M10-P or M10R

[pedaes]

Advertisement (gone after registration)

4 minutes ago, jonoslack said:

But I don't think the highlight issue with the M10 is to do with dynamic range, but with a characteristic of the sensor design (which is fixed in the M10-R).

It is always interesting these things only come to light (excuse pun) when the next upgrade arrives. Same with road testers and cars.😉

[jonoslack]

18 minutes ago, pedaes said:

It is always interesting these things only come to light (excuse pun) when the next upgrade arrives. Same with road testers and cars.😉

Hah! I don't think so. https://www.slack.co.uk/m10-highlights.html

This was published in summer 2018 - long before the M10-R was even a twinkle in anybody's eye

. . . apologies for misplaced sarcasm gratefully accepted 😂

Edited July 14, 2021 by jonoslack

[Advertisement]

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

[Guest Nowhereman]

1 hour ago, jonoslack said:

...I don't know about these figures, but I'm quite certain that in this context shadow recovery is not at all related to highlight recovery - actually, in terms of dynamic range you might assume that better shadow recovery would mean worse highlight recovery (and vice versa)...

If that is the case, there must be an explanation for this, but I've no idea what it is — so, still puzzled.
________________________
Frog Leaping photobook

[jonoslack]

2 minutes ago, Nowhereman said:

If that is the case, there must be an explanation for this, but I've no idea what it is — so, still puzzled.
________________________
Frog Leaping photobook

Hi There 

Of course there is an explanation - although I don't think anyone from Leica is going to provide it, and I'm sure it won't be explained by 3rd party dynamic range measurements (they're almost all more interested in noise - poor highlight recovery is much harder to measure for one thing).

But the fact that you are puzzled as to what the cause might be doesn't alter the fact of the existence of the difference between the M10 and the M10-R. 

Of course, one doesn't want to overstate it, it's fine as long as one is cautious with the highlights - and especially avoid ISO 100 when it's very contrasty. 

all the best

[Guest Nowhereman]

@jonoslack - I start with the premise that there is no point in shooting ISO 100 on the M10 if you're shooting into the light or into strong highlights — so, I simply throw that out of the hopper.

Your statement was, in this context shadow recovery is not at all related to highlight recovery - actually, in terms of dynamic range you might assume that better shadow recovery would mean worse highlight recovery (and vice versa). My conclusion would be the reverse, that better shadow recovery does mean better highlight recovery: as long as you underexpose commensurately on the M10, you recover that much from the highlights. And, in this, I don't see how the better highlight recovery would not be reflected in the two sets photonstophotos graphs discussed. I don't have the knowledge to say more, but wonder if @adan, who has looked into the M10 DNG contrast curve, could enlighten us on this.
________________________
Frog Leaping photobook 

[SJH]

I prefer Jone's reviews but this one reads pretty well re the M10R and why he thinks it's so good https://www.thephoblographer.com/2021/07/12/weve-updated-our-leica-m10r-review-its-amazing/ It also covers some good points regarding colour etc with C1. 

[jonoslack]

Advertisement (gone after registration)

33 minutes ago, Nowhereman said:

@jonoslack - I start with the premise that there is no point in shooting ISO 100 on the M10 if you're shooting into the light or into strong highlights — so, I simply throw that out of the hopper.

Your statement was, in this context shadow recovery is not at all related to highlight recovery - actually, in terms of dynamic range you might assume that better shadow recovery would mean worse highlight recovery (and vice versa). My conclusion would be the reverse, that better shadow recovery does mean better highlight recovery: as long as you underexpose commensurately on the M10, you recover that much from the highlights. And, in this, I don't see how the better highlight recovery would not be reflected in the two sets photonstophotos graphs discussed. I don't have the knowledge to say more, but wonder if @adan, who has looked into the M10 DNG contrast curve, could enlighten us on this.
________________________
Frog Leaping photobook 

Well, from a logical point of view, if you have 7 stops of dynamic range in two cameras and one camera has 1 stop better shadow recovery, then that must mean logically that that camera has one stop less highlight recovery!

But that remark was an aside, and I don't believe it's germane to the issue of the difference in highlight recovery between the M10 and M10-R, because I don't believe that has anything to do with dynamic range (which is why it wouldn't be represented in the Photontophotos graph). 

