Bad times for LEICA S- guess they will be severely hurt

[araucaria]

I've always wondered if the leica glass covers full 645. I guess it does except maybe the wideangle 24 3.5

[paratom]

I have shot 5dIII and Leica S for some years now and find the combination of larger sensor and the S lenses to lead to visible different and better IQ. Not just sharpness but also color and specially rolloff from focus plane to background.
Different look.
I dont think 50MP will change this.

[BobShaw]

If I could use medium format for everything I would. For the occasions that I can't there is 35mm.

[Iluvmycam]

Hi dear Collegues,

what do You think? Guess LEICA will be in trouble on the medium format side of things by Canons recent move.
If the magic Lantern guys open up the EOS firmware, allow 4K video and high ISO/Dual ISO, there´s no need to invest double the budget for almost half the resolution.
They have to react or use another sensor.

Leica seems to always have sensor trouble. The MM has a very sharpsensor, but it has trouble with highlights. The M240 has a second rate sensor. No it is not sour grapes or jealousy. I had 5 Leicas, now 4. I have many other cams as well.

I wish Fuji made a FF Leica rangefinder knockoff for $2500. How I love the Fuji sensor but hate the Fuji controls. The Leica is like a Rolex in build. But build does not produce photos. It is also very overpriced.

[Gigi]

Personally, I feel that large pixels combined with ultrasharp lenses without OLP-filtering is not an ideal combination. More like a recipe for an aliasing machine. I am not shooting Leica-S, but I see a lot of aliasing on my Hasselblad V and P45+ combo, not only on textiles but even on water surfaces, so I feel the problem is real. Stopping down to f/16 works as an OLP-filter and the problem disappears, but so does a lot of apparent sharpness.


From what I have seen, 3.8 micron pixel pitch seems a pretty good compromise, on the S2 that would be 94 MP. That resolution would make the lenses some justice. That pixel size is totally feasible as all 24 MP APS-C cameras have 3.8 micron pixels.

Erik -
I must say this image you posted is rather surprising. I've shot with a Leaf 7 for several years and never ever seen anything like this. I don't have the slightest idea what is causing this problem, but something seems really off. The lens and back you have are of good quality - anyone else have any ideas what's going on here, so we can help Erik?
Geoff

[eronald]

Moiré eez ze new film grain, eet is so artistique ...

Edmund

[ErikKaffehr]

Dear sir,

This is a textbook case of colour aliasing. The small waves have something like pixel dimensions. Demosaic cannot interpolate colour correctly. The image shown is of course a crop. Why you don't see it, I don't know. What I know is that it goes away at f/16, due to diffraction acting as an OLP-filter, but it is still present at f/11. Fine detail, high contrast, correct focus, medium aperture and no vibration are the recipes for this effect.

Best regards
Erik

Erik -
I must say this image you posted is rather surprising. I've shot with a Leaf 7 for several years and never ever seen anything like this. I don't have the slightest idea what is causing this problem, but something seems really off. The lens and back you have are of good quality - anyone else have any ideas what's going on here, so we can help Erik?
Geoff

[Paul2660]

Hi Erik

That is an interesting case of aliasing.  I shoot a lot of water and don't remember seeing this but I may have missed it. I wonder if using a CLPL filter might help as it will eliminate most ofthe glare and the glare may be part of the problem some exasterbating the effect?  I pretty much always use a CLPL when working water especially water in the sun.

Just a thought.

Paul

[Bart_van_der_Wolf]

This is a textbook case of colour aliasing. The small waves have something like pixel dimensions. Demosaic cannot interpolate colour correctly. The image shown is of course a crop. Why you don't see it, I don't know. What I know is that it goes away at f/16, due to diffraction acting as an OLP-filter, but it is still present at f/11. Fine detail, high contrast, correct focus, medium aperture and no vibration are the recipes for this effect.

