Leica M11 - your next camera? {MERGED}

[pgk]

Advertisement (gone after registration)

1 hour ago, aristotle said:

There are even lenses (usually ones with low cutoff frequencies) that have better contrast at low frequencies even though they are significantly worse in terms of resolution.

Can you give an example or is this just philosophical?

[SrMi]

The main point is that one should not worry about a lens because of increased camera resolution. A better lens will always be a better lens.

[Advertisement]

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

[ianman]

On 12/3/2021 at 9:35 AM, Steven said:

Last but not least, there are still some things I havent told you about the M11, including the main thing it will be marketed around. I kept it for myself because some people in the thread have expressed the wish to not know everything and to keep space for surprise, which I will respect and honour. 

It’s solar powered? It’s got a fake winder that generates enough power to take the next picture? Voice control? It’s the last ever M?

[SrMi]

21 hours ago, Jeff S said:

You’re not following the discussion, or the science…

Jeff

On the other hand, there is this excerpt from a talk by Peter Karbe, where he talks about the balance between the lens and sensor performance (SL):

The Megapixel-Race: Can our Lenses keep up

[aristotle]

8 hours ago, pgk said:

Can you give an example or is this just philosophical?

Nothing specific, but on some airborne imaging systems that I've been involved with, the lenses are designed to be super accurate at resolving lower spatial frequencies but they tend to fall off of the cliff pretty quickly.  In these applications, the nature of the things being imaged and detected are well known in terms of limited high-spatial frequency content.  That's not particularly relevant here except to point out that system MTF is more than just a product of two scalar numbers, as might be interpreted by some from the quoted reference.

[Jeff S]

13 minutes ago, SrMi said:

On the other hand, there is this excerpt from a talk by Peter Karbe, where he talks about the balance between the lens and sensor performance (SL):

The Megapixel-Race: Can our Lenses keep up

Yes, I’ve seen.  Karbe wants to sell new lenses that he has designed for high contrast and resolution, effectively ones that will increase the multiplier that Cicala discusses. I wouldn’t expect him to say otherwise. Plenty of older generation, and other brand, lenses will of course still work well on SL bodies.  
 

Jeff

[aristotle]

Advertisement (gone after registration)

7 hours ago, SrMi said:

The main point is that one should not worry about a lens because of increased camera resolution. A better lens will always be a better lens.

That's for sure, but the way I took the question was the other way around.  Given the current lens lineup, is the 60MP M11 going to provide noticeably better resolution than a 40MP M10R.  Still, I think that the answer is yes, but it's not an unreasonable question.

Edited December 6, 2021 by aristotle

[Jeff S]

11 minutes ago, aristotle said:

That's for sure, but the way I took the question was the other way around.  Given the current lens lineup, is the 60MP M11 going to provide noticeably better resolution than a 40MP M10R.  Still, I think that the answer is yes, but it's not an unreasonable question.

The question was whether the lens could ‘keep up’ with 60MP.  We addressed the issues concerning that. To answer the question you posed, the system’s overall resolution will increase. Whether that’s ‘noticeable’ depends on a host of factors, including subject matter, technique, output medium and size, viewing distance, etc.

Jeff

[SrMi]

11 minutes ago, aristotle said:

That's for sure, but the way I took the question was the other way around.  Given the current lens lineup, is the 60MP M11 going to provide noticeably better resolution than a 40MP M10R.  Still, I think that the answer is yes, but it's not an unreasonable question.

I agree with 'yes' and 'not an unreasonable question' though I would have have preferred it without the double negative :).

[aristotle]

14 minutes ago, Jeff S said:

The question was whether the lens could ‘keep up’ with 60MP.  We addressed the issues concerning that. To answer the question you posed, the system’s overall resolution will increase. Whether that’s ‘noticeable’ depends on a host of factors, including subject matter, technique, output medium and size, viewing distance, etc.

Jeff

But they are related questions.  If the lens isn't "keeping up", it will limit the extent to which the increased sensor resolution makes a noticeable difference.  And yes, noticeable depends on how you are looking and what you are looking at, where on the imaging plane you're looking, what the orientation is relative to lens and sensor, etc.  The question is if we are close to the point of diminishing returns associated with increasing sensor resolution, and if so, to the extent that it's lens dependent, which lenses will best be able to take advantage of the increased resolution.

