Why Leica is staying at 37.5MP for the S (Typ 007)

[Theodoros]

MF is certainly in the cross-roads of a major change... I don't believe there will be anymore new CCD sensors introduced. It certainly looks that Leica's decision to use CMOS was calculated years ago (since M240 and S-007 sensors are clearly close related) and that the competition was prepared to have an answer to it.
From this POV, one should consider that the new technology has to be adapted with the different needs MF users have from DSLR users, there is no good solution yet provided for tech and view camera users and sensor size is not picking from where CCD sensors are. But one should consider that Leica invested in Sinar which is a major player in the theca camera market and also produces MF backs for advanced use and of course the competition can't leave tech or view camera market unsupported... IMO, there will soon be an even larger size CMOS sensor (based on the current one) by Leica, that will be used as Sinar's entrance to Cmos technology and the same is to be expected from the other makers...  High Quality LV that is accompanying CMOS technology is a blessing for view  and tech users, but OTOH, cameras with movements don't like small pixels much...
   
I certainly believe that judging resolution based on CCD past experience, is not the right way of looking at things... 6microns is not a huge pixel size by any means as many suggest, I believe it has been chosen having in mind all the above and other considerations too (like low light performance). I believe that Leica's choice of pixel size will prove to be a better choice than competition very soon... and it will prove a winner to most of the possible market that these products aim for... the pros and artist photographers that make a living out of their cameras and the solutions their equipment provides to them....

[bcooter]

that the techies are predominant.

True.

Some use tech to sell you.  Some to unsell you.  Some to talk, some just because I guess they're bored and love charts.

But they're not talking technique . . . they be talking geek and yea I know the geeks have won.

God help us.



IMO

BC

[synn]

True.

Some use tech to sell you.  Some to unsell you.  Some to talk, some just because I guess they're bored and love charts.

But they're not talking technique . . . they be talking geek and yea I know the geeks have won.

God help us.



IMO

BC

Amen.

Tech talk to improve the pictures one takes is constructive and interesting. Tech talk for the sake of tech talk gets real old, real fast.

[hjulenissen]

Amen.

Tech talk to improve the pictures one takes is constructive and interesting. Tech talk for the sake of tech talk gets real old, real fast.

I'd suggest that striving for better pictures or a better technical understanding would both do you and everyone else good, while demanding that people should quit discussing certain things must be the least productive way to spend ones limited time.

One might speculate that you actively find discussion sections (" Equipment & Techniques") and threads ("Why Leica is staying at 37.5MP for the S (Typ 007)") where you know that there will be a technical debate, and then (rather than offer your experiences) tries to derail the discussion. The question is why?

-h

[synn]

I'd suggest that striving for better pictures or a better technical understanding would both do you and everyone else good, while demanding that people should quit discussing certain things must be the least productive way to spend ones limited time.

One might speculate that you actively find discussion sections (" Equipment & Techniques") and threads ("Why Leica is staying at 37.5MP for the S (Typ 007)") where you know that there will be a technical debate, and then (rather than offer your experiences) tries to derail the discussion. The question is why?

-h

Nobody is demanding anyone to do anything.
If the techies have a right to barge into every thread, tech related or otherwise and spill their infinite wisdom, the non techies have an equal right to say it bores the crap out of them.

[ErikKaffehr]

Hi,

Very good observation. I was considering it before posting, but decided not to elaborate. It seems that the discussion continues, so I post a response.

When I made the test with the feather, what I wanted to see if I would see fake detail on natural objects. I tested three cameras the P45+, Sony Alpha 99 and Sony Alpha 77.

	P45+	SLT99	SLT77	Leica S (Typ 007)
Pixel pitch	6.8	6.0	3.8	6
Microlenses	No	Yes	Yes	Yes
OLP filter	No	Yes	Yes	No

What I have seen here was that the P45+ and the Sony Alpha 99 were pretty close. The OLP filter on the SLT99 removed some colour aliasing, but monochrome aliasing was pretty similar. The SLT 77 with it's smaller pixels did show much less aliasing. That makes me feel that small pixels are more important for correct reproduction than OLP-filtering. Another way to say it, I feel the OLP-filters are designed to reduce colour aliasing. Another observation was that shooting distance and focal length did not matter a lot, I started with an 80 mm lens but continued with 150 mm.

Here is a link to that comparison: http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/FakeDetail/comparison1.html

I enclosed the most interesting part, below. The images are left to right P45+, SLT 99 and SLT 77. The SLT 99 has the same pixel pitch as the Leica S (Typ 007) but has OLP filter (Optical Low Pass). So, from where I stand, I would say that it is better to have an image with small pixels and downsize with a good algorithm than having large pixels and upsize with a good algorithm. The feather is a very good test for this, as it is very easy to what is to tell real detail from artificial detail.

I would agree, with posters stating that 37 MP and 50MP is not a large difference. My point is mainly that smaller pixels give better rendition, in the sense they produce less artificial detail.

Jim Kasson had an excellent series on the issue in june 2014, here is a link to one of the final articles in that series: http://blog.kasson.com/?m=20140630 . That article series is worth reading for the technically minded.

