400 Leica photographers agree: we love CCD!

[haroldp]

Advertisement (gone after registration)

Disclosure: I support THE petition because I feel that everybody has a right to send a request to Leica. Leica can then decide whether it makes sense or not. However there is one myth that has to be debunked. Sensor technology has absolutely nothing to do with colour rendering as a sensor is a monochrome device. The colour rendering is imparted by the formulation of the colour filters in the Bayer filter. If the 240 would have had an ON CMOS the colour would have been exactly the same as the M9. If a hypothetical new CCD would be made by CMOSIS, the colour would be like the 240...

 

So you are basically asking the wrong thing.

Now if you were to ask for an M9 type Bayer filter on a future CMOS M we would be getting somewhere.

 

Thank You .... H

[Paul J]

I'm not so au fait with the finer mechanics behind the sensors, nor do I really want/care to delve into them, however I do know I don't like the M240 over the M9, and I don't like the Phase One IQ250 over my P65+ which shares it's 6 year old Dalsa CCD sensor with the IQ260, cut from the same cloth as the IQ280. My colleagues feel the same. CCD at this level, at base ISO, is astonishing. It's an IQ monster and it's ridiculous what you can do with it.

 

If I was to shoot a wedding I might want CMOS, with better high iso, but I rarely shoot over base ISO on both my Leica and Phase. If I wanted to shoot motion, I would be shooting CMOS. I do nether of these.

 

I think the main reason we are now seeing a proliferation of CMOS is because of convergent disciplines, of motion and photography which I have no interest in. The double revenue stream potential for camera companies is too much to resist and I don't like our chances of Leica going back or remaining with CCD for long. In my eyes, this is a shame.

 

The best IQ I see comes from CCD sensors, wether thats a property of CCD or not, I'm not 100% sure, until I see evidence otherwise I believe it to be. I use what tools work best for me and I go for best IQ. I believe a lot more can be squeezed out of CCD. I certainly won't be changing over with the current tools, which to my, and my peers eyes don't compare with the high end CCD offerings.

Edited January 29, 2015 by Paul J

[Advertisement]

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

[NightSun]

 

From posts here it is clear that workflows do change and yet there are those who argue that there is no inherent difference between the final outputs from either type of sensor. The REAL question which I would like answered is that if this is indeed the case, why does the output workflow have to change?

 

Because there are indeed inherent differences among the raw image outputs from different sensors (same scene, same lens assumed). These differences could be (although not exclusively) traced back to the following: the radiometric response curve of the sensor, that along with noise performance determines dynamic range and quality, subtleness of tone transitions at different exposure levels. Then there are the optical effects of the stack of IR and AA filters that influences micro-contrasts and presence/lack of aliasing artifacts. And you also have the Bayer RGB filters in front of the pixels. These can be also different and actually some camera manufacturers tend to use Bayer-filters with a lower absorption, thus resulting in better signal-to-noise performance, but at the same time trading off against color reproduction. Also, depending on the actual spectral transmission characteristics of Bayer-filters, color discrimination can be widely different. And, finally, camera manufacturers many times implement some sort of noise reduction on the chip, so that even the so-called raw data in your image file is already somewhat manipulated and not really raw as it would be coming off directly from the AD converter. And noise reduction also comes with trade-offs, usually with a degree of micro-contrast loss.

 

So, based on these main factors, you could have a quite different raw image with the same lens on a scene compared to another camera sensor output. With post-processing these differences can be mitigated, but a perfect match is not strictly possible. E.g. if one sensor has a much more compressed radiometric curve at dark exposure levels than the other, then you will never be able to get so subtle tonal transitions in shadowy image regions with the first compared to the second, no matter what you do in post-processing.

[Lenshacker]

The CMOS vs CCD debate has become much like the film vs digital debate. Each is a case of one technology being different from the other. With the "right amount of processing", people can make a digital file look like film. With the right amount of processing, you can make an M240 file look like an M9. A lot of people would rather spend time shooting with the technology that gives them results that they like without spending a lot of time processing it. Otherwise- we would never have a market for the M Monochrom.