Probably the technical people at Leica understand it properly, because they cured it with the M10-R sensor, but (in my mind quite rightly) they don't get involved in this kind of discussion. 

I think your puzzlement (your word) is due to the fact that you are assuming an identical linear reduction in the ability to recover highlights on the two cameras as the image is increasingly overexposed. But this is demonstrably not the case, when you overexpose with the M10 you very quickly lose all detail and get blotchy discolouration when you try and reduce the exposure in post processing - when you do the same amount of overexposure on the M10-R you find that there is still detail, and you don't get the blotchy discolouration.

As I say (once again), I don't think this is anything to do with dynamic range, and trying to understand it with relationship to dynamic range is bound to cause confusion. I think it's a minor flaw with the sensor which I'm assuming is a hardware problem (or Leica would certainly have fixed it in firmware). 

 

 

[Guest Nowhereman]

24 minutes ago, jonoslack said:

...I think your puzzlement (your word) is due to the fact that you are assuming an identical linear reduction in the ability to recover highlights on the two cameras as the image is increasingly overexposed. But this is demonstrably not the case, when you overexpose with the M10 you very quickly lose all detail and get blotchy discolouration when you try and reduce the exposure in post processing - when you do the same amount of overexposure on the M10-R you find that there is still detail, and you don't get the blotchy discolouration.

As I say (once again), I don't think this is anything to do with dynamic range, and trying to understand it with relationship to dynamic range is bound to cause confusion. I think it's a minor flaw with the sensor which I'm assuming is a hardware problem (or Leica would certainly have fixed it in firmware). ...

The issue may, indeed, be right here but I can't help thinking that, rather than a hardware problem, it could stem from the contrast curve that Leica placed on the M10 DNG in order to get better SOOC images, more like the M9(?) and not as flat as the M204(?). That's why I'm trying to see if @adan can shed any light on this, as he has looked into what kind of contrast curve Leica may have baked into the M10 firmware.
________________________
Frog Leaping photobook

Edited July 14, 2021 by Nowhereman

[jonoslack]

35 minutes ago, Nowhereman said:

The issue may, indeed, be right here but I can't help thinking that, rather than a hardware problem, it could stem from the contrast curve that Leica placed on the M10 DNG in order to get better SOOC images, more like the M9(?) and not as flat as the M204(?). That's why I'm trying to see if @adan can shed any light on this, as he has looked into what kind of contrast curve Leica may have baked into the M10 firmware.
________________________
Frog Leaping photobook

Okay - If it was in the contrast curve for SOOC images then they would surely have fixed it long ago! I've spent hundreds of hours on this over the years, including with Leica in Wetzlar; I don't know the full explanation for it, but I know that it exists and can easily demonstrate it and I know that it cannot be solved in the firmware. 

But really, it's not that big a deal - I shoot with both the M10 and the M10-R on a regular basis, both cameras take great pictures and of course it's possible to spoil a photo on any camera by getting the exposure wrong!

 

Edited July 14, 2021 by jonoslack

[SrMi]

6 hours ago, Nowhereman said:

Thanks @jonoslack, @Steven, @Jeff S - I take the point you're making on the advantage of the M10-R over the M10 in highlight recovery. What still confuses me is the following statement:

...and the following chart, which shows substantially better shadow recovery at ISO 200 by the M10 compared to the M10-R. I may be misunderstanding this, but I would have thought that if, you shoot at ISO 200, and expose for the highlights, and raise shadows in post, you should end up better with the M10 than with the M10-R — or at least the same. I am not being obstreperous — simply confused by this.

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!

________________________
Frog Leaping photobook

You do not account for the exposure, the main factor of noise.

What the the graph above shows is that the noise is reduced (shadow performance improved) as the ISO is raised. When the graph is flat, it means that increasing ISO does not help any longer. From docs:

"For those shooting raw it can be useful to know when raising ISO in the camera has little or no advantage over applying digital gain in post processing."

Big jumps in the curve are indication of dual conversion gain.

[SrMi]

3 hours ago, jonoslack said:

Well, from a logical point of view, if you have 7 stops of dynamic range in two cameras and one camera has 1 stop better shadow recovery, then that must mean logically that that camera has one stop less highlight recovery!

