Magic of the M9 KAF-18500 Sensor

[jaapv]

Let's go back on thread subject, I.e., assigning some "analogue magic" to the M9 CCD sensor when compared to more modern CMOS sensors (A7 and M240), and comments on the "better" dark tones and color rendering of the M9.

I'm sure the OP really observed these differences but I'm also convinced that there are reasons much more prominent than "magic" sensor differences. For instance:

- different IR response (color rendering)

- wrong exposure. Meaning the OP might not have used the whole capability of his CMOS sensors. 2 more bits DR correspond to a 4 times wider range of dark tones, This has nothing to do with only being able to print 9 bits or less, it has _all_ to do with tonal range.

No marketing hype for digital latest and greatest from me. Some professional semiconductor engineering baggage though.

Roland.

Bits? where did those come from? Yes- it has to do with the tonal range. What tonal range are you able to print? Overkill in the camera can only cause you to either pick a range from the original file or to compress the tonal range to get it on paper - losing the "advantage" you thought you had.

Talking about quacking, the sensor appears to have less impact than the soft- and hardware that follows it up. The Nikons quack a lot better than Sony's own DSLRs, despite sharing sensors, and I spend far less time getting a good result out of the X1 (Sony sensor) than I do out of the NEX 7. And matching the NEX7 images to the results of the M240 is near-impossible in my hands. I know - apples and pears, but I shoot them side by side. I guess Leica knew a thing or two by preferring CMOSIS over Sony (whom they contacted first)

 

[jaapv]

Today my wife took a red velvet cake (deep purple really) out of the oven with a red glove. I'll post a photo tomorrow, taken with the 240. On the photo, the glove is bright red, but the cake is light brown ... Note that I don't mind really. It's just part of the learning curve.

Thanks,

Roland.

Yes - that is logical. The camera responds to IR. The IR filter on the M sensor blocks about 70% (80% on the M9) You are seeing the remaining IR as a colour shift. Slap on an 486 filter and the colour will be correct.

 

[bigeye]

How much work in post to get what you want?

How much work in post to get what you want?

I'm no expert, but I've gotten far enough to find that there is too much 'stuff happening' between the sensor and image file to make useful generalizations about any particular component. Each camera manufacturer manipulates the raw and the jpg output more or less and it either works for you or doesn't.

B&Ws made from the GR can be sublime, but I often wince at its rendering of color.

Three Nikons (D90/D700/D800e) went out on a bright Fall day with me and I have probably spent a day manipulating identically shot raw images from each in post. For that shooting condition, I was surprised at how acceptable the D90 image could be made, how little work was necessary for the D700 image, and how much work it takes me to match the D700 output with the D800e. That's from the same sensor and camera makers.

How long and hard it is to get the image I want in post is what matters. (DxOmark should add Photoshop times for a panel of photographers to their camera tests. 😱)

 

[YYV_146]

The whole DR range argument is the same marketing blah as the megapixel argument. Film had about 13 stops, the M has about 13 stops, and the Sony indeed considerably more. However, you need to be pretty good at printing to get nine stops onto paper, so the only advantage of those high dynamic ranges is that one can afford to expose sloppily....

Not sloppily, but latitude to control shadows and light. Exposing for the highlights isn't sloppiness, it's best practice for digital sensor. You can get images closer to what the human eye sees with greater consistency with a high-DR sensor.

 

[ferider]

Bits? where did those come from?

Just the correct unit of DR of a digital processing system, Jaap. In layman terms, for cameras, continue to call it "EV stops", if you prefer.

Here is the dark purple red velvet cake in case anybody cares:

On the LUF are other examples that show that the 240 is more sensitive to IR than the M9. And I assume the OP's A7r was too.

... there is too much 'stuff happening' between the sensor and image file to make useful generalizations about any particular component. Each camera manufacturer manipulates the raw and the jpg output more or less and it either works for you or doesn't.
: : :
How long and hard it is to get the image I want in post is what matters.

Exactly.

Carry on 🙂

 

[Corran]

I really need to sit down and test the color response of the M9, especially compared to my other Nikon DSLRs.

Initially, after post-processing, I love the color of the M9. But what gets me is the hoops I have to go through to get there. Number 1 on that list is white balance. The M9, set on Daylight, looks atrocious to me. It's like a bad color negative film poorly scanned. I have to set the temperature to like 6500 or more in post to get a "normal" white balance. On a normal day I set it to cloudy, on an overcast day it's set to shade, to get some semblance of proper color. I do shoot/edit RAW 100% of the time.

I haven't seen (and nor have I really looked) at a comprehensive test for color but I see lots of people, as above, say they were "blown away" by the color response of the M9. I don't know what that means. Blown away how? In what respect? Compared to what?

