M11 vs. SL2-S at ISO 12500

[Chaemono]

Advertisement (gone after registration)

vor 1 Stunde schrieb SrMi:

Can you say that for all 24MP sensors (e.g., M10-P, Z 6) or is that for SL2-S (and S5) sensors?

First, it seems to apply in general to 24 MP BSI sensors but I know that it’s the case for the SL2-S and the S5. In the pictures in my link above, the S1 and the α7 III also don’t show white pixels in the high ISO dark areas.

[Advertisement]

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

[bradhusick]

I'd love to see a comparison of the M10-P with the M11 in 24MP mode at high ISO levels.

[SrMi]

56 minutes ago, bradhusick said:

I'd love to see a comparison of the M10-P with the M11 in 24MP mode at high ISO levels.

There is no 24MP mode in M11. It would be nice, but Leica said "math" prevents it.

[bradhusick]

I guess I meant 18 megapixels.

[bradhusick]

I have now read Sean Reid's review and the results from lower level raw modes on the M11 are not encouraging to me from the standpoint of image resolution and clarity, nor dynamic range. It appears they are simply available to save storage space at the cost of IQ.

[SrMi]

40 minutes ago, bradhusick said:

I have now read Sean Reid's review and the results from lower level raw modes on the M11 are not encouraging to me from the standpoint of image resolution and clarity, nor dynamic range. It appears they are simply available to save storage space at the cost of IQ.

That is not how I read it. 
E.g., in Concluding Thoughts, Sean writes that the M-DNG is cleaner than either M10-R's DNG or downsized L-DNG.
IMO: based on his analysis, the differences between different DNG modes are too small to matter. I do not have practical experience, as I always shoot in L-DNG. 

[FlashGordonPhotography]

Advertisement (gone after registration)

4 hours ago, bradhusick said:

I have now read Sean Reid's review and the results from lower level raw modes on the M11 are not encouraging to me from the standpoint of image resolution and clarity, nor dynamic range. It appears they are simply available to save storage space at the cost of IQ.

At a pixel level the M11 files are cleaner at the lower resolutions. Leica claim an increase in DR of a stop, which seems about right from my testing. But it also depends on how you test and what your criteria are. But you can, depending on the algorithm you use, acheive pretty much the same results in post by down sampling the image. So it depends on whether you prefer to do this stuff in camera or in post.

So if you wanted the M11 to shoot at lower resolutions there are no disadvantages to setting either 36 or 18MP in camera and just treating the M11 as such. You'll end up with similar results as if Leica had stuch a lower resolution sensor in there. The SL2-S is probably the best sensor around at a pixel level currently for noise and DR but the real world differences to the M11 are not huge and unlikely to be noticable in the real world. A half stop exposure error would skew the results enough to change it.

Gordon

[Guest]

On 2/19/2022 at 2:03 PM, FlashGordonPhotography said:

At a pixel level the M11 files are cleaner at the lower resolutions. Leica claim an increase in DR of a stop, which seems about right from my testing. But it also depends on how you test and what your criteria are. But you can, depending on the algorithm you use, acheive pretty much the same results in post by down sampling the image. So it depends on whether you prefer to do this stuff in camera or in post.

So if you wanted the M11 to shoot at lower resolutions there are no disadvantages to setting either 36 or 18MP in camera and just treating the M11 as such. You'll end up with similar results as if Leica had stuch a lower resolution sensor in there. The SL2-S is probably the best sensor around at a pixel level currently for noise and DR but the real world differences to the M11 are not huge and unlikely to be noticable in the real world. A half stop exposure error would skew the results enough to change it.

Gordon

Looks like M11 (blue) is doing pretty well regarding Dynamic Range, compared to SL2 and SL2-s (source😞

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!

Interestingly M11 read noise (blue) is best of the 3 up to ISO 400 and between SL2 and SL2-S for higher ISOs:

Overall this seems to indicate better M11 noise (~0.7 stops?) and DR performance (1 stops?) than SL2, despite higher resolution of the M11.

