Ming Thein on the Leica M Typ 240

[UliWer]

Advertisement (gone after registration)

... ISO 6400 is dramatically worse than ISO 2500 on the M9. I was shocked. And he detects an automatic and destructive in-camera noise reduction kicking in beginning as early as ISO 2500. ...

 

You must have read something at Sean Reid's which I missed. I found the following statement when Reid compared the M9 with the new M at ISO 2500:

 

"This is fairly a dramatic difference...At ISO 2500 the M9 file is quite rough - as we have long known. The M240 file on the other hand is just a bit gritty."

 

He further deals explicitly with banding of the M240 at ISO 3200 and ISO 6400, explaining that Leica is willing to address this in the production models (Reids used a pre-production example without LR profiles which he tells at least a dozen times). And he does not forget, that the M9 may show banding at ISO 1250 and ISO 2500.

 

He sums it up with:

 

"...The other thing we see often are makers who allow very high ISO settings but who then apply aggressive smoothing filters before writing the RAW files - often leaving them a mushy mess.

 

Leica is doing neither. They aren't over representing the high ISO capability and they aren't trying to hide the banding with aggressive in-camera smooting (though there does seem to be some mild overall noise filtering being applied to the high ISODNG files). In some lighting, cool daylight for example, one may find that the camera does fairly well even at ISO 6400. But in difficult warm lighting, especially when working for extended periods, one may want to watch for banding at high ISO."

[Cheekyguille]

A thorough analysis. Some sticking points, it says the new M gains 2 to 2.5 stops in high iso performance, and that the color is strange, not quite like that of CCD, yet it concludes that the improvements are not 1 but two generations ahead of the M9? Pardon me, it doesn't make sense. Also, metering in the M9 is off by 4 stops?

 

Clarification here is important, specially for M9 owners who like me see the improvements in IQ as evolutionary, with other more marked improvements in areas that aren't important for rangefinder use (i.e. video, EVF, use of R, F or V lenses on an M body). The main reason for using a rangefinder camera is simplicity, reading this review confirms that the new M is no longer simple, and also not substantially better than the prior generation.

 

The new M now competes with DSLRs and will be appealing to DSLR owners, perhaps for the first time. I hope that the new strategy works Leica and that this move isn't dilutive. A much better M to use with Leica's superior M glass will continue to be a destination that I am interested in.

[Advertisement]

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

[Guest malland]

...CMOS vs. CCD? Let the techies decide the merits of each. Image quality looks great, better , more efficient battery power, lower noise at higher ISO’s. These are the important things we’re looking for, correct? The M240 succeeds!. Let those that love to hear themselves speak, talk all they want about whatever it is they need to talk about...

Ulrich, first, I thought you were a techie. Second, although I would think that Leica will get the best image quality possible from this camera, there is little point, in my view, to speculate on whether that will be "better" — better than what? — or not. Third, to talk down to to "people who like to hear themselves talk" is something that I just call attention without commenting on — except that some people like the current color rendition of the M9 and, rightly or wrongly, associate that with the CCD sensor; and these people have every right to wonder or be concerned about whether the color rendition that Leica produces in the final M240 version will appeal to them as much.

 

—Mitch/Paris

Bangkok Hysteria (download link for book project)

[Dr. Ulrich Rohde]

The production M 240 has enhanced low ISO colors , I have both cameras, not the final one, and I can only repeat, the increased dynamic range makes more difficult pictures possible.

 

And yes many things are subjective, but I would like to remind all of you that the majority uses software to adjust the picture to their liking.

Is there an absolute measure of "best", obviously not

 

An yes my background is technical but picture evaluation is subjective

[Dr. Ulrich Rohde]

Maybe I should clarify two thing:

 

My interest is that the pictures (colors on a calibrated monitor) from the camera are correct. Photo shop can always change them, but I want the real color.

Likewise I am looking for the best dynamic range, this includes low noise. As I have both cameras, the M 240 gives you more opportunities.

 

As far as talking down to people , the M 9 successors with the older sensor are available and nowhere did

I give advise not the buy them.

 

The other huge improvement is the R adaptor and the electronic viewfinder .

 

Why are the discussions here held with attacking many of us, what does the word tolerance mean

 

I talked about hardware not living things

Edited February 24, 2013 by Dr. Ulrich Rohde
one typo

[photomeme]

Some sticking points, it says the new M gains 2 to 2.5 stops in high iso performance, and that the color is strange, not quite like that of CCD,

 

According to Reid, only 1.3 stops.

 

Puts says: "Some reviewers might hope that the M is a quantum leap forward compared to the M9 or M Mono, but that is hardly the case."

[jonoslack]

Advertisement (gone after registration)

According to Reid, only 1.3 stops.

 

See my post about this in the other thread. I'd be really surprised if it wasn't at least 1.5stops in the shipping firmware.

[photomeme]

See my post about this in the other thread. I'd be really surprised if it wasn't at least 1.5stops in the shipping firmware.

 

I've seen your posts speculating about this, but it just doesn't make any sense.

 

Nikon and Canon CMOS cameras have not generally improved their ISO performance between pre-production and final, because the redundant sampling noise reduction is hardwired on the sensor, unaffected by firmware.

 

According to Reid, there are worrisome signs of Leica adding an additional layer of noise reduction -- not of the on-CMOS-sensor variety, but a destructive median filtering sort -- from ISO 2500 and up. His images are pretty convincing, and it's a huge red flag.

 

And the ISO 1.3 advantage to which Sean Reid refers is after giving this mandatory destructive noise reduction a pass.

 

I'm not sure I see how this gets better between now and production. Leica might enable the user to turn off this median filtering (if Reid is correct). But that would reduce ISO performance further.

