M9 Sensor Qualities

[mlu19]

Leica brand is a status symbol, a luxury brand. People who own one are proud and happy users. I see that the analogy between cameras and cars have gotten old. How about let us compare cameras and timepieces. All cameras take pictures; all watches tell time. Both industries have well-established markets for consumer/professional/luxury segments.

I'm happy to see where Leica has positioned itself in the camera industry. In my opinion, it is apples and oranges to compare Leica to any other camera brands on the market. OP is right in regards of the sensor technology and price point. At the end of the day, you buy what you can afford and what makes you happy user.

 

[Corran]

A Nikon D5300 will capture more data under all circumstances than the M9. Much more, despite sensor size.

Who cares?
Then I can assume you've pre-ordered the new Pentax/Hasselblad/PhaseOne camera, as it'll likely blow away any 35mm-style camera in technical IQ? Because you know, that's all that matters, right?

 

[Mark C]

I know it might seem weird, but I look at the pictures. If you break away from the tech mindset you start to notice some really interesting things. That is if your goal is to take pictures not capture data though.

 

[Mcary]

The data is unequivocal, empirically tested, and definitive.

A Nikon D5300 will capture more data under all circumstances than the M9. Much more, despite sensor size.

The digitization of photography has homogenized the capture of photons.

Technical IQ cannot be "boring". It's all just data points quantified. There is no "character" to a sensor. You either have a certain strength Bayer array or AA filter or you do not. You either have a certain signal to noise ratio or you do not. They are all finite and perfectly quantifiable right down to photon and electron counts. It really is just math.

The sad part is that a person with a much lower priced consumer camera system can replicate your work, and even exceed it on pure data capture. Maybe you have a technique or shooting style that you like and think the Leica RF is worth paying a $7,000 premium for, but it is definitively not giving you better data input than the D5300. If your job is to give clients the best files you can, the Leica is not competitive. This is why Mr. Schulz has spoken about why Leica is no longer in favour amongst photojournalists and why Leica as a brand is actively not pursuing the same market you seem to be working in. You are defending what Leica's main spokesperson is not endorsing as Leica's motivation in product development or use anymore. You can be as fervent at defending your work as you want, but your own brand's spokesperson is actively moving away from your own position.

Leica mechanical engineering= excellent.
Leica lass = excellent.
Digital files = mediocre at best.
Value of all those = not very good.

How do you know if I do or do not own a Leica?

OK we give up you win the Nikon D5300 is a much better camera than the Leica M9.

Everyone please refrain from pushing or shoving and form a neat line to sale your M9s and other Leica digital gear in the Classifieds.

 

[uhoh7]

Leica vs not-Leica always brings out the attitudes on all sides.

It's like putting a bowl of mud in a sandbox where some 3 year olds having been playing nicely...

 

[sig]

If only Leica could use the best sensor there is, then everybody could be happy.

Remember that your cameras (both Nikons and Leicas) can not be offended by what people write on ze interwebs. So at least you do not need to worry about what to tell the 'kids'

 

[uhoh7]

I love all the cliches--on both sides

"leica is for rich posers"
"Sony has too many dials" (maybe that's why Leica has such primitive focus aids with the 240--no dials to move the patch)

"DXO says.."
"don't tell me what my Leica can't do, you troll"

and my favorite here: "no one should get a Leica for the image quality"

Now some of the suspects don't have the cameras, but a number do or did, which makes these declarations especially poignant.

"I had a M9, and I can tell you it sucked."

"I had a A7 (and shot some really nice stuff with it) but I could not handle the interface."

In reality I screw up the M9 interface just as often and the Sony--or more so. I leave the caps on. The dial comes off 'A' to "B", I shoot with the SEM 21 coded as a 90 f/4. etc.

With the A7 I just forget my mode, P or S.

I use the M9 in spite of the RF, not because of it. The Sony is alot better for high speed focus.

But it's the M9 I always take these days. Because I prefer the images.

The exception at the moment is shooting birds. I have a bunch of feeders and like to shoot songbirds in flight. For that I use 180, 300 or 400, and 5000/sec on the A7 and it works pretty good.

 

[TennesseJones]

Do still quite like the frame lines though, in those Leicas...

Anyway, will now sell mine forthwith.

 

[hepcat]

The data is unequivocal, empirically tested, and definitive.

A Nikon D5300 will capture more data under all circumstances than the M9. Much more, despite sensor size.