[Adam Bonn]

On 1/20/2022 at 2:45 PM, SrMi said:

.
With Cobalt, you must first buy the Adobe and DNG Basic pack. It is unique and different for each camera model. The film simulation packs can be shared between cameras.

Off topic here, but in case it’s of interest to anyone…. This is also exactly how the relationship between ‘adobe standard’ and ‘adobe color’ works in ACR/LR 

adobe standard is the unique profile that they made for your camera and adobe color (and adobe landscape, neutral, vivid etc) is a generic profile shared between cameras

That’s not to criticise it or say it’s bad, just that it’s not at all tailored for any particular camera

[SrMi]

5 hours ago, Adam Bonn said:

Off topic here, but in case it’s of interest to anyone…. This is also exactly how the relationship between ‘adobe standard’ and ‘adobe color’ works in ACR/LR 

adobe standard is the unique profile that they made for your camera and adobe color (and adobe landscape, neutral, vivid etc) is a generic profile shared between cameras

That’s not to criticise it or say it’s bad, just that it’s not at all tailored for any particular camera

I am surprised to hear that since Adobe switched from Adobe Standard to Adobe Color as default profile.

[Adam Bonn]

38 minutes ago, SrMi said:

I am surprised to hear that since Adobe switched from Adobe Standard to Adobe Color as default profile.

Yup, if you go and find them on  your system, open one up (xmp file so any text editor will do) you'll see that it contains zero camera specific data, works off the back of adobe standard (for comparison the Cobalt ones piggy back off of modular or standard) but most telling of all.. You'll see there's only one file not one each for the thousands and thousands of cameras that adobe support like you see with the dcp files folder. 

The dcp files are all named camera brand, camera model, adobe standard(. dcp), but their display name is always adobe standard. So adobe have quite cleverly implemented a 3D LUT dcp upgrade that works with all their incumbent dcp files 

They must be pleased with it because as you note adobe color is the new adobe standard with default settings 

[FlashGordonPhotography]

15 hours ago, mzbe said:

Looks like M11 (blue) is doing pretty well regarding Dynamic Range, compared to SL2 and SL2-s (source😞

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!

Interestingly M11 read noise (blue) is best of the 3 up to ISO 400 and between SL2 and SL2-S for higher ISOs:

Overall this seems to indicate better M11 noise (~0.7 stops?) and DR performance (1 stops?) than SL2, despite higher resolution of the M11.

P2P measurements are interpolated to a fixed output size (a really small one - 8MP) and not at pixel level, so the results are comparitive but don't always translate to real world use. The graph above doesn't show the SL2 at 47MP or the M11 at 60MP but both interpolated down to 8MP.

The M11 files do have more DR than the SL2. I should, being newer but it's not massively different. It also has a different tone curve, so you can get different results depending on where you push and prod the files.

Gordon

[Guest]

42 minutes ago, FlashGordonPhotography said:

P2P measurements are interpolated to a fixed output size (a really small one - 8MP) and not at pixel level, so the results are comparitive but don't always translate to real world use. The graph above doesn't show the SL2 at 47MP or the M11 at 60MP but both interpolated down to 8MP.

The M11 files do have more DR than the SL2. I should, being newer but it's not massively different. It also has a different tone curve, so you can get different results depending on where you push and prod the files.

Gordon

M11 DR biggest advantage of 2 stops appears to be specifically at the new base ISO 64 (P2P graph doesn't account for ISO 64 being "pulled" with the SL2 = bad extrapolation on the graph). 2 stops is significant, if ISO 64 is appropriate - this seems to be supported by some early real world feedback

[FlashGordonPhotography]

6 hours ago, mzbe said:

M11 DR biggest advantage of 2 stops appears to be specifically at the new base ISO 64 (P2P graph doesn't account for ISO 64 being "pulled" with the SL2 = bad extrapolation on the graph). 2 stops is significant, if ISO 64 is appropriate - this seems to be supported by some early real world feedback

A 2 stop difference at a down sampled 8MP file does not translate to a direct comparison to the native resolutions of the cameras, unless your output is 8MP using the same down sampling method. It's nice to have some comparisons using consistent methodology but they're only valid at that output, using that down sampling routine. My normal prints require from around 50MP and larger as a native resolution. So I need to do my comparisons there, not at 8MP, to get an actual difference between the SL2 and M11. I think the M11 is somewhat better but not 2 stops at native resolution but I have yet to really stress test the M11 sensor.

