The New Leica S Adoption-Rates Question

[Theodoros]

I haven't used Leica S or S2 yet, but I am quite experienced with a variety of sensors used in various MFDBs were I believe the S2 is based on… The general conclusion I have ended up, is that both Kodak and Dalsa sensors are far superior than sensors used in DSLRs for colour accuracy, but quite different between them! I find Dalsa sensors to be more colour accurate (especially the "old" 33mp 36x48mm one), with files that "out of the box" seem to have a "duller" presentation, but with great latitude to achieve the "bite" anyone desires, while Kodak, seem to be "punchier", more "eye-catching - out of the box", with more saturation and contrast, perhaps favouring blues and greens towards their "deeper" (darker) tonality while reds may have a bit more balance towards magenta than they should (obviously to "help" skin tones to appear more pleasing)… I've also find, that both Dalsa and Kodak can be treated (one by adding "punch", the other by subtracting some) to be very close, but Kodak may need more "work" in trying to do so (especially with red accuracy which I find superb with Dalsa)… Now, Leica is based on a Kodak sensor …no? Hence I expect it to yield towards the "family character"… somehow however, I have that feeling that given Leica's tradition for "neutral lenses" in terms of colour balance, they should have preferred a Dalsa sensor for the camera.

[Theodoros]

I missed this post - sorry - I didn't mean to ignore you.

To answer you: no, having done the swap from 135, 6x6 and 6x7 so often it brought no visual problems along with the change in shape; if anything, for a brief period, the swap brings a sense of visual rebirth, which is quite good for one's photography.

I think that what both you and synn are actually referring to is manual familiarity with individual cameras, rather than formats. Are you saying that if you owned both a 35mm and a 6x6 camera long enough to know them both, you couldn't change around at will without running into a visual/conceptual problem?

Rob C

No, what I'm saying is that there are people that do need time to adopt in different aspect ratio, while if one makes the jump from DSLR to Leica S, ...that, can't be the case.  I thought it's obvious in the first post of mine that you quoted on… never understood where did you see me talking on format adaptation.

[eronald]

It would be nice to know whether the base ISO is the same and the sensor was simply rated at base 160 on the first model S2 to protect the highlights more and encourage handheld use, or whether there is a real difference in the way the sensor data is processed.

BTW, color and texture from the original M8 was very good, but with a tendency to magenta on white skins, so I see no reason to doubt the quality of the S files per se, but it would be nice to have some Raw images to process. Surely a company which invests heavily in bespoke cabinets for dealers can afford to donate a few Raw files?

It would also be interesting to see some cotton/wool black next to some synthetic black - as we all learnt, Leica has had some IR sensitivity issues.

Edmund

Richard,

We haven't done any exhaustive side-by-side testing, but having worked extensively with S2 files since a year before the camera was introduced, I did see a difference in the overall look, feel and malleability of the S (Typ 006) files when editing at 1:1. Pixel for pixel comparisons between cameras are notoriously difficult to accomplish. It is near impossible to achieve the exact same camera placement, subject lighting (for naturally lit subjects), exposure and focal distance when shooting a series of images between two cameras, so we often have shied away from this type of evaluation.

[peterv]

Hi Edmund, there are some raw files here:

http://www.s.leica-camera.com/en/S-RAW

I use the S2 and the files come out nice and neutral, not desaturated at all.

[Theodoros]

It would be nice to know whether the base ISO is the same and the sensor was simply rated at base 160 on the first model S2 to protect the highlights more and encourage handheld use, or whether there is a real difference in the way the sensor data is processed.

BTW, color and texture from the original M8 was very good, but with a tendency to magenta on white skins, so I see no reason to doubt the quality of the S files per se, but it would be nice to have some Raw images to process. Surely a company which invests heavily in bespoke cabinets for dealers can afford to donate a few Raw files?

It would also be interesting to see some cotton/wool black next to some synthetic black - as we all learnt, Leica has had some IR sensitivity issues.

Edmund

It is normal for CCD sensors to underexpose them a bit as to protect HLs more and seek max dynamic range in shadows (where there is lots of latitude to "dig" for…), Sinar did the same (what you suggest) back in 2006 by rating (via firmware update) all their E-motion backs a stop higher. My opinion is, that MF isn't exactly aimed for the "fun-boys", but rather to the experienced users that have full control on what they are doing… I was already under exposing the Sinarback E-motion-22 I was using at the days, never needed Sinar to tell me that it was the correct thing to do… Only updated it, when time came to replace it for my current back… which of course (again) I never use in other than manual mode or trust the metering of my camera to judge exposure.