<snip>

Dynamic range is not related to highlight recovery. However, a reduced dynamic range does manifest itself in less flexibility when choosing between blown highlights and noisy shadows.
Blown highlights are related to metering. Some manufacturers tune their metering for less noise in shadows and more blown highlight, while others tune it for safe highlights. There may be other in-camera factors in play, though.

[SrMi]

9 hours ago, jonoslack said:

Hi There

No - I use what I'm told is the base ISO to determine base ISO 😁. it's about 160 on the M10 and about 100 (probably a bit less) on the M10-R. 

Hi Jono,

Thank you for the correction. I assume your source is Leica. Unfortunately, PhotonsToPhotos is missing the ISO 160 data.

I have received conflicting information about native ISO from Leica (e.g., SL2 and Q2): measurements do not correspond with Leica's declaration of native ISO. On the other hand, the lowest ISO bound in Auto-ISO is a good indication of 'predictable' ISO (i.e., one can rely on automatic metering). 

The most important thing is still: know thy camera :), and you know yours well!

Best,

- Srdjan

[jonoslack]

18 minutes ago, SrMi said:

Dynamic range is not related to highlight recovery. However, a reduced dynamic range does manifest itself in less flexibility when choosing between blown highlights and noisy shadows.
Blown highlights are related to metering. Some manufacturers tune their metering for less noise in shadows and more blown highlight, while others tune it for safe highlights. There may be other in-camera factors in play, though.

I shouldn't have said 'recovery' at either end, I just meant that 7 stops is 7 stops!

[jonoslack]

3 minutes ago, SrMi said:

Hi Jono,

Thank you for the correction. I assume your source is Leica. Unfortunately, PhotonsToPhotos is missing the ISO 160 data.

I have received conflicting information about native ISO from Leica (e.g., SL2 and Q2): measurements do not correspond with Leica's declaration of native ISO. On the other hand, the lowest ISO bound in Auto-ISO is a good indication of 'predictable' ISO (i.e., one can rely on automatic metering). 

The most important thing is still: know thy camera :), and you know yours well!

Best,

- Srdjan

Hi Srdjan

There's plenty of conflicting information about native ISO, in this instance I'm pretty sure I have it right for the M10 (from more than one source). I'm not so certain about the M10-R except that it is either 100 or a little less than that. 

Quite agree, the important thing is to 'know thy camera' 

best

Jono

[Dennis]

I messed up a couple exposures/pictures a few days ago. I over-exposed two frames by 4 stops. With My Nikon D5, I was surprisingly able to recover them... But it was a case. I usually expose correctly or so. After many years, it should be the bread of every day, right? My point is that it is great to have a camera that recovery so much. It's not necessary, but it's definitely a plus.
It's not (maybe) the best practice because one always relies on camera boundaries and "extreme" situations. But I understand  Steven's point. He loves to shoot wide open in bright light with its M10-R and then recovery blown highlights. And he doesn't want to use/carry ND filters because he doesn't like them. He wants to see w/o any help or extra piece of glass what a single exposure can do under its own condition and modus operandi. It works like a charm for him. 

What is happening to the "IQ" after a challenging recovery? In a way or another, maybe, we perceive a loss. Don't you think?
10-5 years ago, the differences were very notable. Nowadays, maybe they are minimal, but I guess it is worth trying to get the best image possible from the beginning. Apart from vision, composition, we just have to use a clever and effective combination of three values. Right? It's not a big deal.
. 
I'm not a pixel peeper; I just love to take photos whenever I can. I can't read the photons charts. But I really believe, for example, that Manual ISO will always be better than the Auto-ISO. Why? Because at the lowest ISO, you get more DR, more details, less noise. Am I correct? So, conditions allowing, I always use the lowest ISO possible.
Setting the camera for auto iso, minimum shutter speed at 1/250 (to say one), forces a combination such as 1/250 and 3200 iso, rather than 1/60 800 iso.
If I don't need 1/250 and can handle 1/60 handheld, I get the same image, but with "more genuine" quality to tell it in a way. 