The sad part is that a person with a much lower priced consumer camera system can replicate your work, and even exceed it on pure data capture.

Ahhh... I think I understand the problem now... you're confusing the terms "image quality" and "quality image." They are not synonymous.

The truth is that a competent photographer makes quality images with whatever equipment he has at hand. That work may be copied in concept, but likely not replicated no matter what the "image quality" of the gear.

The millions of mediocre images stored on Flickr that were taken with "high image-quality" gear and the occasional "quality image" taken with a camera with less than state-of-the-art "image quality" proves that you can't minimize the role of the photographer in image making. That merely confirms that the photographer makes the "quality image," not the camera.

 

[willie_901]

The fundamental challenge in digital imaging is to model continuous phenomena with discrete data. The model in our case (with a few exceptions) are Bayer interpolation algorithms.

The best possible outcome depends on two things. Does the data map onto the model in a one-to-one fashion? Does the uncertainty in the data (the noise, or the errors) compromise the model's estimates?

The color-filter array maps the data onto the model. All brands do this equally well because the constructing the required RGB micro lenses spatial arrangement is straightforward.

The the uncertainty in the data is another story.

In photography people describe to the uncertainty in the light-amplitude estimates as noise. When the signal is abundant, only the noise limits the SNR – which in turn limits the DR. Some noise is unavoidable (shot or photon noise). Some noise depends on circuit design and the efficiency (quality) of the circuit components. High uncertainties in the data leads to high uncertainties in the model estimates. Criticism of M8/M9 sensors is based on their relative SNR performance when the signal level is low. In low EV situations, reasonable shutter times for hand-held and. or subject motion use limit the signal. Then the electronic noise level becomes important

Other sources of uncertainty are rarely considered. This doesn't mean they aren't important. Ineffective color-transmission selectivity in the micro-lenses generate errors in the data as well. The real (state of nature), but unknown amplitude of the green, red or blue light is contaminated by contributions from other visible light frequencies. These errors lead to errors in the color rendition. The Bayer interpolation models are not compatible with contaminated photon-amplitude estimates. The models requires the each pixel to contain only amplitudes from narrow-frequency ranges of R, G, or B light.

It is easy to measure and compare SNR.

Measuring and comparing errors in Bayer color contamination is more difficult. As far as I know such measurements are limited to academic journal articles.

A fair judgement of final image quality should include all noise sources. The IR contamination that plagued the early days of the M8 could not be solved by clever rendering solutions. The data was not compatible with the Bayer model. The only solution was to remove the contamination.

Likewise, color rendering problems caused CFA color contamination can not be eliminated by post processing methods. The impact of errors due to IR contamination are obvious. Some black objects turn purple. Foliage color does not render well. Skin color is off. The impact of errors due to visible light contamination are more difficult to analyze. However this doesn't mean they are insignificant.

Sensors can have character.

 

[Mcary]

Realize that things like DXO ratings and other such things play an important part in deciding what camera to use/not use for many people but in all honesty stuff like that just bores the crap out of me.
My reason for shooting with a digital Leica, M8 in my case is I enjoy using it, sure it has numerous issues and short comings but none of that changes the fact that I enjoy using it which for me is as much or even more important the technical image quality. Of course this comes from a person who's happy with the quality of results he gets from old out dated 60 year old lens, so what do I know 🙂

 

[Aristophanes]

OK we give up you win the Nikon D5300 is a much better camera than the Leica M9.

I did not say that.

The sensor is superior. It's about the delivery of data points. Right now the superior sensor is found in the Nikon. Why it is not found in the Leica requires deconstructing the Leica business model.

The issue is why Leica uses sub-par sensors in flagship cameras that have superlative engineering elsewhere, especially in optics. price and value factor in. Thank-you to willie_901 reminding about the M8 sensor issue.

From the history of the company it is indefensible. Loyalists of the older photojournalist legacy are no longer the target market. They are now a marketing tool to sell to the fashionistas. So says Leica (many times over they've said it).

I think it very clever of Leica to ignore the sensor quality issues and basically use the older RF crowd as street cred marketing at full price, while stocking the Dubai airport gift shop with their products. Leica's corporate brain trust does not have to put their money into sensors on par with Nikon's average consumer product because the street cred RF crowd will buy regardless. Fuji, OTOH, puts huge capital into developing its own sensors. Nikon has been a partner with Sony. It shows. Nikon, Sony, Fuji, and even Canon (slipping a bit) have, at this moment in digital photography, decisively out-engineered Leica. from my perspective, the emergence of the all-new mount Leiac T with a recycled Sony APS-C sensor module is the icing on the cake.

