Well, it might interest some...

[mhoersch]

Advertisement (gone after registration)

Does the DXO RAW Converter support the Leica M8 and DMR DNG format? I can't find this information on their website.

 

Perhaps my hint was too subtle: As Jaap already pointed out there may be some doubt as to whether DXO really used the M8's raw format, or rather JPG out of the camera. The DXO software, as far as I know, doesn't support DNG.

 

Comparing M8 JPG files to RAW files of other brands might not be quite fair.

[AlanG]

But Alan, here's no free lunch here: any software that 'removes' vigging is actually increasing exposure and therefore noise in the periphery of the frame. Ditto with edge and corner sharpening routines. You can't make a silk purse out of a sow's ear unless you're willing to half-close your eyes when you look at said purse.

 

Tim

 

You may be correct in theory, but this does not often matter to me in practice. After two years of critical use, I am very aware of what it does and am extremely satisfied with the results. Maybe you and other are more critical than I am.

 

By the way, you know what? I sometimes find myself turning off the "vignetting correction," because the vignetting actually "improves" the image.

[Advertisement]

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

[AlanG]

Perhaps my hint was too subtle: As Jaap already pointed out there may be some doubt as to whether DXO really used the M8's raw format, or rather JPG out of the camera. The DXO software, as far as I know, doesn't support DNG.

 

Comparing M8 JPG files to RAW files of other brands might not be quite fair.

 

Give DXO a little credit here. I think you should consider reading everything at the DXOmark site to get a good understanding of their very thorough methodology. They explain that they make direct measurements of the raw data without using any converter.

 

On the first page, DXOmark explains the concept of their camera Sensor ranking as:"A simple, instructive scale developed by DxO Labs to help users easily visualize the RAW image quality performance of today's digital cameras."

[asfeir]

The site does not explain how the dynamic range is measured. And as pointed out earlier, this depends on the lens unless it is obtained from sensor data.

The M8 uses a KAF-10500 Kodak CCD that has a dynamic range of 71.5 dB or 11.9 ev

[AlanG]

The site does not explain how the dynamic range is measured. And as pointed out earlier, this depends on the lens unless it is obtained from sensor data.

The M8 uses a KAF-10500 Kodak CCD that has a dynamic range of 71.5 dB or 11.9 ev

 

Yes, the site does explain how they do the dynamic range measurement. They have made a high quality step tablet and first measure the noise, then the dynamic range.

 

Noise protocol

[delander †]

I wonder how many other brand/camera model dedicated forums are having difficulty in coming to terms with the results of these tests?

 

Perhaps they should extend the tests to those nice expensive MF cameras and backs.

 

Then the Canon G10 will really shine and Michael Reimann's comparison of it with the Hassleblad P45+ will come through.

 

Jeff

[Isabelle Lenatio]

Advertisement (gone after registration)

I wonder how many other brand/camera model dedicated forums are having difficulty in coming to terms with the results of these tests?

 

Perhaps they should extend the tests to those nice expensive MF cameras and backs.

 

Then the Canon G10 will really shine and Michael Reimann's comparison of it with the Hassleblad P45+ will come through.

 

Jeff

 

I can imagine Nikon and Canon users will be quite pleased with these results, they do come out very nicely

[sean_reid]

Again, the DXO tests are measuring *some* aspects that affect a camera's file quality but certainly not all of them. These kinds of rankings and such are not definitive in any way, they're just interesting data which should be seen in the context of a larger picture. Two aspects of the M8 that make the files so satisfying to me (the lack of AA filter and the ways in which many RF lenses draw) aren't described in the DXO tests.

 

The tests do discuss the camera's S/N ratio (with results that shouldn't surprise any of us) and give one interpretation of its DR (having not removed the confounding variable of lens mounted). With a high contrast lens, the M8 will clip highlights sooner than, say, a 1Ds Mk III and that won't come as a surprise to many professionals who've been working with both.

 

Regarding the M8, there's really not a lot of news in the DXO tests. They give us quantification of things most of us have sussed out by now. The strengths and weaknesses of the M8 are what they are...still. And nothing in the test results take away from how beautiful those M8 files can be.

