue with them is they seem to struggle with extreme Highlights and Shadows between ISO 200 and ISO 800 in the same scene. They either overexpose the Highlights and underexpose the Shadows or underexpose the Shadows and overexpose the Highlights. Stay tuned.

ISO 200 to 800 is hardly a reach. As to the rest - Stay tuned for what? Your re-interpretation of physics?

 

Edited October 24, 2018 by pico

[dkmoore]

Chaemono could very well be the best photographer on this sight but we would never know because all we get is pictures of sticks and walls.

I'm being a hypocrite here (self admitted) as I chastise others for being ornery but I really don't see the point in constantly comparing Sony to Leica.

Stay tuned for monotony...

[david strachan]

Exposure is always a user input...making sure you cover the subject brightness range...as you want. Strictly as YOU want.

 

Blown highlights is not always a disaster...do we want everything to look HDR.

 

One still has to think, as with anything from spade to knife.

When i use my compass i have to think Grid North, Magnetic North or Geographic North.

...

...

[Chaemono]

vor 2 Stunden schrieb pico:

ISO 200 to 800 is hardly a reach. As to the rest - Stay tuned for what? Your re-interpretation of physics?

 

Blown Highlights and crushed Shadows of a DxOMark score 100 sensor, just because you keep insisting. 😀 Stay tuned. 

Edited October 24, 2018 by Chaemono

[jaapv]

On 10/16/2018 at 10:44 AM, evikne said:

I would go for full manual mode with the M10 too. I always do that myself. To me, that is the fastest, most natural, most direct and reliable way to use a Leica M. I may take a look at the exposure metering (red dot and arrows) in the viewfinder before the first shot, but only as a clue for my manual settings.

To preserve the highlights, I simply point the viewfinder against the light source and adjust the ISO/shutter speed/aperture so the right red arrow disappears. If I have time, I may take a test shot and eventually make some adjustments before I go on. As long as the light is the same, I don’t need to bother anymore with the exposure and can rather concentrate on more important things, like capturing the right moments.

And if  something goes wrong, I can only blame myself, not the camera.

Exactly. The M cameras are not made for those who expect automation to do their thinking for them.

[Guest Nowhereman]

Advertisement (gone after registration)

@jaapv - While what you say is true, it doesn't mean that there's no issue with the highlight rendering of the M10. It was evident in the first shots that took when I got the camera from the dealer in Paris. It was a sunny day, and I didn't need to be in the harsh tropical light of Bangkok to see, as I've said earlier, that the M10 blew highlights more easily and also rendered highlights substantially lighter than the M9 and MM — I've never compared it to the Sony A7III, but others, whose judgment I trust more than that of Chaemono, for reasons that can be found in this and another thread, have shown that the M10 is problematic in this respect: essentially, whenever you're not shooting with the light source behind you,  which is most of the time as far as I am concerned. At this stage, I would not recommend the M10 for anyone interested in primarily shooting into the light, or with a strong sidelight. I suspect that the problem could be solved by a firmware fix from Leica.
_______________
Alone in Bangkok essay on BURN Magazine
Nowhereman Instagram

Edited October 24, 2018 by Nowhereman

[pgk]

5 hours ago, dkmoore said:

..... I really don't see the point in constantly comparing Sony to Leica.

Stay tuned for monotony...

Indeed. Until someone comes up with a truly quantitative and objective technical system for genuinely comparing differences which make very little difference to actual photographs then we will all have to suffer the continual condensing of differences into some silly simplistic scoring system. After all, why let complex differences get in the way of describing them in idiot terms?

As for blown highlights, well I get the distinct impression that few here are genuinely interested in learning their metering system and post processing workflow. Genuinely blown highlights would suggest a lower dynamic range (unless someone can tell me differently with a very good technical explanation) and I'm extremely dubious that the dynamic range of the M10 is substantially lower than many other cameras, especially the M9. I shoot into the light (contre jours) with ALL my digital cameras and have had to learn their characteristics in terms of highlight handling and equally importantly lens flare characteristics too. We have never had cameras as good as we have now and we need to enjoy using them as much as we need to whine about their perceived failings.

[Mark II]

The M10's dynamic range is not substantially lower than other cameras (although it is lower than the class leader).

Much of the M10's reputation I think comes from its bogus ISO100 implementation, which is implemented as a software pull from an exposure equivalent to a hardware ISO of ~160 or thereabouts. If you use ISO 100 you need to dial-in -2/3 exposure compensation to get similar highlight headroom to ISO200 or to that from an M240.