But of course, Leica's credibility is less about engineering and more about milking an uncritical street cred heritage coupled with workmanship videos of someone buffing aluminum. The really hard part, the capital investment in photodiode imaging, the uttermost core of image-making in digital, is subsumed. It is Blackstone-driven genius brand and luxury marketing and product pricing at its best.

 

[hepcat]

I did not say that.

The sensor is superior. It's about the delivery of data points. Right now the superior sensor is found in the Nikon. Why it is not found in the Leica requires deconstructing the Leica business model.

The issue is why Leica uses sub-par sensors in flagship cameras that have superlative engineering elsewhere, especially in optics. price and value factor in. Thank-you to willie_901 reminding about the M8 sensor issue.

Except that it won't work in the sensor-to-flange distance of the M series. IIRC one of the technical challenges that had to be overcome for the M sensor is the angle of entry of the light with such a close flange-to-sensor distance.

You still haven't really said what your beef is... you've danced all around the marketing, tried to slam the ability of the sensors, and said they're too expensive for their tech specs. Ok, so what if they are? What's your point? What do you expect out of this? Do you expect I should dump my M9 and get a Nikon? Should I demand a refund from Leica because they're too expensive?

Sorry, but after all these pages, I still don't know what your point really is.

 

[Mcary]

I did not say that.

The sensor is superior. It's about the delivery of data points. Right now the superior sensor is found in the Nikon. Why it is not found in the Leica requires deconstructing the Leica business model.

The issue is why Leica uses sub-par sensors in flagship cameras that have superlative engineering elsewhere, especially in optics. price and value factor in. Thank-you to willie_901 reminding about the M8 sensor issue.

From the history of the company it is indefensible. Loyalists of the older photojournalist legacy are no longer the target market. They are now a marketing tool to sell to the fashionistas. So says Leica (many times over they've said it).

I think it very clever of Leica to ignore the sensor quality issues and basically use the older RF crowd as street cred marketing at full price, while stocking the Dubai airport gift shop with their products. Leica's corporate brain trust does not have to put their money into sensors on par with Nikon's average consumer product because the street cred RF crowd will buy regardless. Fuji, OTOH, puts huge capital into developing its own sensors. Nikon has been a partner with Sony. It shows. Nikon, Sony, Fuji, and even Canon (slipping a bit) have, at this moment in digital photography, decisively out-engineered Leica. from my perspective, the emergence of the all-new mount Leiac T with a recycled Sony APS-C sensor module is the icing on the cake.

But of course, Leica's credibility is less about engineering and more about milking an uncritical street cred heritage coupled with workmanship videos of someone buffing aluminum. The really hard part, the capital investment in photodiode imaging, the uttermost core of image-making in digital, is subsumed. It is Blackstone-driven genius brand and luxury marketing and product pricing at its best.

Shame on them! bad Bad naughty Leica!

 

[Aristophanes]

Shame on them! bad Bad naughty Leica!

It's all about paying Blackstone back.

 

[rivercityrocker]

It's all about paying Blackstone back.

WHY ARE YOU HERE? WHAT ARE YOU TRYING TO PROVE?

Does trolling give you a boner? Is this how you attempt to improve your self esteem? If it is, it's a pathetic way to live your life.

We get it. You think Leica is overpriced junk. Don't buy one. IGNORE LEICA CAMERAS AND MARKETING.

It's as simple as that.

 

[mlu19]

It's all about paying Blackstone back.

In your perfect world, what would you like to see from Leica?

 

[MCTuomey]

But of course, Leica's credibility is less about engineering and more about milking an uncritical street cred heritage coupled with workmanship videos of someone buffing aluminum. The really hard part, the capital investment in photodiode imaging, the uttermost core of image-making in digital, is subsumed. It is Blackstone-driven genius brand and luxury marketing and product pricing at its best.

Uh ... well then don't buy a Leica if their marketing practices turn you off. One point: it's probably better not to presume that your rationalization or prioritization of value matters to anyone but you, or try to convince anyone that it should, since no one else need share your criteria for choosing what camera to purchase and use.

For my part, I think it's perfectly fine for an enterprise, especially a relatively small one in a sea of larger competitors, to source its engineering and parts to specialists. Done all the time, I believe.