Hi Erik,

The difference between f/11 and f/16, is a reduction of the MTF by at least 15% near the Nyquist frequency on a 6.8 micron pitch sensor. That's the Optical Transfer Function's effect from diffraction and a bit of lens/sensor aperture blur alone, which will then interact with the subject contrast and the sensor specific MTF (e.g. coverglass and fill-factor) to reduce some subject contrast to low significance, or even below a relevant detection threshold. The Raw converter will have much less ambiguous data to work on, even though some aliasing can still exist (f/22 would eliminate enough contrast to avoid aliasing altogether on that sensel-pitch).

The 10% loss of contrast in the MTF50 region will have enough remaining contrast to allow local contrast and detail boosts for compensation.

Cheers,
Bart

[torger]

I've had some rippled water moire, but you need relatively short shutter speed and the ripples need to be small enough to create moire. It's a very good example of how moire can occur in natural subjects.

With a tech cam at ISO50 and f/11 and center filter the shutter speeds are often long enough to smooth the waves to avoid moire.

[paratom]

Leica seems to always have sensor trouble. The MM has a very sharpsensor, but it has trouble with highlights. The M240 has a second rate sensor. No it is not sour grapes or jealousy. I had 5 Leicas, now 4. I have many other cams as well.

I wish Fuji made a FF Leica rangefinder knockoff for $2500. How I love the Fuji sensor but hate the Fuji controls. The Leica is like a Rolex in build. But build does not produce photos. It is also very overpriced.

Interesting how taste is different: I have allways prefered Leica S output over Leica M output and Leica M over Fuji output. For me the Fuji colors look like some old prints lying in the sun too long. A bit retro and nice but I prefer the deeper and more saturated colors.

[mbaginy]

Interesting how taste is different: I have allways prefered Leica S output over Leica M output and Leica M over Fuji output. For me the Fuji colors look like some old prints lying in the sun too long. A bit retro and nice but I prefer the deeper and more saturated colors.

Wow, I find your remarks remarkable.  I also prefer Leica's M colors to Fuji's but found Leica's less saturated than Fuji's.  I now use both X-Pro1 and X-T1 and often turn down saturation.  I often find Fuji's colors too velvia like (I disliked velvia).  Interesting, how our perceptions differ.

I also liked the colors Zeiss lenses on my Canon 5D gave me more appealing (cooler) than from Canon lenses.  A shame that manual focusing with the 5D is such a pain.

[lowep]

Ken, I have one.The Kodak sensor just doesn´t cut it any more, so does the handling.
In my daily work I MUST have the option to go beyond ISO 800 at least once a month.

If you already have one Leica S then you are rich enough to have at least two systems so why all the belly aching?

Just buy a Canon or Nikon DSLR for the days when you have to go beyond ISO 800 and keep the Leica S as a weekend shooter.

[EricWHiss]

Erik -
I must say this image you posted is rather surprising. I've shot with a Leaf 7 for several years and never ever seen anything like this. I don't have the slightest idea what is causing this problem, but something seems really off. The lens and back you have are of good quality - anyone else have any ideas what's going on here, so we can help Erik?
Geoff

Geoff,
I think you are onto something here.   It's quite likely this image is one he took specifically to illustrate color moire since that is of interest to him and something he has studied. That is of course much different than saying it shows up in every image, which I think you are are saying which I agree with. Color moire can occur but happens with very low frequency in my images and simple steps such as stopping down or moving the camera to subject distance can avoid it.  Surely he would know that?     

It may well be that finer pixel pitch will reduce the occurrence of aliasing as he argues, but it's not a free lunch either.   You'll see diffraction loses at wider and wider apertures as the pixel size gets smaller and DR can suffer as well.     It's one of the reasons why I am a proponent of larger sensors. 

Back on topic of the Leica… no I don't think the new Canon will cause them any trouble.  Their offerings don't compete on number of pixels anyhow.  That's silly. If all that mattered were just number of pixels then the Nokia n808 cell phone with 41mp would have wiped out all kinds of cameras.   