Again, the answer may be all of them, but I'm not aware of any data that indicates this.  To my knowledge, Leica doesn't publish MTF curves on its sensor, and the example in the quoted reference were just fictitious MTF numbers at an unspecified spatial resolution, wavelength, etc.

[Jeff S]

Read Roger again.  Unless the lens is crap, there’s no concern with ‘keeping up’.  And that’s not even considering all the real world differences beyond the resolution charts.  Even Roger doesn’t rely on the charts for his personal lens choices (see my other earlier link).

Jeff

[dpicciotto]

Hi guys, since we’re now approaching January (and M11 reveal?) how do you feel about prices of used M cameras? I’m an unhappy SL user (it is not what I was looking for) I’m currently looking at 240s, 262s or M10s to move to.
If the difference between now and after won’t be that much then I’d rather move ahead now and use the Christmas holidays to get familiar with the new camera. 
Even if beyond budget I’m inclined towards M10 because of better high ISOs. I’m used to L mount performance (in particular Sigma FP) and sometimes I really push it. 
Thoughts? Cheers

[Tailwagger]

Frankly, if I was worried about about anything lens related in light of even more MPx, it would be with RF accuracy and how well my lenses are calibrated.

[SrMi]

With every resolution upgrade, any manufacturer, there was always a Megapixel-Angst. So far it was always unfounded.

[aristotle]

1 hour ago, Jeff S said:

Read Roger again.  Unless the lens is crap, there’s no concern with ‘keeping up’.  And that’s not even considering all the real world differences beyond the resolution charts.  Even Roger doesn’t rely on the charts for his personal lens choices (see my other earlier link).

Jeff

I did read that post, and he doesn't define "crap".  The numbers that he provides as an example are at unspecified spatial resolutions and don't represent any real-world sensor or lens that I'm aware of at the boundaries of sensor spatial resolution.  The only way that one can say that we aren't at the point of diminishing sensor returns due to lenses not "keeping up" is if one evaluates the lens camera system using real transfer functions at spatial resolutions of interest, or better yet, uses the real camera / lens system in real life.  Again, I doubt that we are at that point, and I'm not saying that I'm concerned whatsoever.  But that post doesn't give any evidence one way or the other unless he indicated that he evaluated a 60MP Leica sensor with Leica lenses and I missed it.

I for one am not concerned with resolution in any case.  My M262 provides me with plenty of resolution.  I happen to really like the overall output of the M10R/M10M better for a variety of reasons, but resolution isn't one of them.  The M10R and M10M do provide clear resolution improvements though (on the order expected) when looking at 100% crops in comparison with the M262 for any lens I've used, and the meager spatial resolution increase associated with 60MP vs 40MP is unlikely to push beyond any boundary.  At some point that boundary will be pushed though, and it may be the case that the lenses "won't keep up" in the sense that it doesn't make sense to spend more on the sensor.  I remember in the early days (when we had 6MP crop sensors) where folks were doing the math to suggest that the limit was someplace around 80MP.  I don't know if that's realistic, or off by 3 dB.  And just as important, the sensor MTF may end up being limited by something other than pixel pitch at some point anyway meaning that we may never get to that point.  But the pedant in me finds the description in the quoted text wanting.  Because at some specified spatial resolution, either the camera or the lens will actually "suck" in comparison with the other.  So far, it's been the lens ahead of the camera.

[IkarusJohn]

1 hour ago, SrMi said:

With every resolution upgrade, any manufacturer, there was always a Megapixel-Angst. So far it was always unfounded.

Not sure about that - shutter slap was an issue on some early high resolution cameras.  It would seem that issue has been resolved.  Also, I doubt there would be the enthusiasm for image stabilisation, whether in camera (SL2) or optical in lens (the two SL zooms).  My view is that the resolution race hit the point of diminishing returns some time ago.  I wouldn’t turn a camera down due to high MP count, but I wouldn’t chose it for its high MP count either.  It seems to just go wit the territory of new cameras.

I would rather they put their effort into improved dynamic range, colour, noise, etc etc

[Jeff S]

24 minutes ago, aristotle said:

I did read that post, and he doesn't define "crap".  The numbers that he provides as an example are at unspecified spatial resolutions and don't represent any real-world sensor or lens that I'm aware of at the boundaries of sensor spatial resolution.  The only way that one can say that we aren't at the point of diminishing sensor returns due to lenses not "keeping up" is if one evaluates the lens camera system using real transfer functions at spatial resolutions of interest, or better yet, uses the real camera / lens system in real life.  Again, I doubt that we are at that point, and I'm not saying that I'm concerned whatsoever.  But that post doesn't give any evidence one way or the other unless he indicated that he evaluated a 60MP Leica sensor with Leica lenses and I missed it.