 

[Takkun]

I'm finding myself agreeing with both sides here, though with less vitriol and more ambivalence.
Yes, Leica bodies are frustratingly expensive for the sensor in them. Yes, its aggravating to see a precision machine sold on celebrity endorsements and past glory, and subsidized by wealthy non-photographers.

But as someone else elaborated in technical details, a sensor may be a sensor, but the chain of processing does differ from model to model. Somehow I doubt all those people claiming a unique color rendition from the CCD models, myself included, aren't all totally deluded.
And any photographer who shoots with some degree of frequency can attest that haptics make getting the shot easier, or at least more enjoyable. The mention of image quality versus quality image reminds me of when I worked at a bicycle shop: It baffled me why non-racers would sacrifice any semblance of comfort for speed. I understand that when you sink several thousand dollars into a piece of machinery that one wants the best deal for their money, but there's different ways of quantifying the value.

There's also a lot of talk about 'critical work' here. That varies wildly among photographers and their work and professions. Why don't press photographers shoot MF, and so on? You can take an argument or generalization to absurd ends. I'll agree that when it comes to daily press photography (where getting something to take back to the newsroom is what's important) or commercial work (where file versatility is key), no, the M system's output isn't going to be competitive. You know what? Those pros aren't going to be shooting with a rangefinder. Or even probably a D5100 or whatever was mentioned here.
So when it comes to creative photography, all bets are off. Yeah, I wish I could afford a brand new M240, but I also wish I could afford a Lotus. Or any car for that matter..

On a personal note, I'm on rivercityrocker's side here, as someone who shoots a lot of concerts. I'm giving myself a little more work to do in post than I did with my Nikon, but with a smaller kit and a focus system that works best for me, I'm getting a lot more keepers than I ever did. *My* critical work, in this case, is more limited to getting the damn thing to focus and just being still without breaking my back with a bagful of gear. I tried the Fuji X-system, I liked it, it didn't work for how I shoot. I don't pretend that there's something magic about Leica (hell, I only have 1 Leitz lens), it just works for me, and I leave it at that.

It's probably safe to say RFF isn't an accurate cross-section of the Leica-buying public, and most people here like to yap about gear, but are competent and enthusiastic shooters to back it up. I get the feeling we're all a bit tired of Leica's shameless marketing but are thankful that they're producing the cameras we like to use regardless.

 

[Aristophanes]

In your perfect world, what would you like to see from Leica?

Perfect?

How about just a product line at premium prices that is justified?

They should be sourcing sensors at the same level as Nikon.

If Leica was still public (especially with the M8 issue) they'd have had a much more difficult time putting uncompetitive sensors in a premium brand. As it is, the private equity of Blackstone can ignore the RF legacy consumer yet still sell to them secure in having coupled-RF monopoly. Blackstone gets paid by your allegiance by sourcing lower-end sensors with you paying the difference. It's a business strategy. Few auto companies could get away with this in a competitive market.

This allows Leica to drive down the quality of non-internal components (that which Leica has to source from others like sensors, ADC, screens, proprietary algorithms live video, etc.) passing those savings on to Blackstone as profit. They charge $6,000 for the camera but you get a sensor from an consumer $800 model equivalent. then the fanboys start a discussion about CCD colour ignoring the laws of physics playing into the hype. It's a study in marketing by technical ignorance.

Leica appears to be substituting luxury brand product placement based on history more than on contemporary engineering prowess at the sensor level. All Leica purchasers pay an extreme price for a sensor that cannot do the optics credit (same for the S2). This allows Fuji, Nikon, etc. to make much cheaper products with lesser optics that at final resolution actually perform better, most noticeably at higher ISOs and in shadows at all ISOs (with AF and stabilization no less). By the time you add it all up, an M-series shoots stops less capacity than its sort-of competitors.

Therefore we can only conclude that the Blackstone/Leica strategy is to sell inferior quality sensors because image quality based on sensor resolution, DR, SNR, etc is less of a factor than milled aluminum blocks. In some ways, Leica is a fascinating company to watch in this era of 1% inequality debate. it seems to be the epitome of a luxury company that uses its history pedigree to sell products that its own engineering cannot support. Great optics, sub-par sensors.

The price to performance ratio for Leica's products is probably at its worst ever. Since the 1% purchaser Leica is aiming for do not require value, the economic rent (Google it) to keep Leica going is coming from elsewhere.