 

No doubt, there may be leagues of over-simplifiers out on the web who will use the DXO results to make sweeping generalizations about this camera vs. that. But those folks do that anyway. Many people don't like the messy mix of pros and cons that really must be considered when comparing camera A to camera B. So the DXO results could be a short cut basis for argument when one doesn't want to think all of the aspects through carefully. They may well be a favorite source, as well, for some people who've never worked with the cameras they are arguing about.

 

Cheers,

 

Sean

[sean_reid]

I can imagine Nikon and Canon users will be quite pleased with these results, they do come out very nicely

 

Nikon people might be a bit happier than Canon folks but my own testing suggests that, in difficult incandescent light, the high ISO performances of the D700 and 1Ds Mk III are a lot closer than those tests suggest. The Nikons hold the luminance noise down well but, pushed hard enough, there's chrominance noise from the blue channel especially.

 

Both cameras wipe the floor with the M8 when it comes to high S/N at high ISO but we knew that already.

 

Cheers,

 

Sean

[rosuna]

The problem isn't with the M8, but with the M8.2.

 

The M8.2 is more expensive and the sensor is the same of the previous model. It isn't competitive for the 2008-2009 standards, specially if you consider the price.

 

The gap between the M8 and the 5D was acceptable two years ago, but it has been widening. Now the sucessor of the M8 is the M8.2, and the sucessors of the 5D are the new CMOS third generation models (5D Mark II, D700, D3, A-900)...

[tomasis7]

well said, sean.

 

I just got to think about processing things inside the body. Leica usually leave all hux flux stuffs off under processing of fresh raw data.

 

It can happen that quality of DNG compressed file does down a bit sacrificing some data for more compact file if Kodak spec says for 12 EV.

 

I wonder if it can be possible to extend DR if it is used with low contrast lens?

[Isabelle Lenatio]

Nikon people might be a bit happier than Canon folks but my own testing suggests that, in difficult incandescent light, the high ISO performances of the D700 and 1Ds Mk III are a lot closer than those tests suggest. The Nikons hold the luminance noise down well but, pushed hard enough, there's chrominance noise from the blue channel especially.

 

Both cameras wipe the floor with the M8 when it comes to high S/N at high ISO but we knew that already.

 

Cheers,

 

Sean

 

Sean, I know, I knew .... as mentioned before though it is quite disappointing to see the differences are so big. I'm still happy with my M8 for it provides me with an image quality I like and appreciate.... but still, you know.... But then I'm not one of those users whom pull out every possible argument and comparison to prove the DxO is wrong, conducted their tests wrong, were incomplete etc. etc. If it was done by this Nikon guy, the one with the power-color images I would doubt the results but DxO in my opinion is a serious enough company to be able to conduct and complete serious enough tests.

[tomasis7]

The problem isn't with the M8, but with the M8.2.

 

The M8.2 is more expensive and the sensor is the same of the previous model. It isn't competitive for the 2008-2009 standards, specially if you consider the price.

 

The gap between the M8 and the 5D was acceptable two years ago, but it has been widening. Now the sucessor of the M8 is the M8.2, and the sucessors of the 5D are the new CMOS third generation models (5D Mark II, D700, D3, A-900)...

 

you forgot that rangefinder is peach, SLR is apple

[sdai]

I have some quick relief to the conundrum here ... if the M8 doesn't look good when compared to the 5D, why don't you just select a Nikon D70 from the pull-down menu instead?

 

It's so easy.

[Isabelle Lenatio]

I have some quick relief to the conundrum here ... if the M8 doesn't look good when compared to the 5D, why don't you just select a Nikon D70 from the pull-down menu instead?

 

It's so easy.

 

Yes well .... that would really make one burst out in tears, I'll stick to the G10 comparison for now

[sean_reid]

well said, sean.

 

I just got to think about processing things inside the body. Leica usually leave all hux flux stuffs off under processing of fresh raw data.

 

It can happen that quality of DNG compressed file does down a bit sacrificing some data for more compact file if Kodak spec says for 12 EV.

 

I wonder if it can be possible to extend DR if it is used with low contrast lens?

 

Yes if the subject contrast is beyond what a sensor/processing engine can record, lower contrast lenses can help increase *effective* DR. By lifting the shadow values (often via a bit of veiling flare) those lenses allow the shadow details to be recorded as a brighter tones (Zone II instead of Zone I for example) and this lifts that shadow detail away from the noise floor. The noise floor is effectively what limits the degree to which a camera can "reach" into the shadows. It is especially relevant for the M8 which produces a bit of noise even at ISO 160.