[ErikKaffehr]

Hi,

You are right about the thing that aliasing interests me. I am always interested in things were theory and practice are in contradiction.

Regarding the idea that the image is intentionally taken to illustrate moiré is a bit absurd. As you say, the phenomenon is not very frequent. How could I know it would occur? It cannot be observed in the viewfinder. It would be possible to spot it on the P45+ display, but how probable would that be? Also, it is not easy to find a shooting position in those coastal areas. In this case I am standing on the edge of a rock, it is something like three meter down to the water, the tripod just fits. If I move back I get a lot of unwanted stuff in the image.

It is simply one of my shoots september 2013. I have it in half a dozen of my exposures on that shoot, and I was not particularly happy when I observed it on my 24" screen.

It was probably shot at f/11 as it is aperture I use mostly on the Hasselblad. Exposure was 1/30 at 50 ISO. From my test chart shoots I know that colour aliasing disappears at f/16, but at f/16 there is a significant loss of sharpness. Some lenses would probably performing better at f/8 than at f/11, but according to my measurements there is little loss from f/8 to f/11 and f/11 gives a bit more of DoF.

Anders Torger, who used to shoot Leaf on a Linhof shoots mostly at f/16, just to avoid aliasing.

This is the full image with affected part marked:


And here is the raw file: http://echophoto.dnsalias.net/ekr/Articles/Shoots/Aliasing/20130920-CF044127.iiq


Regarding the effects of reduced pixel size you are mostly wrong. It is clear that diffraction effect will be more visible with smaller pixel sizes, but that it is only when you pixel peep. Enlargement is larger if you look at the image at actual pixels, but if you downscale a small pixel image to the same size as the large pixel image it will always be better. This is valid under the presumption that imager size is the same .

DR is on the other hand affected. Halving the pixel area, like going from 6.8 micron pixels to 4.8 micron pixels would give a decrease in DR by 1/2 EV. But, it would not have a significant effect on noise levels.

You are right about Canon not causing problem for Leica. I don't think Leica sweats a lot because of Phase One, Leaf or for that part the Pentax 645Z.

On the other hand, I don't understand why Leica develops lenses with what they say extreme sharpness and don't follow up with sensors to match.

Just as a side note, the OP (the guy who started the thread) owns a Leica S2, so he speaks from personal experience.

Best regards
Erik

Geoff,
I think you are onto something here.   It's quite likely this image is one he took specifically to illustrate color moire since that is of interest to him and something he has studied. That is of course much different than saying it shows up in every image, which I think you are are saying which I agree with. Color moire can occur but happens with very low frequency in my images and simple steps such as stopping down or moving the camera to subject distance can avoid it.  Surely he would know that?      

It may well be that finer pixel pitch will reduce the occurrence of aliasing as he argues, but it's not a free lunch either.   You'll see diffraction loses at wider and wider apertures as the pixel size gets smaller and DR can suffer as well.     It's one of the reasons why I am a proponent of larger sensors.  

Back on topic of the Leica… no I don't think the new Canon will cause them any trouble.  Their offerings don't compete on number of pixels anyhow.  That's silly. If all that mattered were just number of pixels then the Nokia n808 cell phone with 41mp would have wiped out all kinds of cameras.  

[JV]

Just as a side note, the OP (the guy who started the thread) owns a Leica S2, so he speaks from personal experience.

No offense but complaining about the ISO performance of a CCD sensor and the price of Leica is pretty close to trolling in my book....

[ErikKaffehr]

Hi Paul,

I don't see this often. It is a combination of small ripples on the water, relatively high contrast and so on. The idea of using a CLPL is interesting. I have a couple of CLPLs but neither fit on the blad, as they have a bayonet. I have an adapter so I could fit my existing filters but than I would not be able to use the lens shade. I don't think Hasselblad still makes filters for the V-series. Food for thought…

As Anders Torger has pointed out in another posting, he has seen it a couple of times, but he normally uses a center filter and that lengthens exposure. A polariser would also do that. Personally, I seldom use polarisers but NDs I use quite often and I have the same problems with NDs.