Today, I would choose M10-P over the R. It's just enough for what I want. And if the blur motion issue (in high MP FF cameras) is there, do I must switch my humble 1/60 to 1/250? Did you see it? It's a conspiracy to shoot anyway at 3200 iso, LOL.

Please tell me what you think. When we recover a lot (up to the camera capabilities) and choose a higher iso (vs. Lowest one), can we observe quality loss or not? 🤔

My2c.

[SrMi]

28 minutes ago, Dennis said:

I messed up a couple exposures/pictures a few days ago. I over-exposed two frames by 4 stops. With My Nikon D5, I was surprisingly able to recover them... But it was a case. I usually expose correctly or so. After many years, it should be the bread of every day, right? My point is that it is great to have a camera that recovery so much. It's not necessary, but it's definitely a plus.
It's not (maybe) the best practice because one always relies on camera boundaries and "extreme" situations. But I understand  Steven's point. He loves to shoot wide open in bright light with its M10-R and then recovery blown highlights. And he doesn't want to use/carry ND filters because he doesn't like them. He wants to see w/o any help or extra piece of glass what a single exposure can do under its own condition and modus operandi. It works like a charm for him. 

What is happening to the "IQ" after a challenging recovery? In a way or another, maybe, we perceive a loss. Don't you think?
10-5 years ago, the differences were very notable. Nowadays, maybe they are minimal, but I guess it is worth trying to get the best image possible from the beginning. Apart from vision, composition, we just have to use a clever and effective combination of three values. Right? It's not a big deal.
. 
I'm not a pixel peeper; I just love to take photos whenever I can. I can't read the photons charts. But I really believe, for example, that Manual ISO will always be better than the Auto-ISO. Why? Because at the lowest ISO, you get more DR, more details, less noise. Am I correct? So, conditions allowing, I always use the lowest ISO possible.
Setting the camera for auto iso, minimum shutter speed at 1/250 (to say one), forces a combination such as 1/250 and 3200 iso, rather than 1/60 800 iso.
If I don't need 1/250 and can handle 1/60 handheld, I get the same image, but with "more genuine" quality to tell it in a way. 

Today, I would choose M10-P over the R. It's just enough for what I want. And if the blur motion issue (in high MP FF cameras) is there, do I must switch my humble 1/60 to 1/250? Did you see it? It's a conspiracy to shoot anyway at 3200 iso, LOL.

Please tell me what you think. When we recover a lot (up to the camera capabilities) and choose a higher iso (vs. Lowest one), can we observe quality loss or not? 🤔

My2c.

A raw image has three channels (red, blue, and green). Once you blow one channel, it is gone, kaputt. However, most post-processing software can reconstruct one or two blown channels, though it may lead to discoloration. IMO, the best way to deal with blown highlights is a clone tool.

How do you know if you have blown a highlight? Adobe software is not reliable in that respect, i.e., what the software marks as blown may or may not be blown highlight. If the highlights are not blown in the raw files, the recovery can be done without any image degradation.

The topic is interesting but off-topic to this thread and sub-forum. If more discussion is necessary, we could start a new thread.

[SrMi]

1 hour ago, jonoslack said:

Hi Srdjan

There's plenty of conflicting information about native ISO, in this instance I'm pretty sure I have it right for the M10 (from more than one source). I'm not so certain about the M10-R except that it is either 100 or a little less than that. 

Quite agree, the important thing is to 'know thy camera' 

best

Jono

Thank you for the information, Jono.

Best,

- Srdjan

[Mark II]

33 minutes ago, Dennis said:

10-5 years ago, the differences were very notable. Nowadays, maybe they are minimal, but I guess it is worth trying to get the best image possible from the beginning. Apart from vision, composition, we just have to use a clever and effective combination of three values. Right? It's not a big deal.

This, more than anything. Pretty much all modern full-frame cameras have very similar dynamic range and noise performance, as the sensors are now pretty much limited by physics rather than engineering. There is perhaps another 1/2 stop at best that could be achieved with some magical newer technology - otherwise the only way to see an improvement is by removing the CFA (monochrome sensors) or moving to a larger sensor size (more light capture).

