Dynamic Range of SL2-S

[scott kirkpatrick]

Advertisement (gone after registration)

seems to be consistently one stop greater than that of the SL2.  See

https://photonstophotos.net/Charts/PDR.htm#Leica M10-R,Leica SL2,Leica SL2-S

 

 

[Advertisement]

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

[SrMi]

12 minutes ago, scott kirkpatrick said:

seems to be consistently one stop greater than that of the SL2.  See

https://photonstophotos.net/Charts/PDR.htm#Leica M10-R,Leica SL2,Leica SL2-S

 

 

Thanks fort the link! Do we know how much ISO settings differ between SL2 and SL2-S? Does ISO 1600 on SL2 corresponds to ISO 1600 on SL2-S (same exposure)?

[Chaemono]

One, to this day, I’m flabbergasted by the DR the chart shows for the SL2. In practice, it’s a lot better. Two, in practice exposure latitude matters more than DR. 

[Jeff S]

1 minute ago, Chaemono said:

Two, in practice exposure latitude matters more than DR. 

And routinely confused in forum discussion.

Jeff

[scott kirkpatrick]

Dynamic range is the number of stops between the lowest exposure that contains any information in the shadows and the highest exposure with any texture in the highlights.  Even that is a bit subjective.  And it says nothing about where the camera's suggested exposure to produce 20% gray (or 18% gray or whatever you want to see in Zone 5) falls in between these limits.  See Andy Piper's careful H&D curves* for a good example of two cameras with the same dynamic range but different amounts of headroom.

In answer to @SrMi 's question, I think the two cameras are always within 1/3 stop in their exposure recommendations.

I don't know how "exposure latitude" should be defined in the digital context.  Slope of the H&D curves, perhaps.

*I'll put the reference or link in when I find it.  It's in the M10 forum, post #93 of 

 

Edited December 29, 2020 by scott kirkpatrick

[Jeff S]

Ctein addressed the topic in a series of TOP articles...

https://theonlinephotographer.typepad.com/the_online_photographer/2012/09/dynamic-range-is-not-exposure-range-part-i.html

Jeff

 

[scott kirkpatrick]

Advertisement (gone after registration)

Ctein takes a lot of words to say that the difference between dynamic range and useful range of exposures is the noise which destroys information, mostly in the low end of the exposure curve. What we want is useful range, the number of stops between texture that you can extract from the shadows to texture that you can pull down from the highlights.  I think that is what Photons... gives, since the bottom of his curves are determined by noise measurements in dark images at base ISO.

[beewee]

2 hours ago, scott kirkpatrick said:

seems to be consistently one stop greater than that of the SL2.  See

https://photonstophotos.net/Charts/PDR.htm#Leica M10-R,Leica SL2,Leica SL2-S

 

 

This was updated in the last 24 hrs. I was looking at it last night and didn’t see the SL2-S yet.

The curve is very similar to the Nikon Z6II but we’re lacking 1/3rd stop data points. I think it is very likely that the SL2-S is using a very similar sensor with the base architecture as the Nikon Z6II, Panasonic S1 and S5, and the Sony A7 III - with some added customization to Leica’s requirements including micro lens array to work better with M lenses, Leica proprietary color filter array, and whatever else that gives the SL2-S a bit of extra ‘magic’ as compared to those other cameras that share a similar sensor.

It’s possible that although it’s a ‘Sony’ sensor, Leica could also license the sensor IP and have it custom made by another company with their own tweaks added.

Edited December 29, 2020 by beewee

[SrMi]

53 minutes ago, scott kirkpatrick said:

Ctein takes a lot of words to say that the difference between dynamic range and useful range of exposures is the noise which destroys information, mostly in the low end of the exposure curve. What we want is useful range, the number of stops between texture that you can extract from the shadows to texture that you can pull down from the highlights.  I think that is what Photons... gives, since the bottom of his curves are determined by noise measurements in dark images at base ISO.

The highest exposure is "fixed" in digital sensors; i.e., it occurs when the sensels are fully saturated. What varies with increased ISO is the amplification of the noise in shadows. Note that highlight headroom/recovery is not related to DR, but to exposure.

 

[Photoworks]

aren't most digital camera sensors basically stupid. my understanding is that the signal gets capture and transmitted to by the sensor to the processor and then all the magic is in the processor. A better sensor just capture a stronger signal that does not have to be amplified as much.

 

[beewee]

This read noise figure is interesting: https://photonstophotos.net/Charts/RN_ADU.htm#Leica SL2-S_14,Nikon Z 6II_14,Panasonic Lumix DC-S1_14,Sony ILCE-7M3_14

It seems the SL2-S is more similar to the Nikon Z6II whereas the Panasonic S1 is more similar to the Sony A7III. The Z6II is a bit like an in-between of the S1/A7III and the SL2-S.

[SrMi]

13 minutes ago, beewee said:

This read noise figure is interesting: https://photonstophotos.net/Charts/RN_ADU.htm#Leica SL2-S_14,Nikon Z 6II_14,Panasonic Lumix DC-S1_14,Sony ILCE-7M3_14

It seems the SL2-S is more similar to the Nikon Z6II whereas the Panasonic S1 is more similar to the Sony A7III. The Z6II is a bit like an in-between of the S1/A7III and the SL2-S.

Do not use that graph to compare cameras. From text below the graphs:

"These raw values are not appropriate for comparing camera models because they are not adjusted for gain or area."