What Jim has found was that about 2 micron pixels were needed to pull best information of the Zeiss Distagon 55/1.4 (Otus) lens he was using for reference.
 
Best regards
Erik

Never the less, one can't avoid noticing, that Erik uses examples with comparison of Sony's 24mp Cmos 24x36 sensor and Kodak's old 39mp CCD MF sensor ...and then addresses his (questionable IMO) findings as if Leica's new sensor will behave exactly the same as the old CCD MF sensor!!!  All this of course is despite the fact that Leica's new sensor has much more many commons with... Sony's sensor (both of exactly the same pixel size, both Cmos, both with micro lenses)!!!!

[hjulenissen]

... an equal right to say it bores the crap out of them.

You can tell me that I am boring. I can tell you that you are boring. And so forth. What good does it do to any of us? Please feel free to PM me.

-h

[BernardLanguillier]

Nobody is demanding anyone to do anything.
If the techies have a right to barge into every thread, tech related or otherwise and spill their infinite wisdom, the non techies have an equal right to say it bores the crap out of them.

Indeed. I think that this whole tech talk on the 007 results from the fact that people see a gap between the marketing sensor tech talk of Leica and their understanding of its actual merits relative to the competition. In other words, Leica marketing may have made a mistake by trying to further rationalize the value of the 007 in terms of sensor technological advance that may not be as clear cut as they claim.

My personal view remains that the S system is a fascinating offering, mostly thanks to is remarkable lenses.

In that context, the 007 has closed a lot of the gaps that existed relative to its DSLR competitors (645Z, D810, a7r) and that in itself means it is much better value than the previous S cameras. The system now has capabilities that make it universally appealing.

Whether the pricepoint makes sense or not depends on everyone's needs and situation. Not needing shutter leaf lenses, I personally think it is way over priced.

Cheers,
Bernard

[synn]

Sure. And how many times will you repeat yourself before you bore yourself out?

-h

At least as many times as people find the time to point it out.

[EricWHiss]

Theodoros,

 I would hope he is more careful than Leica engineers, because if reactors had the same rate of defects as Leica cameras, life near them would be dangerous - I had 4 M8 bodies!

Edmund

Yes, Edmund, but you know what they say….  "Those that can do, do, those that can't teach"

[Chris Livsey]

In the spirit of technical accuracy, which this thread seems to have centred on:-

He who can, does. He who cannot, teaches.

Man and Superman (1903) "Maxims for Revolutionists"
George Bernard Shaw

[Theodoros]

This all. doesn't take into account that different sensors are designed to serve different tasks... In other words, there are only some that think of MF as DSLR replacement. In my view, if I am to use MF only for what a DSLR can do, I may as well use a DSLR and save the money.... But a DSLR, won't provide me with MS, it will have limited use when movements are needed, doesn't offer the modularity as to change finders or screens or backs a traditional MF system offers and doesn't offer the colour accuracy needed for repro work... That's why I'm stuck with C645 and Fuji GX680 or others with Mamyia RZ or Hassy-V or Rollei, or are blaming Hasselblad for "closing" the H system... (which they now seem to try and open back bit by bit).

One may easily notice that in industry the larger the sensor the larger the pixels, in other words, pixel density is tighter the smaller the area, there is a good explanation for this since a sensor receives photons not only on its centre but on the edges and corners too and of course, these edge pixels must perform up to a level and the entrance angle increases with larger formats... let alone if lens movements come into the equation...

As a result one has to understand Leica-S positioning into the market and link it with their decisions... For people that upgrade from DSLR just to improve their IQ in common DSLR use, it does exactly that, for people with MF that want to keep their systems and only want to upgrade their back and abandon their DSLR system altogether thus shrinking the equipment needed while increase quality at the same time... it does exactly that.. For those that want to upgrade their MFDBs or want to use a camera with movements, they now own Sinar and Leica has their MFDBs to consider too...

Clearly, the camera provides a link that one may consider if he wants to both replace his DSLR system altogether and integrate it into his MF system keeping the rest of it... That's why the adapters! ....and of course it is sensible for them to decide on a sensor that would provide ONE base to serve all they are planning... a sensor base that will be able to cope with m, S or Sinar... It's simple isn't it? It really is! I believe we'll be surprised with the quality of their sensor in four months time when it will be introduced! I believe High ISO performance will exceed anything we are used up to now, but it will also perform very well on a Leica-P3 view camera and colour accuracy and DR will be right at the top as one would expect from a similarly priced MFDB... I also expect great LV performance and a new Sinarback based on the same sensor technology soon...

[EricWHiss]

Chris,
Perfect thanks for the quote and reference too. :-)

Though there is indeed a lot of technical chatter, I'm wondering how much of it is truly relevant?   I mean what percentage of images are ruined by color aliasing?  What percentage of those can't be fixed with simple photoshop work?  When you shop for a camera system - at what level should aliasing come up in the purchase decision criteria?

There is one forum member particularly concerned with it, almost to obsession, but does color aliasing impact the rest of you in any real way?   I've never had a concern over it because it doesn't seem to come into play in my images, even those of feathers, because I can control it with distance and aperture. On the few occasions its present, it seems most viewers never notice it.  

