carstenw wrote
but they have gained some experience with both autofocus and stabilisation in the Digilux 3, so it doesn't seem too far fetched that they could design something useful.



A bit more than that - Leica was the original inventor of autofocus, decided that it was of little or no use for general photography, did not fit their philosophy and subsequently lost it in the whole Minolta/Kyocera saga.

If someone's interested, a second-hand shop in Milan sells a mint-cond. DMR for 2400€, 2 years warranty.
The only reason I've never bought it is AF lackness. Colour rendition much better than my Canon (1Ds2, 5D).
http://www.newoldcamera.com/Scheda.aspx?Codice=GU2325&Tipo=DIGITAL%20MODUL-R

Colour rendition much better than my Canon (1Ds2, 5D)

This old chestnut. Yet we still see superb photographers like Arne posting DMR images with seductive yet very implausible colour, just like in the old days. Regardless of whether it is possible to produce realistic colours with a DMR, very experienced FMers routinely don't. Turquoise skies and yellowy greens abound. If I want that kind of colour inaccuracy I prefer to add it myself in photoshop. Colour accuracy matters when you can't change film.

Probably your're right, that was my impression using the DMR for a few hours. At least, say I'm not completely satisfied with my Canon colour rendition and I strongly WANTED the DMR to be better

Yes, the saturation of the posted images seems high. But I was rather blown away by the color of the DMR that I had for about a year. It went in exchange for an M8, far more portable, more to a lightweight shooing style. But the colors of the DMR.... hard to replace. I think they might have been better than the M8, by a nose.

But the colors of the DMR.... hard to replace.

Impossible! But necessary ;D

Anyone can mess up color rendition, it's much more difficult to get it right. I've been completely delighted with the DMR's color quality.

brainiac wrote:This old chestnut. Yet we still see superb photographers like Arne posting DMR images with seductive yet very implausible colour, just like in the old days. Regardless of whether it is possible to produce realistic colours with a DMR, very experienced FMers routinely don't. Turquoise skies and yellowy greens abound. If I want that kind of colour inaccuracy I prefer to add it myself in photoshop. Colour accuracy matters when you can't change film.

There is some validity to this. However, the fact that I and others manipulate the colours in our images, has little or no bearing on the question of wether the DMR is capable of accurate colour reproduction. Modyfying colours is a creative tool. Having said that, I would also prefer a a priori correct reproduction as a starting point. Unfortunately no digital camera to date has 100% correct colours, and neither has film.
As for turquoise skies, I'm a bit puzzled by this statement since we have been going over it before, and came to the conclusion that in the cases discussed it was caused by the RAW converter and not the DMR. My further experience does not confirm excessive turquoise in the skies. Here's an example straight out of the box (C1). No manipulations, no sharpening.

The second point you mentioned specifically, the yellow greens, is in my experience, correct. It is in my opinion perhaps the weakest piont of the DMR, but it is often a quite pleasant aberration that many users seem to like. I don't, but I don't much like the bluish cast I get with my 1DsII either...
Sometimes, surprisingly, the DMR-greens can be almost spot on, like in this image. Again, no manipulation to the colours.

That green does appear too yellow and the bushes on the left have some brown in them as well. I like the look though and it reminds me very much of the color palette my old Fuji S2 gave. With all the color controls available in the RAW processors today, I find it very hard to believe one could not get whatever look they wanted. I for instance was able to tame my Canon 5D to give me those Fuji colors I liked with the correct amount of WB/Tone settings..

Thanks Arne, your shots demonstrate exactly what I was talking about. Of course profiling can solve a multitude of evils, but the question remains how you separate a green object that looks yellowy from a yellowy green object that also looks yellowy green. As I understand it, this can only be done with guesswork and careful brushstrokes.

Routine Canon 5D shots seem, to my eye, to render much more natural colour without adjustment. Of course it is always possible to shift towards turquoise and mustard in a 5D shot, but it seems likely to me that the return journey would be tonally lossy.

These colour issues discouraged me from going the DMR route as they seem to outweigh the claimed advantages.

Among those who have actually owned and used both the 5D and DMR the opinions are unanimous: they prefer the DMR's image quality, not just for the color quality but for the dynamic range and for the ease of extracting the most from the RAW files, in particular the ability to pull details out of shadow areas where the 5D files fall apart like cheesecloth stretched too far.

I've found that I mostly like the DMR rendition of sky. I don't have the 5D, so I can only comment on the 1DsII and 400D, both of which I find render the sky with a magenta cast. Magne's profiles have improved the general colour reproduction of the 1DsII, but not fully eliminated this objectionable cast. Perhaps I'm more sensitive to magenta than cyan..
It would have been quite interesting to see what a Magne Nielsen profile could do with the DMR.
With the recent FW update Leica seems to have tweaked the colour response a little, but I haven't used the DMR enough to comment on it; nice that they still issue improvements though.
I agree with Douglas' comments on the DR and the ability to push the RAW files around without breaking up. Probably related to the 16 bit signal processing.

telyt wrote:
Among those who have actually owned and used both the 5D and DMR the opinions are unanimous: they prefer the DMR's image quality, not just for the color quality but for the dynamic range and for the ease of extracting the most from the RAW files, in particular the ability to pull details out of shadow areas where the 5D files fall apart like cheesecloth stretched too far.

As a 5D and a 1DIII owner I can say that Canon has definitely improved both color quality (richness?) and shadow detail with the newest generation. This isn't going to be noticed particularly with web evaluations. But most people here will definitely appreciate the difference.

