Bug in highlight metering (Firmware 1.3.0)

[mosh1]

Advertisement (gone after registration)

I think I found an issue with Highlight Metering in firmware 1.3.0. It only works correctly for me when these two conditions are true:

- Live View enabled

- Exposure Preview setting set to permanent in Capture Assistant.

When both are true you can see in the Live View the camera taking a second to adjust exposure so highlights aren't blown. When shooting with Live View disabled or Exposure Preview not set to Permanent then Highlight Metering behaves essentially the same as multi-field metering. I think this bug is why many are seeing Highlight Metering not working as expected. Even when working as expected, Highlight Metering takes a second to calculate the exposure and if you shoot too quickly it can still blow highlights.

Can someone else experiment and confirm?

 

 

[hdmesa]

I can confirm it does not work correctly when using the rangefinder. It's worse than Multifield and easily overexposes highlights. Completely screwed up implementation that should have never made it to consumers.

[hmzimelka]

Yeah, this has been discussed in a different thread. 

It's an annoying issue for me too... Also a problem for me is Multi-Segment metering has a different exposure between Live View and Rangefinder mode. 

I've reported this to info AT leica-camera DOT com. I would recommend to everyone to do the same...

[SrMi]

1 hour ago, hmzimelka said:

Yeah, this has been discussed in a different thread. 

It's an annoying issue for me too... Also a problem for me is Multi-Segment metering has a different exposure between Live View and Rangefinder mode. 

I've reported this to info AT leica-camera DOT com. I would recommend to everyone to do the same...

I never tried info AT leica-camera DOT com, but had good feedback from technicalinfo@leicacamerausa.com.

[hdmesa]

I think we're going to find that there will be compromises to metering accuracy when metering off the sensor in the lower power mode used for the rangefinder versus the normal power mode that drives Live View to the LCD and Visoflex. Leica may be doing this so that the M11 in rangefinder mode draws less power like the rangefinder metering mode on the M10 does. Perhaps Leica needs to give us the option in firmware to meter two different ways with the rangefinder – battery saving mode (but with lower accuracy) and normal mode (with accuracy identical to Live View).

[hdmesa]

Reported to both.

1 hour ago, hmzimelka said:

Yeah, this has been discussed in a different thread. 

It's an annoying issue for me too... Also a problem for me is Multi-Segment metering has a different exposure between Live View and Rangefinder mode. 

I've reported this to info AT leica-camera DOT com. I would recommend to everyone to do the same...

9 minutes ago, SrMi said:

I never tried info AT leica-camera DOT com, but had good feedback from technicalinfo@leicacamerausa.com.

 

[SrMi]

Advertisement (gone after registration)

There is a very slight difference in exposure whether shooting live view or not (multi-field, 1/80 vs. 1/90, 1/250 vs. 1/320). Do we know which one is better?

[Kwesi]

15 minutes ago, hdmesa said:

I think we're going to find that there will be compromises to metering accuracy when metering off the sensor in the lower power mode used for the rangefinder versus the normal power mode that drives Live View to the LCD and Visoflex.

I would imagine that Leica would make sure that the optimal amount of power needed for metering accuracy is available regardless of viewing mode.

[hdmesa]

Just now, Kwesi said:

I would imagine that Leica would make sure that the optimal amount of power needed for metering accuracy is available regardless of viewing mode.

You would think so, but we've seen stranger things before.

[hdmesa]

7 minutes ago, SrMi said:

There is a very slight difference in exposure whether shooting live view or not (multi-field, 1/80 vs. 1/90, 1/250 vs. 1/320). Do we know which one is better?

I think whichever most closely matches how the SL2/SL2-S would meter the same scene with the same adapted M lens. It would be nice to have similar exposure logic across platforms.

[Kwesi]

So true.

The behavior seems quite odd though given that the metering is done off the sensor in both cases.

[SrMi]

FWIW, the difference is much smaller than it is with previous digital M cameras  (center-weighted vs. multi-field). 

Still, it is curious that there is a difference at all in M11. However, it doesn't bother me much as the multi-field exposure is always only approximate (automatic). 

[hdmesa]

5 minutes ago, SrMi said:

FWIW, the difference is much smaller than it is with previous digital M cameras  (center-weighted vs. multi-field). 

Still, it is curious that there is a difference at all in M11. However, it doesn't bother me much as the multi-field exposure is always only approximate (automatic). 

Maybe one method is metering off RAW instead of JPEG — like we’ve always wanted to have. Wouldn’t that be something 😎

[SrMi]

For the record, I am one of (few?) people who are so far happy with the highlight weighted metering. On sunny days, it prevents blown highlights (compared to multi-field) and does not unnecessarily lower the exposure.

[hdmesa]

40 minutes ago, SrMi said:

For the record, I am one of (few?) people who are so far happy with the highlight weighted metering. On sunny days, it prevents blown highlights (compared to multi-field) and does not unnecessarily lower the exposure.

I see the opposite – Highlight-weighted consistency exposes brighter than Multifield, at least in situations where I need it, for example: patchy, contrasty lighting. On my SL2-S with Highlight-weighted, it can't be fooled – nothing is ever blown except specular highlights as you'd expect.

Edited April 17, 2022 by hdmesa
grammar

[hdmesa]

Here's an example of what I get with the different metering modes when using the rangefinder. Image composition was the same for each: patchy, high-contrast light. One of brightest areas in the scene was placed in the center of the frame. All of the metering modes blew the highlights. Highlight-weighted as implemented on the SL2-S would have delivered a much lower exposure than the other two modes in a situation like this.

Aperture was f/2 for all three. Shutter speed and ISO set to Auto.

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 April 17, 2022 by hdmesa

[tashley]

I have given up expecting accurate metering on all cameras, Leica included, years ago, because there’s no such thing. 

Some of us expose to the right and some don’t. If you do, you have varying tolerances  for blown highlights. What’s specular versus what’s a white patch where you might want to retain some detail? If you want detail, in how many colour channels? Does the brightest bit of the sky matter if blown and your rendering intent is high key or B&W?

Then there’s the fact that a lot of scenes where the DR of the scene > the DR of the sensor even allowing for specular highlights. 

My method is simple. Know the camera well enough to get it right despite the quirks of the metering and if in doubt, usually, set some negative comp if you have time. In practice this means I set mine to centre weighted and -1/3rd and if it’s a grab shot that’s it and I work with the result. If there’s time to tweak I take several shots and edge it to the right as I go though with the M11 not having a colour histogram that’s an issue.

What all that requires is less ACCURATE metering (because accurate for what?) and more CONSISTENT metering.

So these sorts of bugs are pretty irritating.

Odd how Leica is so high-end professional in some ways and so prone to amateurish mistakes in others.

When the product catches up with the brand positioning, it’ll be killer. Until then it’s like a gorgeous bipolar date.  

 

[Pelu2010]

Hey if you only show us the histogram without context, that can mean anything and nothing. 
 

[hdmesa]

9 minutes ago, Pelu2010 said:

Hey if you only show us the histogram without context, that can mean anything and nothing. 
 

I mentioned the content was the same for all three, that the scene was patchy, high-contrast light with a high-contrast patch right in the center. Does it matter? It's just a crap shot of a cactus that's ugly to look at.

 

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Tailwagger]

50 minutes ago, hdmesa said:

It's just a crap shot of a cactus that's ugly to look at.

Seems relevant to me. Given the variance we just need to make sure it was the camera and not the peyote. 😆