Where some sensors still do better than Leica today, it is primarily in their lower ISO settings. For example, the Z7 has a native (non-pull) ISO 64 with enough well capacity that it can yield higher dynamic range and lower noise. However, at ISO 200 or so and higher, pretty much all full-frame 35mm colour sensors perform the same today. All that changes is software that tweaks the relative highlight/shadow recovery balance, and occasionally differences in the colour-filter-array that trade off colour characteristics for sensitivity (for example, see the Q2 vs M10).

Even resolution is now hitting limits, where the only way to avoid diffraction softening from limiting the effective resolution is to use ever wider apertures. This is responsible for the current generation of monstrously large mirrorless prime lenses, which are trying to hit apertures of f1.2 or wider yet still resolve sharply on a 50MP FF sensor (think the latest fast 50mm primes from Canon and Nikon). For Leica M cameras this is particularly a problem, as the huge lens as a tradeoff for resolution is not really viable with a rangefinder - making compact lenses which are free of aberrations either impossible for a 50MP sensor, or prohibitively expensive due to the exotic glasses and manufacturing tolerances that may be required.

As to the original question, M10-P to M10-R is only a 30% increase in linear resolution. The main issue should be whether or not the increased resolution is worth the increased file sizes and shorter battery life. As to highlight recovery, I am fairly certain that simply dialling in some negative exposure compensation on the M10-P and processing RAW accordingly would eliminate most differences - albeit with more effort and care.

[SrMi]

5 hours ago, Mark II said:

This, more than anything. Pretty much all modern full-frame cameras have very similar dynamic range and noise performance, as the sensors are now pretty much limited by physics rather than engineering. There is perhaps another 1/2 stop at best that could be achieved with some magical newer technology - otherwise the only way to see an improvement is by removing the CFA (monochrome sensors) or moving to a larger sensor size (more light capture).

Where some sensors still do better than Leica today, it is primarily in their lower ISO settings. For example, the Z7 has a native (non-pull) ISO 64 with enough well capacity that it can yield higher dynamic range and lower noise. However, at ISO 200 or so and higher, pretty much all full-frame 35mm colour sensors perform the same today. All that changes is software that tweaks the relative highlight/shadow recovery balance, and occasionally differences in the colour-filter-array that trade off colour characteristics for sensitivity (for example, see the Q2 vs M10).

Even resolution is now hitting limits, where the only way to avoid diffraction softening from limiting the effective resolution is to use ever wider apertures. This is responsible for the current generation of monstrously large mirrorless prime lenses, which are trying to hit apertures of f1.2 or wider yet still resolve sharply on a 50MP FF sensor (think the latest fast 50mm primes from Canon and Nikon). For Leica M cameras this is particularly a problem, as the huge lens as a tradeoff for resolution is not really viable with a rangefinder - making compact lenses which are free of aberrations either impossible for a 50MP sensor, or prohibitively expensive due to the exotic glasses and manufacturing tolerances that may be required.

As to the original question, M10-P to M10-R is only a 30% increase in linear resolution. The main issue should be whether or not the increased resolution is worth the increased file sizes and shorter battery life. As to highlight recovery, I am fairly certain that simply dialling in some negative exposure compensation on the M10-P and processing RAW accordingly would eliminate most differences - albeit with more effort and care.

FYI, here is an article with a table for diffraction vs. resolution:
Diffraction and ultimate FF pixel count
E.g., stopping down to f/8 limits the resolution to 45MP on monochrome and 179MP on Bayer sensor.
Takeaway: "We are a long, long way from having sensors whose resolution would be limited by diffraction at apertures like f/5.6."

[Advertisement]

Hallo,

Look here: Leica M

[beewee]

7 hours ago, SrMi said:

But I really believe, for example, that Manual ISO will always be better than the Auto-ISO. Why? Because at the lowest ISO, you get more DR, more details, less noise. Am I correct?

This depends on the sensor design. In most cases, no. Especially when higher ISO is needed.

In many modern sensor designs, M10 and M10-R included, they take advantage of a dual-stage gain architecture. This means that when you set the ISO above a certain threshold (as governed by sensor design), you will see a sudden drop in noise and bump up in dynamic range versus under exposing by even 1 stop and cranking up the exposure slider in post processing. This is because, at some threshold that is dependent on the sensor design, the sensor will switch over to a different low noise amplifier circuit on the sensor that is optimized for higher amplification needed due to lower signal (light).