I'm pretty much convinced that most of the quality improvements we see the DMR and M8 are due to CCD. Canon newest models close this gap some. But it seems when looking at IQ alone CCD has fundamentally superior properties.

I don't think Leica cares all that much about color accuracy. It seems to me they manipulate blue/green to be pleasing to western eyes. I like the look. it's not good or bad, just a design choice.

I would hope the DMR file was better than 5D at twice the price.

brainiac wrote:
Colour rendition much better than my Canon (1Ds2, 5D)

This old chestnut. Yet we still see superb photographers like Arne posting DMR images with seductive yet very implausible colour, just like in the old days. Regardless of whether it is possible to produce realistic colours with a DMR, very experienced FMers routinely don't. Turquoise skies and yellowy greens abound. If I want that kind of colour inaccuracy I prefer to add it myself in photoshop. Colour accuracy matters when you can't change film.

Shooting the 5D and 1DIII together, the 5D's colors really suck. I thought they were pretty good before the 1DIII. Currently, the file improvements with Canon's newest cameras is unappreciated on the web.

dcmiller wrote: As a 5D and a 1DIII owner I can say that Canon has definitely improved both color quality (richness?) and shadow detail with the newest generation. This isn't going to be noticed particularly with web evaluations. But most people here will definitely appreciate the difference.

Sounds promising, and agrees well with other reports I've read of the 1DIII. Hopefully the same improvements have found their way into the new s-camera as well, I have a 1DsIII on order

The point regarding CCD vs CMOS is interesting, as it has been claimed that the Kodak CCD chip in the DMR is essentially a (small) MF sensor and that this goes some way to explain its high dynamic range and colour rendition.

Yes! The first time I shot my 5D I was shocked by that magenta tainted blue sky and it took me a while tinkering with the in camera tone settings to get things to the point that an out of camera jpeg and an ideal RAW conversion looked the same, or at least very very close. My Fuji S2 had much better color. One gripe I still have with the 5D is its rendering of reds. Even using all the various profiles including Magne Nielson's and regardless of using C1 or ACR, the reds are just not as good nor accurate as from the S2. I partially suspect some of this could be with the Canon lenses as I have noticed a more accurate and better Red when using Zeiss lenses.

photoArne wrote:
I've found that I mostly like the DMR rendition of sky. I don't have the 5D, so I can only comment on the 1DsII and 400D, both of which I find render the sky with a magenta cast. Magne's profiles have improved the general colour reproduction of the 1DsII, but not fully eliminated this objectionable cast. Perhaps I'm more sensitive to magenta than cyan..
It would have been quite interesting to see what a Magne Nielsen profile could do with the DMR.
With the recent FW update Leica seems to have tweaked the colour response a little, but I haven't used the DMR enough to comment on it; nice that they still issue improvements though.
I agree with Douglas' comments on the DR and the ability to push the RAW files around without breaking up. Probably related to the 16 bit signal processing.

Seems to me that the color differences are more likely due to some other part of the data capture process than whether or not it is done with CCD or CMOS.

Weird. My 5D's seem to nail sky colour as accurately as I could ever hope. Meanwhile I see turquoise skies all over FM. Maybe there really is some physiological difference in how each of us sees the spectrum. I see Arne's shot of the red and yellow building as having a very slightly too green sky.

How's this?:

brainiac wrote:
Weird. My 5D's seem to nail sky colour as accurately as I could ever hope. Meanwhile I see turquoise skies all over FM. Maybe there really is some physiological difference in how each of us sees the spectrum. I see Arne's shot of the red and yellow building as having a very slightly too green sky.

I think uncalibrated monitors, monitors that are calibrated but still can't render colours accurately, poor browser colour management and ignorance of colour space issues in general contribute to this.

OT: I'll say there's also the issue of 'taste'. I see a bias towards heavy saturation in the US, you can see it in their TV programmes that feature colours and captions that I find large, garish and ugly but they all do it so it must be the acceptable norm over there.

But Arne's from Norway...

Oh I was just going off on a Friday tangent ... nothing aimed at Arne, who's photo's I greatly admire.

brainiac wrote:
Weird. My 5D's seem to nail sky colour as accurately as I could ever hope. Meanwhile I see turquoise skies all over FM. Maybe there really is some physiological difference in how each of us sees the spectrum. I see Arne's shot of the red and yellow building as having a very slightly too green sky.

How's this?:

foxbat wrote:
brainiac wrote:
Weird. My 5D's seem to nail sky colour as accurately as I could ever hope. Meanwhile I see turquoise skies all over FM. Maybe there really is some physiological difference in how each of us sees the spectrum. I see Arne's shot of the red and yellow building as having a very slightly too green sky.

I think uncalibrated monitors, monitors that are calibrated but still can't render colours accurately, poor browser colour management and ignorance of colour space issues in general contribute to this.

OT: I'll say there's also the issue of 'taste'. I see a bias towards heavy saturation in the US, you can see it in their TV programmes that feature colours and captions that I find large, garish and ugly but they all do it so it must be the acceptable norm over there.

NTSC.

> Mind you, these can be subtle differences.

Yes, they can be, but the olive grass thing on the DMR is quite strong. And still nobody has explained how to find the patch of mustard that I spilled on the lawn. I accept that people love their DMR colours, but people also love bass and treble, until they hear a human voice. Does anybody want to show us some unaltered DMR skin tones?