[beewee]

3 minutes ago, SrMi said:

Do not use that graph to compare cameras. From text below the graphs:

"These raw values are not appropriate for comparing camera models because they are not adjusted for gain or area."

They all have the same sensor area. Gain will shift the curves left and right relative to each other. The more interesting bit: “The shape of the curve can tell you something about the amplifier circuitry of the camera.”

[Chaemono]

vor 1 Stunde schrieb scott kirkpatrick:

Ctein takes a lot of words to say that the difference between dynamic range and useful range of exposures is the noise which destroys information, mostly in the low end of the exposure curve. What we want is useful range, the number of stops between texture that you can extract from the shadows to texture that you can pull down from the highlights.  I think that is what Photons... gives, since the bottom of his curves are determined by noise measurements in dark images at base ISO.

If Photons gives that, it should have picked up that at ISO 100 the α7RIII/IV and S1R kind of suck vs. the SL2. #21 and onwards: https://www.l-camera-forum.com/topic/303396-leica-sl2-dynamic-range/page/2/#comments

 

 

[la1402]

55 minutes ago, beewee said:

This read noise figure is interesting: https://photonstophotos.net/Charts/RN_ADU.htm#Leica SL2-S_14,Nikon Z 6II_14,Panasonic Lumix DC-S1_14,Sony ILCE-7M3_14

It seems the SL2-S is more similar to the Nikon Z6II whereas the Panasonic S1 is more similar to the Sony A7III. The Z6II is a bit like an in-between of the S1/A7III and the SL2-S.

Very much so. I have done a comparsions with the Z6 and the SL2S with a test chart and as these and other tests and videos show - they are super close. Which can be expected as they probably share the same sensor.  Here just a snipping from the center at ISO 6400. SL2S left, Nikon right.

As always, the Z6 is slightly brighter and slightly warmer at the same settings.

Raw files from ISO 1600-12800 can be downloaded here:

https://1drv.ms/u/s!AsvTwjTJIhXbhIZNjwguBCZhQFh9jw?e=ArvjM7

Unforturtunately it also shows that in the corner the Sigma 24-70 2.8 is much worse than the Nikon 24-70 4.0 (both at f 5.6). I need to save a bit 💶

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 December 29, 2020 by la1402

[Chaemono]

vor 1 Minute schrieb la1402:

Very much so. I have done a comparsions with the Z6 and the SL2S with a test chart and as these and other tests and videos show - they are super close. Which can be expected as they probably share the same sensor.  Here just a snipping from ISO 6400. SL2S left, Nikon right.

As always, the Z6 is slightly brighter and slightly warmer at the same settings.

Raw files from ISO 1600-12800 can be downloaded here:

https://1drv.ms/u/s!AsvTwjTJIhXbhIZNjwguBCZhQFh9jw?e=ArvjM7

Unforturtunately it also shows that in the corner the Sigma 24-70 2.8 is much worse than the Nikon 24-70 4.0 (both at f 5.6). I need to save a bit 💶

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 Z6 shows pushed shadows banding at ISO 100. https://www.dpreview.com/articles/6705906050/nikon-z6-image-quality-and-dynamic-range-impress-but-not-without-caveats

[SrMi]

6 minutes ago, Chaemono said:

The Z6 shows pushed shadows banding at ISO 100. https://www.dpreview.com/articles/6705906050/nikon-z6-image-quality-and-dynamic-range-impress-but-not-without-caveats

That is the advantage of not having OSPDAF :), no need to worry about the stripes.

[la1402]

50 minutes ago, SrMi said:

That is the advantage of not having OSPDAF :), no need to worry about the stripes.

Yeah. Not that it was a big one in reality. 

Edited December 29, 2020 by la1402

[Advertisement]

Hallo,

Look here: Leica L-Mount

[SrMi]

1 hour ago, la1402 said:

Very much so. I have done a comparsions with the Z6 and the SL2S with a test chart and as these and other tests and videos show - they are super close. Which can be expected as they probably share the same sensor.  Here just a snipping from the center at ISO 6400. SL2S left, Nikon right.

As always, the Z6 is slightly brighter and slightly warmer at the same settings.

Raw files from ISO 1600-12800 can be downloaded here:

https://1drv.ms/u/s!AsvTwjTJIhXbhIZNjwguBCZhQFh9jw?e=ArvjM7

Unforturtunately it also shows that in the corner the Sigma 24-70 2.8 is much worse than the Nikon 24-70 4.0 (both at f 5.6). I need to save a bit 💶

<snip>

Thank you for sharing the files.
I used RawDigger to analyze the noise in the gray squares*. The brightest and darkest squares have the same noise for both cameras. However, the middle gray square has more noise in SL2-S files than in Z 6 files. This is probably caused by the strange noise patterns of SL2-S files.

*Select a square and compute average/std of green channel.

[scott kirkpatrick]

7 hours ago, Photoworks said:

aren't most digital camera sensors basically stupid. my understanding is that the signal gets capture and transmitted to by the sensor to the processor and then all the magic is in the processor. A better sensor just capture a stronger signal that does not have to be amplified as much.

No.  The signal from each photosite has to be amplified (by different amounts at different ISOs, and digitized without introducing too much noise beyone the shot noise that is inherent in capturing photons.  Then it leaves the imaging chip as a digital signal, whose processing is noiseless and can be "smart."  It takes much magic to do the analog part with as little distortion and delay as possible.