P.S. (edit): Personally, I find no CCD back rated "correctly" for sensitivity… They should all (IMO) be rated higher than they are, although by different amount more or less… not that it makes any difference to those that trust their knowledge and experience above anything else.

[KevinA]

No experience myself, but I did have a chat with someone who borrowed one from Leica. Dust was the problem they encountered and found it impossible to keep clean.
Do those of you that have find this to be a big problem?

[eronald]

deleted

[Telecaster]

Personally, I find no CCD back rated "correctly" for sensitivity… They should all (IMO) be rated higher than they are, although by different amount more or less… not that it makes any difference to those that trust their knowledge and experience above anything else.

I agree...I could care less what the stated values are. Just use the thing & figure out for yourself what its limits are. Then evaluate with a calibrated external meter if you must (I'll pass, thanks.)

Re. "accurate" color...I don't even know what that is. Pleasing color catches my eye. So does unexpected color. Photos that attempt to reproduce reality don't do much for me. Photos that reveal someone's unique take on reality...those I'm interested in.

-Dave-

[Theodoros]

I agree...I could care less what the stated values are. Just use the thing & figure out for yourself what its limits are. Then evaluate with a calibrated external meter if you must (I'll pass, thanks.)

Re. "accurate" color...I don't even know what that is. Pleasing color catches my eye. So does unexpected color. Photos that attempt to reproduce reality don't do much for me. Photos that reveal someone's unique take on reality...those I'm interested in.

-Dave-

About Iso rating… I think the makers rate the MF sensors having the use of strobes in mind… About "colour accuracy"... I think that they "tune" the MFDBs having fashion photography in mind, where dresses have to be both "colour accurate" plus as appealing as possible, while dress/model skin combination has to be as attractive as possible… (I believe the later explains Kodak's sensor tuning).

[eronald]

I like cameras with "usable, flexible color" which give me files I can then retouch into the color I want. Which means a camera which sees subtle gradations of color, and makes files which don't "break" when retouched.

And then there are images with color I really really don't like, however "unique" the colors are. An example of such a horrible-color unique-color image is doubtless the first one on this page

http://www.s.leica-camera.com/en/S-RAW

which reminds me of "The Scream".

Edmund

[/quote author=Telecaster link=topic=86673.msg706219#msg706219 Re. "accurate" color...I don't even know what that is. Pleasing color catches my eye. So does unexpected color. Photos that attempt to reproduce reality don't do much for me. Photos that reveal someone's unique take on reality...those I'm interested in.

-Dave-
[/quote]

[Theodoros]

Sometimes I wonder what your monitor is and how do you calibrate it…. 

[eronald]

Sometimes I wonder what your monitor is and how do you calibrate it….  

I can't remember how I calibrated my Macbook Pro
But my on-screen colorchecker (colorfilled) images match my printed Colorchecker, and the Xrite colorchecker which I measured (i1Pro2 spectro) to create the image .
Is that what you wanted to know?
I'll be glad to run a print profile of anything including transparency media for you for free so you can review the quality of my profiles

However, now you say it, I will check the image on something else to verify that my monitor isn't doing something strange.

Edmund

[Theodoros]

I can't remember how I calibrated my Macbook Pro
But my on-screen colorchecker (colorfilled) images match my printed Colorchecker, and the Xrite colorchecker which I measured (i1Pro2 spectro) to create the image .
Is that what you wanted to know?
I'll be glad to run a print profile of anything including transparency media for you for free so you can review the quality of my profiles

However, now you say it, I will check the image on something else to verify that my monitor isn't doing something strange.

Edmund

Thanks for the offer Edmund, but since I do painting reproduction for a living, I'll pass the free offer… What would interest me however, is if you can remember how you've calibrated your Macbook pro… Surely I1,X-rite can't be successful on calibrating a Macbook's pro screen. It just surprised me what is that you found "awful" in Leica's promote image..   
I also use a Macbook pro and have managed to calibrate it to an extend that "matches" surprisingly well (some call it "perfect" match - to my scale it's 98%) my Eizo CG241w monitor which I use to print on my Epson 9900… but haven't done that by using my I1 x-rite (which I also own), so I really am interested if you may share your method of calibrating it. 

[rgmoore]

T.Dascalos,

May I ask you which MF system (lens, camera and back) that you have found useful in your work in photographing paintings.  With my D800 and 5DII in the past I have had inconsistent results even with Xrite
Colorchecker.  I am seriously considering moving up to a MFDB for the sake of improved color accuracy in photographing my own oil paintings and portrait work in the studio.

Any thoughts and suggestions would be much appreciated.

Richard

[eronald]

It just surprised me what is that you found "awful" in Leica's promote image.