Now my Nikon D700 colors were pretty good, if stale and boring (like most digital, to me). The D800E I have is wacky with a magenta tinge inside w/flash and a yellow tinge outside. I have to do all kinds of gymnastics in Photoshop to get really good color, like with the calibration trims on the bottom of the settings. The Adobe Standard setting doesn't work at all for the D800E. And even at my best, with lots of post-processing, I still seem to have a slight yellow/orange, waxy look to skin tones with the D800.

 

[Huss]

Corran, the best colour response I have received from my M-E is by using Auto WB, and correcting as needed.

Of course, that is if I have not used a WB card and set it manually before shooting. It takes just a moment.

 

[Corran]

AWB for me makes it even more blue (i.e. it "auto" corrects it the wrong way).

I'd love to see more images posted here in this thread from users showcasing what they consider the "magic" of this sensor, out of curiosity, with specific color setting information (WB, trims, ISO, etc.)

 

[Huss]

AWB for me makes it even more blue (i.e. it "auto" corrects it the wrong way).

Interesting that our cameras react differently.

 

[jazzwave]

Interesting that our cameras react differently.

....or our eyes/brain see/interpret differently 😀

 

[Pioneer]

If I am shooting RAW I really don't worry about the white balance so I leave it on Auto. If I am intending to shoot jpg, I set the white balance with my little card. Quick and easy.

As for trying to illustrate color differences on screen I learned a long time ago that the monitor you are using to view the examples will have an over riding effect on any examples. Not to mention the lighting around your work station, etc, etc.

The print is actually the best way to do this and even that can be difficult unless we both view in the same location, under the same lighting.

Not to mention, of course, that your perception of color and mine can be different, sometimes dramatically different. This can be quickly appreciated by reading different people's responses to various films, like Ektar, Portra, Velvia, Ektachrome, and so on.

As for being "blown away", you are right to take me to task for that as it is entirely subjective and is based on what I may have been used to before. But the term does fairly successfully relate that I was very pleasantly surprised with the results. I can get results that I like very much, far more quickly with the M9 than any other digital camera I use, with the possible exception of my Pentax *ist DL and K10D. I sold a very nice Canon set that included some terrific L glass and an EOS 1Ds MkII, as I far preferred the results from the M9 and a pre-Asph Summilux. And the Leica kit was far more portable.

 

[willie_901]

In my view the enjoyment and appreciation of the color rendition in digital M bodies is an authentic characteristic of the total design of these cameras. By total design I mean the optical components in front of the sensor, the analog ISO amplification technology, the properties of the analog-to-digital converter and the proprietary firmware used to produce the in-camera raw files. Leica clearly does these things well and Leica owners rightfully consider the final raw data to be a value-added feature of these cameras.

However the CCD can not play a significant role. There is no magic involved. There is only physics and engineering.

The OP wrote, "But basically I’m beginning to understand that CCD is analog and CMOS is digtial, as camera sensors go. "

This statement is a fundamental misunderstanding. While both CMOS and CCD sensors use digital timing to control their output, both technologies generate analog signals. The only difference is the CCD technology requires separate chips to complete the data stream output while CMOS technologies integrate all of the data stream electronics within the main sensor chip.

Both CCD and CMOS employ pinned photodiodes to utilize the photoelectric effect. The output of both CCD and CMOS photo sites are analog DC voltages, which are the signals. DC voltages alone can not have any impact on color rendition whatsoever. Instead they are just an estimate of the light energy a particular pixel collects (and the lower the signal-to-noise ratio, the less certain the estimate becomes). Other factors such as the transmission properties of the color filter array, the IR filter transmission properties and even the frequency response of the PIN diodes themselves could affect the DC voltage amplitudes and could impact Bayer image reconstruction. But these variables will have the same impact for both CCD vs CMOS.

If you think I am misrepresenting the CCD vs CMOS story please take the time to read these links.

This is a fairly non-technical general comparison of CCD and CMOS sensor technologies.

Here is a detailed technical review article from a peer-review journal.

link 1

About midway through this presentation the same author shows data comparing the quantum efficiencies of CMOS vs CCD sensors. QE determines the sensors sensitivity to light. It affects the signal component of the signal-to-noise ratio.


This presentation summarizes the history of image sensor development and explains in detail the physics and engineering behind the superior signal performance (which is not color performance) of CMOS devices.

These links explain why it is extremely unlikely Leica will ever make another camera with a CCD sensor.

Because the signal-to-noise ratios of the signals recorded by the sensor directly affects the data quality, SNR is far from a deceptive a marketing ploy. It turns out a sensor's analog dynamic range is directly dependent on the SNR. DR is not a deceptive marketing ploy either.

Leica photographers make magical images with their digital M bodies. Leica's digital M8/9 designs may empower their owners to produce magical results. For all I know Leica m8/M9 cameras may have the best color rendition in the history of digital still photography. But the facts tell us CCD pinned diode sensor technology alone can not be responsible for the magical images Leica M8/M9 owners create.

 

[Corran]

I didn't mean to imply that I was "taking you to task" Pioneer, I just wasn't sure what that meant.

jazzwave is correct, our perceptions differ. I find it interesting, Pioneer, that you sold a Canon kit after using the M9. In my perception, anyway, I do in fact find Canon colors awful and would not doubt the M9 "blows it away!"