[Bart_van_der_Wolf]

In the spirit of technical accuracy, which this thread seems to have centred on:-

He who can, does. He who cannot, teaches.

Man and Superman (1903) "Maxims for Revolutionists"
George Bernard Shaw

And from the same author:
"A fool's brain digests philosophy into folly, science into superstition, and art into pedantry. Hence University education."



Cheers,
Bart

[Bart_van_der_Wolf]

Though there is indeed a lot of technical chatter, I'm wondering how much of it is truly relevant?   I mean what percentage of images are ruined by color aliasing?  What percentage of those can't be fixed with simple photoshop work?  When you shop for a camera system - at what level should aliasing come up in the purchase decision criteria?

Hi Eric,

That would probably depend on the type of subjects one shoots. I can imagine that those into fashion/fabrics and product related photography, and those in architecture would certainly want to have a more predictably usable image quality. Those who shoot birds (the feathered variety), would probably also prefer something that produces less aliasing artifacts.

Post-processing costs time/money, so if one has a choice it's better to avoid issues to begin with. Being educated and aware helps, ignorance or even denial doesn't.

Cheers,
Bart

[BernardLanguillier]

As a result one has to understand Leica-S positioning into the market and link it with their decisions... For people that upgrade from DSLR just to improve their IQ in common DSLR use, it does exactly that...

Does it?

I'll wait for the DxO mark results before deciding if a 007 + 70mm f2.8 is superior to a D810 + Otus 55mm f1.4.

Based on what we know today, I see little reason to think it would be at all and none to think it would be significantly.

Cheers,
Bernard

[ErikKaffehr]

Hi,

What I see is a fairley large percentage of my shots with the P45+. Typical areas affected:

- Architecture
- Water surfaces
- Tree tops and vegetation

It seems that there will be colour aliasing at f/11 and below but it goes away at f/16. But I don't want to stop down to f/16, because I loose sharpness and sharpness is what I am using MF for.

There is probably a lot of aliasing in nature shots, too, but it is not easily seen. Personally, I don't enjoy fixing artefacts in Photoshop.

With birds, I don't know. I have shown that we can have artificial effects on individual feathers, but I guess that some artificial detail can enhance the image many times. I often see some detail in animal fur that I am pretty sure is beyond the resolution of the sensor.


Best regards
Erik

Hi Eric,

That would probably depend on the type of subjects one shoots. I can imagine the those into fashion/fabrics and product related photography, and those in architecture would certainly want to have a more predictably usable image quality. Those who shoot birds (the feathered variety), would probably also prefer something that produces less aliasing artifacts.

Postprocessing costs time/money, so if one has a choice it's better to avoid issues to begin with. Being educated and aware helps, ignorance or even denial doesn't.

Cheers,
Bart

[eronald]

Nobody is demanding anyone to do anything.
If the techies have a right to barge into every thread, tech related or otherwise and spill their infinite wisdom, the non techies have an equal right to say it bores the crap out of them.

Indeed my dear, it really is so frightfully sad that those servant techies get uppity and are seen above stairs these days. But you know, the world isn't what it used to be.

Here in Paris we still know how to do things:

- The Photographer is that handsome man with the leather jacket who is standing next to the AD. All the girls at Elle love him, and he is seen at every nightclub, gets the jobs, writes the invoices and as long as he's here he gives the essential instructions like "Run down to Louboutin and get us some matte pumps" or "Liliane, time to sparkle for the next shot, do go and powder your nose". He'll make sure to be seen leaving with a couple of models as soon as the shoot wraps.

- The  assistant is that miserable mousy thing standing next to the camera cart, who has just stepped back from the camera. He really needs to wash his hair, in fact it looks like he should take a bath.  But he won't have much time for that tonight because after the shot he needs to pack everything and then he has to get the first batch of previews retouched.

 You see?  Artists and techies each have the life they deserve. Now, isn't that the way it should always be?

Edmund

[Theodoros]

Does it?

I'll wait for the DxO mark results before deciding if a 007 + 70mm f2.8 is superior to a D810 + Otus 55mm f1.4.

Based on what we know today, I see little reason to think it would be at all and none to think it would be significantly.

Cheers,
Bernard

You'll make your decision depending on DXO tests?  I'll trust my decision based on my own tests alright...

[Theodoros]

Hi,

What I see is a fairley large percentage of my shots with the P45+. Typical areas affected:

- Architecture
- Water surfaces
- Tree tops and vegetation

It seems that there will be colour aliasing at f/11 and below but it goes away at f/16. But I don't want to stop down to f/16, because I loose sharpness and sharpness is what I am using MF for.

There is probably a lot of aliasing in nature shots, too, but it is not easily seen. Personally, I don't enjoy fixing artefacts in Photoshop.

With birds, I don't know. I have shown that we can have artificial effects on individual feathers, but I guess that some artificial detail can enhance the image many times. I often see some detail in animal fur that I am pretty sure is beyond the resolution of the sensor.


Best regards
Erik

still... iT'S NOT YOUR P-45 ON TEST!