Getting back to the original topic: When Leica released the Leica S2 there was considerable enthusiasm about the new lenses that were designed to support much higher resolution. When Leica moved to next generation I was assuming they would increase resolution. IMHO it makes little sense to build a very high resolution lens system and combine with a low resolution sensor. I presumed they would use the Sony sensor or that they would scale down the pixels on their CMOSIS designed sensor.

That sensor has been designed for Leica M wide angles, with shallow wells. I guess that they may have given up some full well capacity to keep the wells shallow. Comparing the Canon 5DIII noise levels in DxO mark, it seem that the Canon has slightly less noise (shot noise) at most ISOs and goes lower in ISO and has significant  advantage at base ISO. This would be what I would have expected from reduced well capacity. But I will not pretend I know about CMOS sensor design. Regarding DR, the CMOSIS sensor runs circles around the Canon.

The CMOSIS sensor would be a good alternative on technical cameras and Leica owns Sinar, so I would have expected the CMOSIS designed sensor to show up in digital backs for Sinar, but has not been the case this far.

I would presume that Leica has done their market research and arrived at the conclusion that they have chosen a feasible strategy within the constraints they probably have.

Best regards
Erik

Hi Erik

That is an interesting case of aliasing.  I shoot a lot of water and don't remember seeing this but I may have missed it. I wonder if using a CLPL filter might help as it will eliminate most ofthe glare and the glare may be part of the problem some exasterbating the effect?  I pretty much always use a CLPL when working water especially water in the sun.

Just a thought.

Paul

[tsjanik]

Erik:

I really like the image, moiré notwithstanding (I'd be tempted to remove the contrail however).  I would love to be there with a canoe and a tent.  I assume this is in Sweden; should you ever have the chance to visit Georgian Bay in Canada, you will feel very much at home.

Tom

[ErikKaffehr]

Thanks for the suggestion! I sort of liked that image but got frustrated with the moiré stuff and sort of ignored it.

If you ever come to Sweden, the place is Stendörren a natural preserve near Nyköping. Canoes are OK but tents are not allowed within the preserve. Sweden is a good place for outdoors.

That place is just 5 minutes from office :-).

The coordinates are here: 58°44'54" N 17°22'38" E, taken from an other shot. The real shooting position is marked.

Best regards
Erik

Erik:

I really like the image, moiré notwithstanding (I'd be tempted to remove the contrail however).  I would love to be there with a canoe and a tent.  I assume this is in Sweden; should you ever have the chance to visit Georgian Bay in Canada, you will feel very much at home.

Tom

[EricWHiss]

It is clear that diffraction effect will be more visible with smaller pixel sizes, but that it is only when you pixel peep. Enlargement is larger if you look at the image at actual pixels, but if you downscale a small pixel image to the same size as the large pixel image it will always be better. This is valid under the presumption that imager size is the same .

I'm not sure one can get away with such a presumption that with more pixels the final image size would stay the same with more and smaller pixels.  Rather, with more pixels people will crop more, print bigger, or just zoom into 100% and pixel peep.  I don't think Canon can sell the new 5Ds with more and smaller pixels and tell their customers that they can only expect better quality if they keep the prints the same size as with their 5Dmk2.

You yourself claimed that you'd be limited by diffraction in the posted image if you stopped down to f/16 to avoid aliasing.   I doubt with a p45 its enough to spoil an image more so than aliasing since I shoot my 80mp camera frequently at f/16 and do fine, but in any case you've just proven my point about smaller pixels.  If you were not willing to stop down to f/16 due to diffraction concerns in the above image and a p45, then you would really wouldn't be happy with a similar camera with your own optimum pixel pitch where you would be limited to a much larger aperture, than you are now.