If people are doing comparisons based on screen sized output I'm not surprised the M11 seems vastly superior to the SL2. At the sizes I work at the things might be very different.

For example P2P shows the ISO 64 of the A7R4 being as good as the X1D. Down sampled to 8MP it may be but at native resolution the X1D sensor is noticeably cleaner and more robust. Prints are noticeably different. Things that show at full resolution (like PDAF banding) often disappear when files are down sampled.

Gordon

[SrMi]

On 2/21/2022 at 2:19 PM, mzbe said:

M11 DR biggest advantage of 2 stops appears to be specifically at the new base ISO 64 (P2P graph doesn't account for ISO 64 being "pulled" with the SL2 = bad extrapolation on the graph). 2 stops is significant, if ISO 64 is appropriate - this seems to be supported by some early real world feedback

Bill Claff measurements (e.g., read noise) point to the fact that ISO 50 is native ISO for SL2. Leica representatives told me that ISO 50 is artificial ISO. I can see more DR with SL2's ISO 50, though the metering must be adjusted in order to prevent blowing highlights.

ISO 64 is incredible on M11 :). 

[SrMi]

On 2/21/2022 at 8:38 PM, FlashGordonPhotography said:

A 2 stop difference at a down sampled 8MP file does not translate to a direct comparison to the native resolutions of the cameras, unless your output is 8MP using the same down sampling method. It's nice to have some comparisons using consistent methodology but they're only valid at that output, using that down sampling routine. My normal prints require from around 50MP and larger as a native resolution. So I need to do my comparisons there, not at 8MP, to get an actual difference between the SL2 and M11. I think the M11 is somewhat better but not 2 stops at native resolution but I have yet to really stress test the M11 sensor.

If people are doing comparisons based on screen sized output I'm not surprised the M11 seems vastly superior to the SL2. At the sizes I work at the things might be very different.

For example P2P shows the ISO 64 of the A7R4 being as good as the X1D. Down sampled to 8MP it may be but at native resolution the X1D sensor is noticeably cleaner and more robust. Prints are noticeably different. Things that show at full resolution (like PDAF banding) often disappear when files are down sampled.

Gordon

P2P uses a common CoC not a downsampled file.

[Photoworks]

4 minutes ago, SrMi said:

Bill Claff measurements (e.g., read noise) point to the fact that ISO 50 is native ISO for SL2. Leica representatives told me that ISO 50 is artificial ISO. I can see more DR with SL2's ISO 50, though the metering must be adjusted in order to prevent blowing highlights.

Humm.....

so ISO 50 and underexpose 1-2 stops so that you don't blow the highlight, does it mean iso 100 or 200?

[SrMi]

Just now, Photoworks said:

Humm.....

so ISO 50 and underexpose 1-2 stops so that you don't blow the highlight, does it mean iso 100 or 200?

It is a half to one stop correction that is needed, AFAIR. To get better results, one could use the exposure setting from ISO 100 metering but with ISO 50. The difference may be small, though.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[FlashGordonPhotography]

3 hours ago, SrMi said:

P2P uses a common CoC not a downsampled file.

Is there a difference? Both describe a modified output, not pixel level examination.

Personally I don't shoot high resolution cameras for small prints or screen output. For that I shoot lower resolutions. Having equalised files like this is of little use to me.

Gordon

[SrMi]

24 minutes ago, FlashGordonPhotography said:

Is there a difference? Both describe a modified output, not pixel level examination.

Personally I don't shoot high resolution cameras for small prints or screen output. For that I shoot lower resolutions. Having equalised files like this is of little use to me.

Gordon

It affects when comparing cameras with different ratios, I guess.

Equalized files are the only way to compare cameras, though uncalibrated ISO must be taken into account (curves shift left/right).