 

The difference between CMOS and CCD sensors is "in the noise".

That's really funny to an optical engineer.

[KenFaulkner]

Its a shame that some of you don't like change

You can't hold onto old technology for ever or we all would still be using Box Brownies.

 

You need to take the new technology and adapt, that is what some don't want to do.

 

Its not of a case well we have always done and like it this way.

 

A Company will not stand still for a few hundred hard core old school

 

Sorry if you don't agree but this is life

 

Regards

[Prosophos]

Its a shame that some of you don't like change

You can't hold onto old technology for ever or we all would still be using Box Brownies.

 

You need to take the new technology and adapt, that is what some don't want to do.

 

Its not of a case well we have always done and like it this way.

 

A Company will not stand still for a few hundred hard core old school

 

Sorry if you don't agree but this is life

 

Regards

 

Yes, of course.

 

We don't like change.

 

That's it.

 

I shall quickly modify the title of the letter to read: "Sign here if you don't like change".

 

Thank you Mr. Failkner for helping me "see the light"... so to speak.

[Prosophos]

Advertisement (gone after registration)

Processed properly the 240 files will show superior micro contrast and transitions.

 

Jaap,

 

Do you know where Leica keeps these M240 images locked up? Are they in a secret vault together with the "superior colour" images I keep reading about, but never see?

 

On a serious note...

 

In most instances, the M240 is celebrated for its physical refinements but very few people seem to be passionate about its output, say like the M8 or M9/M-E or Monochrom. Don't you find that interesting in the least?

[01af]

With the "right amount of processing", people can make a digital file look like film. With the right amount of processing, you can make an M (Typ 240) file look like an M9 file. A lot of people would rather spend time shooting with the technology that gives them results that they like without spending a lot of time processing it.

You're confusing "the right amount" with "the right way" ... insinuating an awful huge amount. However, while the right way may take some time to get worked out, once you know the right way, it'll take no more processing time than the wrong way.

 

 

Otherwise we would never have a market for the M Monochrom.

Uh oh—you're blatantly missing the point of the M Monochrom ... but that's a different topic entirely.

 

 

The difference between CMOS and CCD sensors is "in the noise".

Exactly that's where the difference is indeed.

[Lenshacker]

You're confusing "the right amount" with "the right way" ... insinuating an awful huge amount. However, while the right way may take some time to get worked out, once you know the right way, it'll take no more processing time than the wrong way.

 

 

 

Uh oh—you're blatantly missing the point of the M Monochrom ... but that's a different topic entirely.

 

 

 

Exactly that's where the difference is indeed.

 

I get the point of the Monochrom. I get the point of processing. I've been using Digital imaging sensors since 1981. It was nice to get paid to write image processing code for a decade. I like the M9 and the M Monochrom, will use them for a long time. I have a Nikon Df- the sensor in it is amazing, has triple the saturation count of the M240.

 

Next time you accuse someone of missing the point, try to understand where they are coming from.

[01af]

Are they in a secret vault together with the "superior colour" images I keep reading about, but never see?

"Superior colour"? Are you kidding?

 

The M (Typ 240)'s colour rendition is rotten. In this particular respect, it's one of the worst digital cameras I know—together with the M9 which also is terrible, albeit in a slightly different way. You can get good colours out of these two cameras if you care about colour but it will take some effort: make your own custom profiles. Those provided in Lightroom are mostly unusable; they must have been prepared by colour-blind people. Those provided in Capture One are better; at Phase One, they have a guy who cares about and understands digital colour.

 

 

... but very few people seem to be passionate about [the M (Typ 240)'s] output, say like the M8 or M9, M-E, or M Monochrom. Don't you find that interesting in the least?