 

There are trade-offs, to be sure, with lower contrast lenses (flare, mid-tone separation that not all will like, etc.) but they do have this advantage for subjects in contrasty light. That's why I started calling them "sunny day lenses" a few years back.

 

Some complain about the high contrast of ZM (and some Leica ASPH) lenses but the fact is that those lenses show such high contrast in large part because they control falre so well.

 

Cheers,

 

Sean

[AlanG]

I wonder if it can be possible to extend DR if it is used with low contrast lens?

 

No, that would reduce the photo's dynamic range. Just like shooting on a foggy day. The DXO test measures the dynamic range of the sensor not the lens. A higher contrast lens would help assure that the maximum contrast of the scene is reaching the sensor, but I'd think any reasonable quality lens would do as the target has a much greater range than the sensor is capable of recording.

 

In other words, if the target has a 14 stop range, and a poor lens causes enough flare or other problems to knock a stop off of that range, there is still a 13 stop range of brightness reaching the sensor. I think the assumption is that most modern lenses can pass higher than a 12 stop range to the sensor. Especially considering the test is of small white holes in an otherwise black subject. (So the flare potential is reduced.)

 

If you consider a snow scene with very dark shadow areas, the actual scene brightness might be 18 to 20 stops and most lenses can handle this even if film and sensors can't. That is why you can produce HDR images with multiple exposures.

[sean_reid]

Yes well .... that would really make one burst out in tears, I'll stick to the G10 comparison for now

 

Gosh, it shouldn't. None of those cameras is a rangefinder. The M8 wasn't at the top of the technical charts even when it was first introduced but, to me at least, that's not the end of the world.

 

There are technical improvements many of us would like to see in the M9. Meanwhile, the M8 still does what it does and I still love it even though the D700 kills it at high ISO.

 

Cheers,

 

Sean

[sean_reid]

The DXO test measures the dynamic range of the sensor not the lens. A higher contrast lens would help assure that the maximum contrast of the scene is reaching the sensor, but I'd think any reasonable quality lens would do as the target has a much greater range than the sensor is capable of recording.

 

In other words, if the target has a 14 stop range, and a poor lens causes enough flare or other problems to knock a stop off of that range, there is still a 13 stop range of brightness reaching the sensor. I think the assumption is that most modern lenses can pass higher than a 12 stop range to the sensor. Especially considering the test is of small white holes in an otherwise black subject. (So the flare potential is reduced.)

 

If you consider a snow scene with very dark shadow areas, the actual scene brightness might be 18 to 20 stops and most lenses can handle this even if film and sensors can't. That is why you can produce HDR images with multiple exposures.

 

Hi Alan,

 

We need to get together sometime and test these two opposing theories in practice. My observations, having used lower contrast lenses on DRFs for several years now, are not consistent with what you've written.

 

I do agree that it is not the high contrast lenses that affect the DR results. If the lens contrast is high enough it is the sensor DR that is measured (the gate is wider than the cow, so to speak). It is the lower contrast lenses that can affect the result. BTW, I spoke with Peter Karbe about this and he agreed though he's not about to start intentionally designing lower contrast lenses for Leica.

 

Any other photographers who've been using moderate and lower contrast lenses on DRFs want to chime in?

 

Cheers,

 

Sean

[Advertisement]

Hallo,

Look here: Leica M

[sean_reid]

No, that would reduce the photo's dynamic range.

 

Perhaps there's some confusion here. Lower contrast lenses reduce contrast so yes, of course, the photo's contrast will be lower (unless it is changed in post). But by lowering the subject's contrast range, a sensor is better able to record detail from highlight to shadow (because of the noise floor).

 

It's a lot like shooting Tri-X at 250 and pulling development by 15% to get a longer scale negative.

 

I know you know this stuff, Alan so maybe there was some confusion in the Q and A exchange.

 

Will the M8 be able to handle the brightness range of a contrasty scene a bit better with a lower contrast lens? Yes, it will but there are trade-offs that come with that.

 

Given equally high contrast lenses on both cameras, will the Canon 1Ds III record detail across a broader dynamic range than the M8. According to DXO - yes. And that matches my experience.

 

Does an older, lower contrast Nikkor, have the same affect on, say, a D200? Yes, it does.

 

Cheers,

 

Sean