Why not more pixels in the M camera?/ 36 MP {merged}

[pgk]

Advertisement (gone after registration)

But, the ultimate question remains. Is it possible to design a sensor with the correct micro lens structure. I would REALLY like to see someone who actually knows and does not just speculate or believe to tell about it. The answer would in my opinion provide the end to the discussion. If it is impossible, then the "there are better tools for that" -argument would actually be valid.

 

Obviously not - or at least not economically currently. If it was then Leica would no doubt do so. In the future perhaps it will be. But this doesn't end any discussion because there are many other factors which could influence the viability of an M camera with more MPixels. The fundamental (ultimate?) question remains open; are more MPixels relevant on an M camera?

[Advertisement]

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

[Hannes Lummes]

Obviously not - or at least not economically currently. If it was then Leica would no doubt do so. In the future perhaps it will be. But this doesn't end any discussion because there are many other factors which could influence the viability of an M camera with more MPixels. The fundamental (ultimate?) question remains open; are more MPixels relevant on an M camera?

Well...I don't know about others, but as long as I see moire and false color pixels in casual hand held photos the answer for me is "yes".

[Hannes Lummes]

It is not a matter of lens or astigmatism, just thickness of the sensor stack. Well known issue. I have all the bodies quoted above, and that since 2004 when the exact same size and pixel density on my D70's Sony sensor caused more moiré than that of the similar Sony sensor of both Nikon D100 and Epson R-D1 bodies. The extra-thin sensor stack of the M8 caused the same effect together with the IR issue we all know well here.

To clarify and extend my fairly simple comment: The sensor cover glass thickness causes change in astigmatism (and focus change and field curvature) but the real reason for less moire is always less resolution (compared to sensor chip resolution), and less resolution in sensor stacks is made by anti-alias, aka low pass, aka moire filters.  The thickness of the cover glass can be *anything* as long as it is properly calculated to the lens formulas. If it is not correct, like when putting M lenses on Sonys, the effect is bad astigmatism in outer zones of wide angle lenses. (which might, of course, mitigate moire)

You see different amounts of moire probably because of different strength anti-alias filters.

As long as a lens is very sharp, and sensor stack thickness matches it and sensor does not have anti-alias filter, there will be maximum amount of moire.

 

IR problems and cover glass thickness are of course related. And the reason Leica suffers from it in all M digital cameras to some extent is because they needed to keep the cover glass thin (and therefore not very effective for IR cut) *to not cause astigmatism* with wide angle lenses from the analog, i.e. non-cover-glass era.

 

(Edited typos)

Edited February 4, 2018 by Hannes Lummes

[Tailwagger]

Well, in a way, Leica is telling you. They appear to have decided that, with current technology, 24 MP  -or, possibly, something between approximately 20 and 30-  about is the sweet spot for a 135 class sensor in a reportage-style camera.

 

I do sense in this discussion a subtle undercurrent of mistrust.  Most unfailingly respect Leica's decisions with respect to optical formulation, yet when evaluating their ability to make the optimal set of decisions about sensor characteristics some seem to smell a rip-off.  Perhaps if they had the resources of Sony, we might have seen closer to the higher end of the range Jaapv cited. OTOH, beyond micro lensing, given all the criticism of the M240, the requirements necessary for producing a slimmer, lighter, the tightest possible packaging as well as faster startup, increase continuous frame rate, higher res external EVF, etc. may have imposed constraints on power and packaging that did not allow for MP levels much higher than what they landed on.

 

All engineering involves a set of compromises, both from a technical perspective as well as budgetary one. The M10 is no different. Was every decision made the optimal one? Of course not. But I trust the company was earnest in its development and handed us the best product they could. Clearly no camera ticks every box, addresses every need, fits every budget. But for many the M10 is the right tool. Other might wish the tradeoffs had fallen in an alternate direction that accommodated an MP spec that was competitive with the best currently available, but then I do wonder if any of those critics are pounding on Nikon or Sony in their respective forums as to why they've yet to make a range finder. 

Edited February 4, 2018 by Tailwagger

[pgk]

The thickness of the cover glass can be *anything* as long as it is properly calculated to the lens formulas.

 

An M series Leica is still expected to work with most (not all) M lenses designed for film cameras and LTM lenses when fitted with an adapter. Do you see any problem with this argument?

[jaapv]

Indeed. It’s very simple: buy into a legacy system, expect compromises to be slanted towards retro-compatibility.

[Hannes Lummes]

Advertisement (gone after registration)

I wrote: "The thickness of the cover glass can be *anything* as long as it is properly calculated to the lens formulas." Because Ict seemed to think that cover glass thickness is reducing moire.
You commented:

An M series Leica is still expected to work with most (not all) M lenses designed for film cameras and LTM lenses when fitted with an adapter. Do you see any problem with this argument?

You obviously did not read my post very carefully. In the end I specifically say: "they needed to keep the cover glass thin (and therefore not very effective for IR cut) *to not cause astigmatism* with wide angle lenses from the analog, i.e. non-cover-glass era."

It is good idea to first try to read and understand the whole post before comments.

[pgk]

What you are talking about is compromise. "Thickness can be anything ..... but has to be thin." and this is the point surely. And consequently cover glass thickness will have an impact on corner resolution. We've gone over and over this subject here on LUF. The current set of compromises are working well. All need to be adjusted to make gains and it isn't as simple as simply changing the sensor to a higher MPixel one due to micro-lens requirements, cover glass thickness/strength and a lack of lens to body communication. I do not envy the engineers at Leica who are balancing legacy issues with current high aspirations for the M system. My guess is that we are getting towards hi-fi territory - small gains require a great deal more effort than the cost implications might make viable.