Well—at or near base ISO, the noise level of CCD sensors is so low they can get away without any in-camera noise reduction which will kick in only at higher ISO settings. In contrast, the native noise level of CMOS sensors is so high they desperately require on-chip noise-reduction stages on the actual sensor, in the A/D converters, and in the image processor even at base ISO. When disabling in-camera noise reduction in the camera's menu, you'll only affect the last stage, not the earlier stages. Without these in-camera noise reductions, CMOS images at base ISO would look like CCD images at, umm, ISO 3200/36° or so. So with these in-camera noise reductions, CMOS images at base ISO are basically noise-free while CCD images at base ISO keep a very low, subliminal noise floor.

 

It's this noise floor that gives base-ISO CCD images their apparent 'magic'—superior microcontrast, natural-looking tonal transitions, better three-dimensionality. People who don't understand where this magic is actually coming from tend to praise CCD technology. But then, you can have the very same magic in base-ISO CMOS images, too—just add a very small dose of noise. It really is just as simple as that.

 

In Lightroom, you can use the Film Grain feature from the FX panel (try Amount 5 - 15, Size 10 - 15, Roughness 40 - 50, for example). Or in Photoshop, you can use the Noise > Add Noise filter (Uniform, Monochromatic, Amount 1 - 3 %) through an inverted luminance mask (so the shadows will receive more, the highlights less additional noise). The optimal amount of noise to add depends on the image size, the image content, and the beholder's preference.

 

You might be temped to say, "but that's not the real thing, that's fake!" No, it isn't. It's just exactly the same as what you'd get out of a CCD camera at base ISO. Try it, and you'll see.

[Prosophos]

"Superior colour"? Are you kidding?

 

The M (Typ 240)'s colour rendition is rotten. In this particular respect, it's one of the worst digital cameras I know—together with the M9 which also is terrible, albeit in a slightly different way. You can get good colours out of these two cameras if you care about colour but it will take some effort: make your own custom profiles. Those provided in Lightroom are mostly unusable; they must have been prepared by colour-blind people. Those provided in Capture One are better; at Phase One, they have a guy who cares about and understands digital colour.

 

 

 

Well—at or near base ISO, the noise level of CCD sensors is so low they can get away without any in-camera noise reduction which will kick in only at higher ISO settings. In contrast, the native noise level of CMOS sensors is so high they desperately require on-chip noise-reduction stages on the actual sensor, in the A/D converters, and in the image processor even at base ISO. When disabling in-camera noise reduction in the camera's menu, you'll only affect the last stage, not the earlier stages. Without these in-camera noise reductions, CMOS images at base ISO would look like CCD images at, umm, ISO 3200/36° or so. So with these in-camera noise reductions, CMOS images at base ISO are basically noise-free while CCD images at base ISO keep a very low, subliminal noise floor.

 

It's this noise floor that gives base-ISO CCD images their apparent 'magic'—superior microcontrast, natural-looking tonal transitions, better three-dimensionality. People who don't understand where this magic is actually coming from tend to praise CCD technology. But then, you can have the very same magic in base-ISO CMOS images, too—just add a very small dose of noise. It really is just as simple as that.

 

In Lightroom, you can use the Film Grain feature from the FX panel (try Amount 5 - 15, Size 10 - 15, Roughness 40 - 50, for example). Or in Photoshop, you can use the Noise > Add Noise filter (Uniform, Monochromatic, Amount 1 - 3 %) through an inverted luminance mask (so the shadows will receive more, the highlights less additional noise). The optimal amount of noise to add depends on the image size, the image content, and the beholder's preference.

 

You might be temped to say, "but that's not the real thing, that's fake!" No, it isn't. It's just exactly the same as what you'd get out of a CCD camera at base ISO. Try it, and you'll see.

 

Honestly, I don't mind if the files come out of a sausage maker, as long as they look good and behave well under manipulation. Then I'm happy.

 

The M9 sensor provides that, the A7S and Df sensors do that, and film does that.

 

In the digital realm, the M9/M-E/M8 would be my first choices because they do that the best at base ISO.

 

Ooops, I'm repeating myself.

[NightSun]

It's this noise floor that gives base-ISO CCD images their apparent 'magic'—superior microcontrast, natural-looking tonal transitions, better three-dimensionality. People who don't understand where this magic is actually coming from tend to praise CCD technology. But then, you can have the very same magic in base-ISO CMOS images, too—just add a very small dose of noise. It really is just as simple as that.

