Exposing for Highlights

[jonoslack]

Advertisement (gone after registration)

4 hours ago, Chaemono said:

Good. I’ll keep benchmarking the M10 against that DxOMark score 100 sensor camera α7R III and we’ll keep this thread going until at Christmas. 😀

Fine - but make sure you don’t use 100 ISO on the M10 (of course you know that already)!

Everybody missed this one (Lloyd Chambers / Sean Reid), but the pull on the M10 has a nasty roll-off into highlights - fine by ISO 160 (native ISO)

[Chaemono]

Jono, thanks. I had read your article before I embarked on these tests. That’s why I avoided ISO 100 comparisons. I’m really impressed with ISO 200 to 800 on the M10, especially since I have also an α7R III. I’ll take the treatment of Highlights of the M10 in that ISO range any day over that of the Sony. I start to wonder whether smaller pixels is the issue. I’m so curious now that I want to try Z7 with M lenses vs. M10. See which camera treats Highlights better. It could be that the Sony sensor is just a bit dated. 

[schattenundlicht]

yawn.

[Chaemono]

Z7 vs. M10 will be interesting to some. 

Edited November 7, 2018 by Chaemono

[Mark II]

On 11/5/2018 at 9:41 PM, jonoslack said:

Fine - but make sure you don’t use 100 ISO on the M10 (of course you know that already)!

Hi Jono,

Isn't the best M10 configuration ISO 100 but with an extra -2/3 stop of exposure compensation dialled in? I have shot like this quite a lot in bright light, and it has not been a problem.

The gain in image quality (DR, noise) over ISO 200 is pretty marginal, however!

[jonoslack]

1 minute ago, Mark II said:

Hi Jono,

Isn't the best M10 configuration ISO 100 but with an extra -2/3 stop of exposure compensation dialled in? I have shot like this quite a lot in bright light, and it has not been a problem.

The gain in image quality (DR, noise) over ISO 200 is pretty marginal, however!

Hi There Mark

I'm pretty sure that in high contrast lighting you'll find that ISO 100 is always worse than ISO 200 (whatever exposure compensation you dial in). The 'base' ISO is probably around 160, but there's something nasty which happens to highlights at ISO 100

100 ISO and blown highlights

[pgh]

Advertisement (gone after registration)

8 hours ago, Chaemono said:

Z7 vs. M10 will be interesting to some. 

Given that the Sony trounces the M10 so easily to pretty much everyone else, and the Z7 is a slight improvement on that, I don't think it will be. You can keep posting tests for yourself though.

I fear that with all of your experiments you have missed the whole point that with the Leica you have to actively protect the highlights and actively underexpose, whereas with other cameras you can make proper (average) exposures for the scene and have a significantly wider range of light (about 1.5-2 stops) that is represented and/or recoverable without any noticeable loss in quality. 

[pgh]

4 hours ago, jonoslack said:

Hi There Mark

I'm pretty sure that in high contrast lighting you'll find that ISO 100 is always worse than ISO 200 (whatever exposure compensation you dial in). The 'base' ISO is probably around 160, but there's something nasty which happens to highlights at ISO 100

100 ISO and blown highlights

This is my experience - there's no getting around it. And it is why I wish we could shoot at actual base ISO of 160. With this sensor every little bit would help. 

[jonoslack]

5 minutes ago, pgh said:

Given that the Sony trounces the M10 so easily to pretty much everyone else, and the Z7 is a slight improvement on that, I don't think it will be. You can keep posting tests for yourself though.

I fear that with all of your experiments you have missed the whole point that with the Leica you have to actively protect the highlights and actively underexpose, whereas with other cameras you can make proper (average) exposures for the scene and have a significantly wider range of light (about 1.5-2 stops) that is represented and/or recoverable without any noticeable loss in quality. 

But only at 100 ISO You don't need to protect the highlights at 200 ISO (or even 800 ISO come to that). It's an unexpected spin-off from the 'Pull' (lets face it even Sean Reid didn't see it). You can make "proper (average)" exposures with the M10 as well . . as long as you don't shoot at ISO 100!

4 minutes ago, pgh said:

This is my experience - there's no getting around it. And it is why I wish we could shoot at actual base ISO of 160. With this sensor every little bit would help. 

Well, I guess I agree, But I don't find an issue shooting at 200 . . 

[Chaemono]

vor einer Stunde schrieb pgh:

Given that the Sony trounces the M10 so easily to pretty much everyone else, and the Z7 is a slight improvement on that, I don't think it will be. You can keep posting tests for yourself though.