Shooting in London this weekend, the strongly contrast low winter sunllight was a *major* problem with the M10. For black and white it was fine, but for colour I was struggling to retain shadow information while avoiding colour-casts in the transitions around over-exposed areas of the sky. All digital cameras struggle in the environment I was shooting in, but the M10 is clearly worse than the best Sony or Nikon systems.

[pgk]

9 minutes ago, Mark II said:

The M10's dynamic range is not substantially lower than other cameras (although it is lower than the class leader).

All digital cameras struggle in the environment I was shooting in, but the M10 is clearly worse than the best Sony or Nikon systems.

And?

[Chaemono]

Raw files are fun. I know that in reality even those here saying DR of the M10 is clearly worse actually love to look at them.

[Mark II]

56 minutes ago, pgk said:

And?

And so I used live-view periodically to check for ETTL exposures - not ideal given that this was sometimes rapid shooting in changing lighting (the People's Vote March in London, shooting with a 28mm lens). I do not have a problem with burning highlights or crushing shadows in high contrast lighting: the difficulty is when I am not 100% confident of what the metering is telling me relative to what I am trying to achieve.

Perhaps tellingly, it is *much* easier to get good exposures with the EVF (with a wide-angle lens and messy, rapidly changing high contrast lighting) than it is when using the rangefinder. But that kind of bypasses the entire point of using an M. Also tellingly, shooting film in the same conditions would have been less problematic, not because it has greater DR but because it handles overexposure more gracefully.

[Michael Naylor]

5 minutes ago, Mark II said:

Also tellingly, shooting film in the same conditions would have been less problematic, not because it has greater DR but because it handles overexposure more gracefully.

In a previous life I was an audio engineer.  Undesirable dynamic peaks (audio overshoots) can be controlled by the use of what is called "soft clipping".  If only something like this could be implemented within camera firmware.  It's not rocket science - it's just having the "highlight" feature of PS/LR/CO built into the camera.  What's more, the severity of the highlight reduction can easily be automated (dynamic) according to picture content.  It was my experience when working in broadcast television, video engineers and audio engineers were from different planet.  Such a shame, because they should have, and still could, learn from each other.

[pgk]

1 hour ago, Mark II said:

Also tellingly, shooting film in the same conditions would have been less problematic, not because it has greater DR but because it handles overexposure more gracefully.

Actually film can produce spurious highlight information due to High Intensity Reciprocity Law Failure. The best way to illustrate this is that sometimes the Sun is shown on transparency film as being darker than the surrounding highlights. The problem is that we got used to the way that film rendered extreme highlights, and the colour shifts and banding which can occur in extreme situations on digital are less acceptable visually.

When the contrast of a scene exceeds the sensors ability to record both highlight and shadow detail, exposure is always going to be a compromise. Simple meter readings don't really help because they will be fooled by the high contrast into giving dodgy information. You have to experiment and learn what works best for the result you want. These days I generally ignore the meter and set what I think will work best for shooting into the light based on my experience, take a shot and adjust as needed. I don't worry about trying to achieve a 'perfect' exposure because more often than not there simply isn't one. Being pedantic it might be possible to use grads and so on but that's not how I use a rangefinder.

Edited October 24, 2018 by pgk

[Simon]

On 10/16/2018 at 5:27 AM, SMAL said:

I am pretty happy with how it exposes. I never had an issue that it under exposed to a point that it wasn’t recoverable. 

The only time when I need a negative exposure compensation is when shooting right into the sun.

Yesterday I’ve shot some photo at ISO6400 and it was basically pitch black at the beach and the M10 nailed it.

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!

Really gorgeous shot!  😍

[oldwino]

3 hours ago, mkenaylor said:

In a previous life I was an audio engineer.  Undesirable dynamic peaks (audio overshoots) can be controlled by the use of what is called "soft clipping".  If only something like this could be implemented within camera firmware.  It's not rocket science - it's just having the "highlight" feature of PS/LR/CO built into the camera.  What's more, the severity of the highlight reduction can easily be automated (dynamic) according to picture content.  It was my experience when working in broadcast television, video engineers and audio engineers were from different planet.  Such a shame, because they should have, and still could, learn from each other.

Isn’t this sort of what the Fuji X cameras do with their “Auto DR”?