I for one am not concerned with resolution in any case.  My M262 provides me with plenty of resolution.  I happen to really like the overall output of the M10R/M10M better for a variety of reasons, but resolution isn't one of them.  The M10R and M10M do provide clear resolution improvements though (on the order expected) when looking at 100% crops in comparison with the M262 for any lens I've used, and the meager spatial resolution increase associated with 60MP vs 40MP is unlikely to push beyond any boundary.  At some point that boundary will be pushed though, and it may be the case that the lenses "won't keep up" in the sense that it doesn't make sense to spend more on the sensor.  I remember in the early days (when we had 6MP crop sensors) where folks were doing the math to suggest that the limit was someplace around 80MP.  I don't know if that's realistic, or off by 3 dB.  And just as important, the sensor MTF may end up being limited by something other than pixel pitch at some point anyway meaning that we may never get to that point.  But the pedant in me finds the description in the quoted text wanting.  Because at some specified spatial resolution, either the camera or the lens will actually "suck" in comparison with the other.  So far, it's been the lens ahead of the camera.

Then maybe you should read Roger’s myriad other posts on related (and other) lens testing, where he goes into technical detail. The one I linked, at a basic level, merely explains a common myth about lens/sensor resolution relationships in a way most can understand.
 

But to the important point about real life results (as I mentioned in my last post), perhaps this one is worth your time…

https://m.dpreview.com/opinion/4042117089/roger-cicala-why-i-dont-use-an-mtf-bench-to-test-my-own-lenses

Jeff

[SrMi]

3 minutes ago, IkarusJohn said:

Not sure about that - shutter slap was an issue on some early high resolution cameras.  It would seem that issue has been resolved.  Also, I doubt there would be the enthusiasm for image stabilisation, whether in camera (SL2) or optical in lens (the two SL zooms).  My view is that the resolution race hit the point of diminishing returns some time ago.  I wouldn’t turn a camera down due to high MP count, but I wouldn’t chose it for its high MP count either.  It seems to just go wit the territory of new cameras.

I would rather they put their effort into improved dynamic range, colour, noise, etc etc

The shutter shock is unrelated to high resolution. Olympus m43 cameras were probably the first to exhibit the issue (fixed in 2014). It was solved by replacing the first mechanical shutter (the closing shutter) with an electronic one. Leica M-s do not have that kind of shutter.  
The first IBIS was implemented for a 5MP camera (Minolta). It is not a high MP necessity. Hasselblad X1D works very well handheld, even though it does not have IBIS. The same goes for M10-R.
If I choose between a lower and a higher resolutions camera, and everything else is equal, I always choose the high-resolution camera. IMO, the only negative is the price.

 

[IkarusJohn]

12 minutes ago, SrMi said:

The shutter shock is unrelated to high resolution. Olympus m43 cameras were probably the first to exhibit the issue (fixed in 2014). It was solved by replacing the first mechanical shutter (the closing shutter) with an electronic one. Leica M-s do not have that kind of shutter.  
The first IBIS was implemented for a 5MP camera (Minolta). It is not a high MP necessity. Hasselblad X1D works very well handheld, even though it does not have IBIS. The same goes for M10-R.
If I choose between a lower and a higher resolutions camera, and everything else is equal, I always choose the high-resolution camera. IMO, the only negative is the price.

 

I first encountered shutter shock in what was considered to be a high resolution camera (Nikon d800e, "only" 37.5MP) - not actually a problem I've encountered since.  But then, I don't have a high resolution camera by today's standards - happy with 18MP (Monochrom), 24MP (M10-D, SL & TL2) and ?MP? (M-A) 😀

The X1D II was not a problem at all handheld, but then it's sensor is 50MP on a medium format size.  Never had an IBIS camera, and have no interest in getting one.

The largest I print is 1 metre long side, and resolution isn't a problem.  High MP seriously isn't a driver for me as I don't crop much more than straightening horizons.

Edited December 7, 2021 by IkarusJohn

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Jeff S]

6 minutes ago, IkarusJohn said:

 (Nikon d800e, "only" 37.5MP) -

Close… actually 36.3 MP. The Leica S006 and 007 are 37.5.

Jeff