[AlanJW]

I've seen your posts speculating about this, but it just doesn't make any sense.

 

Nikon and Canon CMOS cameras have not generally improved their ISO performance between pre-production and final, because the redundant sampling noise reduction is hardwired on the sensor, unaffected by firmware.

 

According to Reid, there are worrisome signs of Leica adding an additional layer of noise reduction -- not of the on-CMOS-sensor variety, but a destructive median filtering sort -- from ISO 2500 and up. His images are pretty convincing, and it's a huge red flag.

 

And the ISO 1.3 advantage to which Sean Reid refers is after giving this mandatory destructive noise reduction a pass.

 

I'm not sure I see how this gets better between now and production. Leica might enable the user to turn off this median filtering (if Reid is correct). But that would reduce ISO performance further.

How many stops are you looking for? Somehow, I trust the subjective views of people like Jono and Chris Tribble.more than purported lab " measurements" that don't even include a final or even mature raw conversion.

[jip]

thanks for sharing, pre-ordered mine today.

[Paul J]

Hi Paul,

 

Interesting opinion. Thanks.

Question: From what we have seen so far, in your view, does the image quality of the M match that of the M9 for low ISOs?

Thanks again.

 

 

Hey K-H,

 

It's fair to say the M exceeds the M9 in many regards. A noticeable but small amount of resolution, less base noise, more DR, higher ISO performance.

 

In terms of adjusting the photo "any way you like" I can't agree with that just yet. I can spend hours adjusting a Canon file and be left unhappy with the result. I've had the same results with Nikon too. I put down the Canons for Phase One. But last year I put down the Phase One countless times for the M9.

 

There is just something about the new M's tonal and colour response that still looks flat. The Phase one I own with 12.5 stops of Dynamic Range doesn't share the same look. That's not to compare the two different systems but the look associated with the dynamic range. I think the flatness is from the CMOS sensor properties not the firmware and not the added Dynamic range. It does not seem to share the same brilliance or richness. I will need to wait till I have my M in my hands and thoroughly test it to be sure though.

 

It's a good upgrade. But what I was expecting, what the point I'm trying to make, is that I think it should be a ground breaking upgrade and one that is in keeping with that brand and it's price which is twice what the competition is offering for seemingly and reportedly better IQ. The M9 succeeded in that so why shouldn't the M? What am I spending twice the price on now? Even if I was the sort that bought into Nostalgia, the M is certainly not that. So it can potentially alienate it's existing user group and not fulfill the needs and expectations of the new user group....think about that for a little while.

 

All the science and numbers can sound like it means alot. Real world shooting is another thing entirely though. Personally, I rarely go over 160 ISO. I've spent my life working with 6 stops of DR in Transparency so I can light and meter accordingly. I own alot of batteries for my M9 and never run out, I don't want to shoot movies, I can focus perfectly well with the optical finder and don't own a bag full of R Lenses. Lastly, I just don't like flat homogenised images which look more like they were shot in a Canon or Nikon than ever before. But it seems the M upgrade was not aimed at those like me.

 

Either way, it's too early to tell. Let's get a hold of the camera and see what happens.

[FPB]

.....

 

There is just something about the new M's tonal and colour response that still looks flat. The Phase one I own with 12.5 stops of Dynamic Range doesn't share the same look. That's not to compare the two different systems but the look associated with the dynamic range. I think the flatness is from the CMOS sensor properties not the firmware and not the added Dynamic range. It does not seem to share the same brilliance or richness. I will need to wait till I have my M in my hands and thoroughly test it to be sure though.

 

.....

 

All the science and numbers can sound like it means alot. Real world shooting is another thing entirely though. Personally, I rarely go over 160 ISO. I've spent my life working with 6 stops of DR in Transparency so I can light and meter accordingly. I own alot of batteries for my M9 and never run out, I don't want to shoot movies, I can focus perfectly well with the optical finder and don't own a bag full of R Lenses. Lastly, I just don't like flat homogenised images which look more like they were shot in a Canon or Nikon than ever before. But it seems the M upgrade was not aimed at those like me........

 

Nearly exactly my impression too. We will see how it develops with the final FW and the production M. No final decision before that point.

 

Bill

[k-hawinkler]

Hi Paul,

 

Thanks a lot for your explanations. I have to agree.

 

Often, when I shoot the same scene with different cameras but all with Leica lenses, the M9 images stand out with vibrant colors and require not a lot of work if WB and exposure are correct.

 

I don't have any experience with Canon DSLR gear. But I do with Nikons. Often, I am not able to process .NEF with CS6 so that the images look as pleasing as the M9 ones. My best bet for .NEF files, especially from the D800E, is using the latest version of Nikon's Capture NX 2. If I am not mistaken, the Nikon images seem to appear somewhat warmer.

 

I wonder whether I shouldn't work into my workflow a color card for all my cameras, including the M240 when I get it.

 

Thanks again.

[Paul J]

I wonder whether I shouldn't work into my workflow a color card for all my cameras, including the M240 when I get it.

 

The thing about a colour card is colour and tonality work like a spiders web. You change one and it effects the others. You can shrink the gaps in some parts of the web but in other parts those gaps grow at the same time. Creating Noise, banding etc. That is the problem of not being able to adjust the cameras we are speaking of and it ends up being about compromise and spending a fair amount of time to learn what that compromise is and how it fits into your style. Also it's not just a question of the accuracy of the colour but the brilliance of the colour in how it works across the entire tonal range and gamut. The M at least seems has more latitude for adjustment but how much one can make at the moment is speculation until we see the final production unit.