 

Well, the CCD has some advantages in noise performance from the very outset based on the underlying architectural differences, but we should not exaggerate too much on this, either. In reality image noise has many sources, and some of them are dependent on the sensor architecture itself, e.g. dark current noise and fix pattern noises. However, except for the very darkest image parts, the dominating noise source is photon shot noise, which does not directly concern the image sensor, because it is solely dependent on the number of photons you collect. And with the same photon flux on a sensor surface, the single most factor that influences this is pixel size. Whatever sensor architecture you have behind it, if your pixels are of the same size, you will count the same number of photons in that well, and the square root of this number will be noise, no matter what. Of course, the issue gets a bit more subtle, because still it's a question that how many photons you can effectively utilize from their total number hitting the pixel (quantum efficiency). But in this regard, CMOS sensors went through quite much development in the last few years. So, the difference is not so that big between CCD and CMOS sensors, when it comes to noise performance, as it was years before.

[Antonio Russell]

I think that its very hard to decide one way or the other (not owning either camera).

 

Peter,

 

If the issue is such a slam dunk as you say then why is it you will not publish any evidence (i.e. images) to back up your claims for CDD vs CMOS?

 

We all see things differently which is fine, but if you are intending for a manufacturer to change such a core construction component then should you not at least provide some evidence to back up your claims?

 

Surely this should be relatively easy given that they are both digital cameras and you have proven yourself very adept at online publishing.

 

Not wishing to pick a fight or anything but surely this would add weight to your case?

 

Antonio

[jaapv]

Jaap,

 

Do you know where Leica keeps these M240 images locked up? Are they in a secret vault together with the "superior colour" images I keep reading about, but never see?

 

On a serious note...

 

In most instances, the M240 is celebrated for its physical refinements but very few people seem to be passionate about its output, say like the M8 or M9/M-E or Monochrom. Don't you find that interesting in the least?

 

Do you seriously think that I would allow Leica to store my images in a vault ?

 

As I wrote it took me half a year to get my 240 images to the point that I prefer them over my M9, but then it took me longer to get the M9 at a satisfactory point.

[Prosophos]

I think that its very hard to decide one way or the other (not owning either camera).

 

Peter,

 

If the issue is such a slam dunk as you say then why is it you will not publish any evidence (i.e. images) to back up your claims for CDD vs CMOS?

 

We all see things differently which is fine, but if you are intending for a manufacturer to change such a core construction component then should you not at least provide some evidence to back up your claims?

 

Surely this should be relatively easy given that they are both digital cameras and you have proven yourself very adept at online publishing.

 

Not wishing to pick a fight or anything but surely this would add weight to your case?

 

Antonio

 

It's a fair question Antonio, and I'll respond to you because I believe you are genuinely asking and not trying to be clever.

 

In truth, the differences are obvious to me and it amazes me that many people can't see the difference. Many others can, of course.

 

For those that cannot see a difference, I don't think I'll change their minds with any demonstration I come up with. It's a trap to think otherwise, and it will create a forum thread-zilla that's even larger than this current thread.

 

To be blunt, I don't want to put the effort into it. I'd rather enlist like-minded photographers/artists/enthusiasts who wish to see CCD continue - who are willing to sign my letter. As for the rest, they've decided to enter the LUF Leica M9/M-E/Monochrom sub-forum and create conflict. Why? Not sure, other than they enjoy the sport of it.

[Jeff S]

 

In most instances, the M240 is celebrated for its physical refinements but very few people seem to be passionate about its output, say like the M8 or M9/M-E or Monochrom. Don't you find that interesting in the least?

 

As I wrote earlier, I've reached the point…in the context of the total workflow... where my prints using the M are as good, and sometimes better, than with my M8.2…color and b/w. My M8.2 sits idle as back-up, which wouldn't be the case if I felt IQ was degraded, despite my preference for the M as a better built RF camera.