[jaapv]

We are not getting towards - this has been the case since the M8.

[Lelmer]

"Why not more pixels in the M camera?"

 

Because I am perfectly happy with 24 Mpix and don't need more  

[Martin B]

Opinions can differ, but for me the 24 MP sensor is the main drawback of the M10, otherwise I would seriously consider. Why? Because I used a 22 MP sensor in my old Canon 5D MkII from 2009 and can easily compare the files with the ones obtained from my Sony A7R with 36 MP. The difference is very noticeable. I would love to use my Leica M lenses with their high resolution capability on an equal high MP/high DR FF sensor - I really benefit of the properties of such sensor in my photography with the A7R. 

[Herr Barnack]

Well, in a way, Leica is telling you. They appear to have decided that, with current technology, 24 MP  -or, possibly, something between approximately 20 and 30-  about is the sweet spot for a 135 class sensor in a reportage-style camera.

 

My thoughts exactly. 

 

I cannot imagine the amount of research, the sleepless nights, relentless obsession and the heartache that the Leica research and development folk went through to get this right in the M240 and M10 cameras. 

 

It's not like there's a megapixel roulette wheel in the Leica R&D department that they spin and the little ball just happened to drop in the slot numbered 24...

Edited February 5, 2018 by Herr Barnack

[Martin B]

My thoughts exactly. 

 

I cannot imagine the amount of research, the sleepless nights, relentless obsession and the heartache that the Leica research and development folk went through to get this right in the M240 and M10 cameras. 

 

It's not like there's a roulette wheel in the Leica R&D department that they spin and the little ball just happened to drop in the 24 MP slot... 

 

I doubt this. I find another reason more plausible: there are currently not many sensor suppliers out there which could deliver such high MP/high DR FF sensor. It was likely simply a cost and availability concern for Leica to stick with a (now outdated) 24 MP sensor design and older image processor. 

[Bill W]

I doubt this. I find another reason more plausible: there are currently not many sensor suppliers out there which could deliver such high MP/high DR FF sensor. It was likely simply a cost and availability concern for Leica to stick with a (now outdated) 24 MP sensor design and older image processor. 

 

Being an "older image processor" myself, and somewhat outdated my wife would say, I am satisfied with my 24 MP.

[Martin B]

Being an "older image processor" myself, and somewhat outdated my wife would say, I am satisfied with my 24 MP.

 

Nothing wrong with this - I used my 22 MP sensor for many years, too. My point above is more that I personally want to get the best tech out there for the $$$$ spent when I go for a new camera. We all can take fantastic photos with all kind of older cameras and sensors - but when it comes to investing into a new camera, I want better sensor tech than 24 MP. 

[jaapv]

I doubt this. I find another reason more plausible: there are currently not many sensor suppliers out there which could deliver such high MP/high DR FF sensor. It was likely simply a cost and availability concern for Leica to stick with a (now outdated) 24 MP sensor design and older image processor. 

Since when has cost held Leica back? And is "not many could deliver" and "outdated" not a contradiction?

Leica sets the specifications and finds a supplier. Don't forget that Sony (who can hardly be accused of supplying obsolete products) was unable to supply a sensor for the M240, Leica had to look elsewhere.

Nor do I think that pixel count makes a sensor "better" or "worse"

[pgk]

..... but when it comes to investing into a new camera, I want better sensor tech than 24 MP. 

 

And at the moment you cannot have it. If Leica could simply buy an off-the-shelf high MPixel sensor and slot it into to the M10 then all would be hunky dory. But ever since the 'perpetual upgrade' of the M8 this has not been possible. The reasons are numerous and as anyone who has been on this forum for a few years will tell you, they regularly need repeating. FWIW I too have been using a 22MPixel camera for a fair time and yet to me the 18MPixels of the M9 still produce comparable prints at near 30" x 20" and they are equal to those from a higher MPixel Nikon used by a colleague too.

 

So there are two questions: one is technical - can it be done to which the answer is currently, obviously no, but the other is about need. Whether you actually 'need' more MPixels, assuming the answer to the first question is no longer no, is about the relevance of more MPixels. How many people as a percentage of Leica M owners genuinely print to larger than 30" x 20" (cropping in my book is a lazy way to justify more MPixels)? I suspect not that many.

[lct]

I doubt this. I find another reason more plausible: there are currently not many sensor suppliers out there which could deliver such high MP/high DR FF sensor. It was likely simply a cost and availability concern for Leica to stick with a (now outdated) 24 MP sensor design and older image processor. 

 

As a Leica and Sony user, i can understand why Leica doesn't want the same kind of results A7 bodies can achieve with M lenses. If they share my viewpoint, Leica users don't want compromise on image quality and prefer a thin and light 24MP body to a bigger and heavier 42MP the same way as i prefer my 12MP Sony A7s mod to the bulkier A7r2 i've been using for my job.

[Advertisement]

Hallo,

Look here: Leica M

[sblutter]

I print M240 prints up to 40x60" (the limit of the printers available to me) with no loss whatsoever, even under a magnifying glass.  Its all about preparing a very well exposed file properly in Photoshop with up-rezing in several steps and so on

 

You can learn a lot in a multi day workshop, they're worth it if you print a lot

[pedaes]

.... prefer a thin...

 

Good point. Lets not forget the high priority given by M users to have a digital body of the same thickness as the film M's. Well, we got it, thanks Leica Designers and Engineers, and I for one are happy with any other 'features' negated by this move.  The 'back to the essentials' theme was very appealing  and the sensor size  sits well with this. 

Edited February 5, 2018 by pedaes