I fear that with all of your experiments you have missed the whole point that with the Leica you have to actively protect the highlights and actively underexpose, whereas with other cameras you can make proper (average) exposures for the scene and have a significantly wider range of light (about 1.5-2 stops) that is represented and/or recoverable without any noticeable loss in quality. 

You sound like you’re curious to see more Raw file comparisons. Stay tuned. 😀

[Chaemono]

'Please, more. Post them, again. We can't get enough of Sony clipped Highlights.' Sure, your wish is my command. 

α7R III + 35 Summicron-M ISO 640 f/4.0 @1/2000 sec. ARW file: https://cc2032.zenfolio.com/img/g761517904-o750076470.dat?dl=2&tk=l3rEuOAvvRqh-F6oMO3hWGwYqf1pvpBE1wbakLiON5Y=

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 + 35 Summicron-M ISO 640 f/4.0 @1/2000 sec. DNG file: https://cc2032.zenfolio.com/img/g935169348-o750076470.dat?dl=2&tk=DU9n7eVRVMNoy_L_7vpJCnZ_XUZxCIsm3ydH9DttY4k=

[Chaemono]

And two more. Enjoy!

α7R III + 35 Summicron-M ISO 640 f/2.0 @1/2000 sec. ARW file: https://cc2032.zenfolio.com/img/g1048879071-o750076470.dat?dl=2&tk=fGWbSsPNQyQp9s1QY90PaZUcr4g2Bh_dmueiqOewKic=

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!

 

And now the same scene much better exposed with the M10. M10 + 35 Summicron-M ISO 640 f/2.0 @1/2000 sec. DNG file: https://cc2032.zenfolio.com/img/g905440933-o750076470.dat?dl=2&tk=hRFREWU9Yho2pkCal4FKgSqGCBqLFurzTVvEwY8bckw=

[mdemeyer]

15 hours ago, Chaemono said:

'Please, more. Post them, again. We can't get enough of Sony clipped Highlights.' Sure, your wish is my command. 

α7R III + 35 Summicron-M ISO 640 f/4.0 @1/2000 sec. ARW file: https://cc2032.zenfolio.com/img/g761517904-o750076470.dat?dl=2&tk=l3rEuOAvvRqh-F6oMO3hWGwYqf1pvpBE1wbakLiON5Y=

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 + 35 Summicron-M ISO 640 f/4.0 @1/2000 sec. DNG file: https://cc2032.zenfolio.com/img/g935169348-o750076470.dat?dl=2&tk=DU9n7eVRVMNoy_L_7vpJCnZ_XUZxCIsm3ydH9DttY4k=

I think, though, by showing images with the same EXPOSURE on the two cameras is showing that the SENSITIVITY of the two differs, which is a well known fact.  The M10 has a lower sensitivity to light than the Sony at a given ISO - some claim the ISO rating is exaggerated.  I’ll not get into that argument.  You can see this by looking not at the highlights but the mid-tone pipe coming out of the pump - it’s clearly darker in the M10 image.

Secondly, this whole discussion is more about metering, in my opinion.  So comparing images at the same exposure is meaningless to the point of the duscussion.  

My $0.02.

Edited November 8, 2018 by mdemeyer

[Chaemono]

Thanks for the feedback. See post #98 and the subsequent posts on page 5 and on page 6 for ISO 500 on the Sony vs. ISO 640 on the M10.

 

 

[indergaard]

1 hour ago, mdemeyer said:

I think, though, by showing images with the same EXPOSURE on the two cameras is showing that the SENSITIVITY of the two differs, which is a well known fact.  The M10 has a lower sensitivity to light than the Sony at a given ISO - some claim the ISO rating is exaggerated.  I’ll not get into that argument.  You can see this by looking not at the highlights but the mid-tone pipe coming out of the pump - it’s clearly darker in the M10 image.

Secondly, this whole discussion is more about metering, in my opinion.  So comparing images at the same exposure is meaningless to the point of the duscussion.  

My $0.02.

I've been trying to say that for a couple of pages in this thread. And it doesn't seem to sink in. The A7rIII consistently over-exposes simply because it has better sensitivity. This means that the Sony's 42mp BSI sensor is outdated compared to the M10's traditional 24mp CMOS sensor. Ironically, the BSI technology should result in better sensitivity, which again might just explain why the Sony over-exposes when compared with the exact same exposure settings on a non-BSI camera.