 

Passion? I reserve that for photographing and attempting to make superior pics and prints. I haven't found an M camera….film or digital…that got in the way of that passion. And frankly, most of the incremental gains in pure print quality that I've experienced have come more from the back-end of the workflow….processing, papers, display conditions, etc…than from any camera or lens. That was true in my darkroom days, and still in the 'lightroom' era.

 

Jeff

Edited January 29, 2015 by Jeff S

[Antonio Russell]

It's a fair question Antonio, and I'll respond to you because I believe you are genuinely asking and not trying to be clever.

 

In truth, the differences are obvious to me and it amazes me that many people can't see the difference. Many others can, of course.

 

For those that cannot see a difference, I don't think I'll change their minds with any demonstration I come up with. It's a trap to think otherwise, and it will create a forum thread-zilla that's even larger than this current thread.

 

To be blunt, I don't want to put the effort into it. I'd rather enlist like-minded photographers/artists/enthusiasts who wish to see CCD continue - who are willing to sign my letter. As for the rest, they've decided to enter the LUF Leica M9/M-E/Monochrom sub-forum and create conflict. Why? Not sure, other than they enjoy the sport of it.

 

 

Fair enough. I must say that I am one of those who have looked at your site myself and can't really see any difference between the two, so I guess that if you are pushing for change based on your preference for CCD rather than anything objective that can be proven one way or another then I am thinking you will need a lot more signatures, and probably a lot of folks willing to put down some cash in advance. Leica are keen on their special editions aren't they! Good luck.

Edited January 29, 2015 by A Aparicio

[01af]

... I'll respond to you because I believe you are genuinely asking and not trying to be clever. [...] As for the rest, they've decided to enter the LUF Leica M9/M-E/M Monochrom sub-forum and create conflict. Why? Not sure, other than they enjoy the sport of it.

So people either agree with you or create conflict!? This level of negativity surprises me—I wouldn't have expected this from you. I, for one, would genuinely like to know if you see something that I cannot see. So far, you have always refused to back up your claims with some substance; instead, you are always backing out with platitudes. Hence, my impression so far is, no substance from you. You don't see anything I don't.

 

 

In truth, the differences are obvious to me and it amazes me that many people can't see the difference. Many others can, of course.

 

For those that cannot see a difference, I don't think I'll change their minds with any demonstration I come up with.

You divide people into two groups: those who see the difference and those who don't. You keep ignoring a third group—those who see the difference but don't consider it important because they know how to deal with it. (There may be a fourth group, for sake of completeness: those who see a difference but simply don't care.)

 

Of course there are differences; they are easy enough to see. The question is, how to assess those differences? You believe they are substantial; I believe they are not. I have even backed up my claim with pictures ... but your reaction to this was exactly what you're expecting from others: blunt refusal to accept a proof. By the way: the first crop was from the M (Typ 240); the second from the M9.

 

And I don't mean to slag CCD technology. In fact, my digital cameras with CCD sensors outnumber those with CMOS sensors, and I don't plan to get rid of any of them anytime soon. I just want to debunk the misconception that CCD sensors have anything to them that cannot be had with CMOS sensors also.

 

Finally, I didn't leave a comment against your open-letter campaign on your own web-site because that's what I'd consider unfriendly and disrespectful. But if you come over here to the LUF and open a thread about your request then you have to deal with some discussion and contradicting viewpoints, and I expect you to handle that without getting snotty.

[pgk]

I just want to debunk the misconception that CCD sensors have anything to them that cannot be had with CMOS sensors also.

But a fundamental question remains. If the CMOS sensor can do everything that the CCD can, why is it that the workflow has to change so significantly? I would again suggest that for consistency the output from new camera models should match that of previous model,s which it does not. In fact all my camera models have output images differently, not due apparently not to sensor according to many in this thread. Does a lack of output consistency not strike anyone else as an oddity?

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

Because the filter array is different the output is different.

It is a misconception that post processing is a one size fits all. You would not be surprised if you had to adapt your darkroom procedures to different films to get the same results