Like I said I need to look at this and test closer. Frankly I usually am shooting film, not digital, so I haven't been as methodical as I could have been, despite owning the M9 for a while now.

Regardless, there certainly is some "magic" in the skintones, when it all comes together. A pretty girl helps too 😉

Leica M9, Nikon 85mm f/2 S-mount lens + Amedeo adapter, Nikon SB-800 flash

 

[Monochrom]

well now i´m happy because my m9 is an analogue digital camera 😀

 

[jaapv]

On the LUF are other examples that show that the 240 is more sensitive to IR than the M9. And I assume the OP’s A7r was too.

M8 50%, M9 80%, M240 70% of IR suppression. I use filters on the 240 quite often. Last night I took dinner shots, candlelit, and all colours were fine, except the mint coloured synthetic curtains, they were beige 🙁 No IR filter...)

 

[uhoh7]

well now i´m happy because my m9 is an analogue digital camera 😀

LOL, I thought you would like that. 🙂

Two words have got in into a lot of trouble in this thread: "magical" and "analogue". But I'm sure I deserve a hard time.

really a great discussion going, with only minor sniping, excellent content, and real discussion.

Willie, you have made a fantastic post up there, and it will take me some time to really get a basic grasp of your points. Your response was just the kind of technical argument I hoped to draw with my post, thank you.

While no doubt there are layers of technology, hard and soft which make the final DNG output of both the M9 and 240, and each layer will have an effect, I don't think anyone can really expect two sensors which are completely different in every respect are going to make the same output, no matter how hard the engineers try. Tri-X and Velvia are more similar than the M9 and M240 sensors.

@Victor. As I'm shooting the A7 RAW, there is no high ISO NR. Funny thing is the A7 always jumps up to higher ISO than M9...but I do have the M9 limited to 800.

Now quite a bit of my mythical 'magic' could be chalked up to WB. Auto WB on the M9, which I always use, is often excellent, but highly variable with light source and orientation. Shooting raw with both machines I feel I can correct, but in real life I only do this in 15% of shots.

What comes across with the M9 is clarity and richness of color, while auto WB on the A7 is often muddy by comparison. Untouched M9 shots are crazy clean. Colors are very distinct.

Only with major PP stopped down do the M240 and M9 shots start to look more similar. Even then lenses like the 28 cron and 50lux do not make the same results, and I think this is widely confirmed, again, also by people who prefer 240.

Frankly I think this is why the prices have fallen so much on these two lenses.

@Roland I'm very much enjoying your posts, but I do get the idea you are representing the large faction of RFF users who just feel digital sucks LOL

But you are making some pretty good technical points nonetheless, esp IR sensitivity, and of course exposure.

I shoot both cameras at least 2/3 stop dark AE, which I suppose is sacrilege, esp with the M9, but most of the time it works really well. The Sony can recover blown highlights RAW to a surprising degree, but the M9 is very sensitive that way, and you must be really careful or you can loose a whole shot. - 2/3 is pretty safe. By then the blacks are black but they push great from uncompressed DNG, way better than sony, which builds heavy noise with a push in LR.

But I can see why many love film over digital, with the great gamuts, and very forgiving exposure behaviors, and many respects. I may start shooting more myself if I can figure a decent workflow. Obviously film is the true "analogue".

Anyway, TY to everyone so far for an interesting and very civil thread 🙂

 

[LCT]

Magic of nostalgy 😉

 

[Scheelings]

The whole DR range argument is the same marketing blah as the megapixel argument. Film had about 13 stops, the M has about 13 stops, and the Sony indeed considerably more. However, you need to be pretty good at printing to get nine stops onto paper, so the only advantage of those high dynamic ranges is that one can afford to expose sloppily....

At what ISO? I thought that the real advantage is to have those 9 stops at a higher ISO. Does the M9 have 9 stops at 1600 or even 800?

 

[willie_901]

...
Now quite a bit of my mythical 'magic' could be chalked up to WB. Auto WB on the M9, which I always use, is often excellent, but highly variable with light source and orientation. Shooting raw with both machines I feel I can correct, but in real life I only do this in 15% of shots.

What comes across with the M9 is clarity and richness of color, while auto WB on the A7 is often muddy by comparison. Untouched M9 shots are crazy clean. Colors are very distinct.

...

Does this mean you primarily use out-of-camera JPEGS instead of DNG?

 

[kbg32]

I had the M9 and the M8 together for a short period. For pure "sparkle", I found more magic in the output of the M8's KAF-10500 than in the M9--in color as well as B&W.

I found differences between the two as well. The M8 sensor seemed to have a certain "sharpness", "pop", then the M9. The M9 just seemed a little bit more "laid back". I kept my M8 as a backup. I use my M9 almost exclusively, mostly for the full frame and the more discreet shutter.