[thrid]

7 hours ago, mkenaylor said:

In a previous life I was an audio engineer.  Undesirable dynamic peaks (audio overshoots) can be controlled by the use of what is called "soft clipping".  If only something like this could be implemented within camera firmware.  It's not rocket science - it's just having the "highlight" feature of PS/LR/CO built into the camera.  What's more, the severity of the highlight reduction can easily be automated (dynamic) according to picture content.  It was my experience when working in broadcast television, video engineers and audio engineers were from different planet.  Such a shame, because they should have, and still could, learn from each other.

Arri the movie camera manufacturer down in Munich, solved this problem 10 years ago in their Alexa camera with the ALEV III sensor.

The ALEV III has a dual grain readout at each receptor that takes a simultaneous bracketed exposure and combines the two readings on the fly. The result is a butter smooth film like roll off in the highlights and massive dynamic range (15 useable stops). When the Alexa clips or blows out it does it gracefully like film. That's the kind of sensor Leica should put in the M11. 

http://www.arri.com/camera/alexa/technology/arri_imaging_technology/alexas_sensor/

Basic explanation of the Alexa ALEV III sensor

The Science behind the sensor

Although the science behind the breakthrough performance of ALEXA's custom designed CMOS sensor is complex, the use of large photosites and a Dual Gain Architecture are its two main principles. 

By employing unusually large photosites (in today's world of tiny cell phone sensors), ALEXA's sensor exhibits high dynamic range, high sensitivity and low crosstalk. The larger a photosite is, the more light it can capture and the lower the noise. 

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!

The Dual Gain Architecture simultaneously provides two separate read-out paths from each pixel with different amplification. The first path contains the regular, highly amplified signal. The second path contains a signal with lower amplification, to capture the information that is clipped in the first path. Both paths feed into the camera's A/D converters, delivering a 14 bit image for each path. These images are then combined into a single 16 bit high dynamic range image. This method enhances low light performance and prevents the highlights from being clipped, thereby significantly extending the dynamic range of the image. 

 

Here is a +/- 5 stop exposure test on the Alexa with the . Notice how the Alexa never blows the highlights. Most of the other cameras go to pieces when you overexposure. 

 

 

 

Edited October 24, 2018 by thrid

[Chaemono]

Okay ladies and gentlemen,  the topic today is...'Exposing for Highlights.' How did you guess? Let's post more Raw files to explore it further (I know some here love to look at them even if they pretend to hate it).

We saw on page 2 of this thread here:

that using the same ISO on the α7R III, our control camera (definition of control group: the standard to which comparisons are made in an experiment), as on the M10 leads to blown Highlights in the Sony pictures. Therefore, the images below will be with ISO 500 on the α7R III and ISO 640 on the M10. It's actually the M10-P now and LR does not support it with the Adobe profiles, yet. So, I had to make my own based on the stock profile but that's a different matter.

We are exposing for the candle here and the links to the Raw files are in the following post.

α7R III + 35 Summicron-M ISO 500 f/2.0 @1/30 sec. with LR adjustments of Vignetting +30, Exposure +1.90, Shadows +100.

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!

 

M10 + Summicron-M ISO 640 f/2.0 @1/30 sec. with LR adjustments of Vignetting +30, Exposure +1.90, Shadows +100.

Edited October 28, 2018 by Chaemono

[Chaemono]

And this is what they looked without Exposure adjusted and Shadows lifted (using the Adobe standard profile on the Sony and the stock profile on the M10). Check out the noise in the files after they are brightened up. The M10-P image looks suspiciously clean but it's probably just the larger MPx. It has the same M10 sensor.

α7R III + 35 Summicron-M ISO 500 f/2.0 @1/30 sec. ARW: https://cc2032.zenfolio.com/img/g913874405-o750076470.dat?dl=2&tk=-zHG1XCsWqSaO0aC-YKJgVncSCA_4kmjD9wsXRGegNw=

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!

 

M10 + Summicron-M ISO 640 f/2.0 @1/30 sec. DNG: https://cc2032.zenfolio.com/img/g1046047836-o750076470.dat?dl=2&tk=vciKxJzXHXdc_PWujXKnpr4D8-xiejJlQT6NGnoIQag=

[Advertisement]

Hallo,

Look here: Leica M

[sebben]

I can post a comparison between the M10 sensor and the ALEV III if anyone is interested to see just how weak the M10 is in the highlights...