Dear Mr. Dascalos,

 Here is a composite with the original preview from Leica, and a slightly better color balanced version. I don't like my over-red version too much, it looks like an old Playboy, but at least it doesn't look like the greenish Leica preview which turns the model into a fashionable cross between a vampire and an alien. Which is by the way what you get if you render "as shot".  

 I ran this image past another color consultant to the industry, as you indicated my perception was bad, and we ran the image through a couple of converters and agreed about the color issues. The causes are interesting to us, but I don't think they are relevant to you as your taste differs from our perception of color.

Edmund

[Rob C]

Dear Mr. Dascalos,

 Here is a composite with the original preview from Leica, and a slightly better color balanced version. I don't like my over-red version too much, it looks like an old Playboy, but at least it doesn't look like the greenish Leica preview which turns the model into a fashionable cross between a vampire and an alien. Which is by the way what you get if you render "as shot".  

 I ran this image past another color consultant to the industry, as you indicated my perception was bad, and we ran the image through a couple of converters and agreed about the color issues. The causes are interesting to us, but I don't think they are relevant to you as your taste differs from our perception of color.

Edmund

Well, Edmund, the wood certainly looks better.

It's a product of thin women: the light glances off at a sharper angle. Combine that with the spectral diffusion quality of tobacco and there you have it: a flawed test subject for larger Leica.

;-)

Rob C

[torger]

The Leica sample has indeed an obvious green cast, you don't need a calibrated monitor to see that. Casts are generally evident even on uncalibrated monitors. Even if color blind you can with a color picker check values on various near-overexposed areas which indeed are green.

I was assuming that was a look preferred by the photographer though, but if you want "wrong" color balance I think a more yellowish tone is much more pleasing though for most subjects, especially people.

Camera manufacturers that present mediocre sample images on their web is a well-established tradition though, mastered by the Japanese. And Leica makes a quite good job on that too it seems .

[eronald]

Well, Edmund, the wood certainly looks better.

It's a product of thin women: the light glances off at a sharper angle. Combine that with the spectral diffusion quality of tobacco and there you have it: a flawed test subject for larger Leica.

;-)

Rob C

Rob,

As always I appreciate your rigorous scientific explanations  

Edmund



Torger,

In my opinion, the camera is seeing red, assuming the world is grey, and therefore adopting a white balance that induces this greenish cast.

Iliah says that this image contains fluorescent lighting while being shot at 1/90, which I guess would also help make it look good.

Now if I could only understand why a photographer would want to make a woman from Venus look like a green man from Mars, I would be more in tune with the way the world works.

Edmund

[Theodoros]

T.Dascalos,

May I ask you which MF system (lens, camera and back) that you have found useful in your work in photographing paintings.  With my D800 and 5DII in the past I have had inconsistent results even with Xrite
Colorchecker.  I am seriously considering moving up to a MFDB for the sake of improved color accuracy in photographing my own oil paintings and portrait work in the studio.

Any thoughts and suggestions would be much appreciated.

Richard

I use Contax645 and Imacon 528c MFDB in 16x multishot (microstep) mode only, for painting reproduction, the lens I use is the Zeiss 120mm f4 Apo, but sometimes, when I have to reproduce a byzantine wall painting in an old monastery, I have to go wider than that… I also use Fuji GX680 (with the same back - have Contax adapter on the Fuji) for whenever I need movements. MO is, that even the best automatic calibration is good for photography, but really poor for painting reproduction, you really have to build your own profiles to be able to reproduce accurately a painting and you have to avoid interpolated colour, you have to "start" with "true colour" capture (although this is only 10% of the path)… aside the detail of course, where in large paintings, no single-shot camera or back can deliver.
Another thing to consider, is that if you have to work outside the studio, you really have to calibrate a laptop to match as much as possible your printing monitor/printer calibration… and that's not easy at all, It can only be done "manually" and needs lots of experience on how to "tweak" an ICC profile to compensate for the inaccuracies with laptop screen technology and (very important) to make it stable... Then there is lighting, control of reflections, tripod choice (very important), polarisation of lighting & lens, having even lighting temperature on the painting and tenths of other small factors that can ruin the result, for which one can wright a book about…. In other words, it's different to take pictures of paintings, than to reproduce a painting…

[RVB]

T.Dascalos,

May I ask you which MF system (lens, camera and back) that you have found useful in your work in photographing paintings.  With my D800 and 5DII in the past I have had inconsistent results even with Xrite
Colorchecker.  I am seriously considering moving up to a MFDB for the sake of improved color accuracy in photographing my own oil paintings and portrait work in the studio.

Any thoughts and suggestions would be much appreciated.

Richard

Hasselblad multishot excels at this type of work...