Quote

In a BSI CMOS sensor, the wiring is moved behind the light-receiving surface. Doing this makes the sensor more light-sensitive. The more sensitive it is, the less light is needed to get a properly exposed photo and the less noise is created.

So yes. Expose the traditional CMOS sensor of the M10 with the newer BSI sensor of the Sony at identical settings, and the Sony will always create a brighter exposure. This is a good thing if you expose correctly!

Edited November 8, 2018 by indergaard

[sebben]

ISO values are interpreted by the manufacturer of the camera. Thus what maybe ISO800 to one is ISO640 to another. The best one can really do is to compare at the base ISO of the camera where the performance of the sensor is optimal. Everything else is a waste of time really.

 

 

[pgh]

On 11/7/2018 at 1:26 PM, jonoslack said:

But only at 100 ISO You don't need to protect the highlights at 200 ISO (or even 800 ISO come to that). It's an unexpected spin-off from the 'Pull' (lets face it even Sean Reid didn't see it). You can make "proper (average)" exposures with the M10 as well . . as long as you don't shoot at ISO 100!

Well, I guess I agree, But I don't find an issue shooting at 200 . . 

I just don't have that experience.

Even at ISO 200 (which is significantly better than 100 and is fine for many circumstances but fine isn't what this is about) you still have to meter to protect highlights much more than with the Sony. I've had my M10 at ISO 200 almost exclusively for a year now, and when the highlights get too hot in an average exposure you're just not pulling them back into a decent tonal range like with the Sony. No amount of tests will change my mind because I shoot with these cameras every week, deal with this in post every week and and print almost every week. The problem has been discussed in myriad ways but the DxO score is the only numerical data and it reflects my experience. I've tried everything on the M10 to maximize the range of light it can capture, and it just cannot capture as much in a raw file as current Sony/Nikon/ etc. 

[Chaemono]

vor 1 Stunde schrieb pgh:

No amount of tests will change my mind...

Got it. But don’t you still love it when people post links to Raw files?

[M11 for me]

I get some troubles to follow here and understand the messages. With either 100 or 200 ISO I have anyway to carefully measure for the highlights. Most often with landscape this is the sky that blows out. So it is not the case that with ISO 200 we can easier avoid this. 

Since I read here that we should set ISO to 200 (and it AUTO Mode they are set to 200 anyway) I always do this of course. But still I can not make out any difference to before. I am just extremely carefull with the skys. Isay that without having „tested“ that.

On the other hand I experience every day that with very dark areas in the picture as a consequence of not clipped skys, I have seldom problems to pull up those areas in Lightroom. All in all the results are most pleasing. As an alternative camera I have the 5D IV which behaves not really in adifferent way (more my feeling not a test result). With both I very much mind the highlights very (but I must say that the Canon I only use at big events inside. For such venues the canon is a real machine in terms of AF and light measurement that ensures proper exposure times  with any focal length of a zoom lense and still keeping ISO as low as possible. All of this is not possible with the M10 in an environment where I have to react fast). My everyday camera is the M10.

[Advertisement]

Hallo,

Look here: Leica M

[jonoslack]

7 hours ago, pgh said:

I just don't have that experience.

Even at ISO 200 (which is significantly better than 100 and is fine for many circumstances but fine isn't what this is about) you still have to meter to protect highlights much more than with the Sony. I've had my M10 at ISO 200 almost exclusively for a year now, and when the highlights get too hot in an average exposure you're just not pulling them back into a decent tonal range like with the Sony. No amount of tests will change my mind because I shoot with these cameras every week, deal with this in post every week and and print almost every week. The problem has been discussed in myriad ways but the DxO score is the only numerical data and it reflects my experience. I've tried everything on the M10 to maximize the range of light it can capture, and it just cannot capture as much in a raw file as current Sony/Nikon/ etc. 

Well, I don't currently have any Sony or Nikon cameras, but what I can say is that over the last few weeks I've been going through all the images shot this year with the M10 group of cameras (about 20,000) and there isn't a single shot which has unsightly blown highlights - lots of rubbish, but nothing spoiled by the exposure. (there are some underexposed shots, but it's easy to bring up the shadows and noise isn't a problem either).

There is a fault with highlight roll-off at 100 ISO - simple as that. but at 200 and above normal care with exposure seems to me to be all that's required. 

When digital cameras had 8 or 10 stops of DR then this would have been a real issue, but with current sensors I just don't see it as an issue.

Of course - the Sony files may be better - but in this case it seems to me that the M10 is good enough - and that's all that's needed!