Aristophanes
Well-known
Ever since Hermes took controlling shares of Leica in 2000 in its dark days, the brand has mostly been a luxury brand. The key here is "well-heeled". Just looking at their sensors Leica has been a marginal player in moving IQ forward, and is more interested in making a status symbol. So we see videos of people with a file. Its the premium stance of the brand that matters.
Leica feeds off the status buzz now rather than the real workmanship of making images. Those who were teethed on the M3 and M6's, Leica left you long, long ago.
"Stephan Schulz: Leica had a long history with professionals throughout the 20th century. Most of these professionals were photojournalists. But we realized that in the 21st century, the image of our brand was becoming weaker because Leica was no longer as strongly represented among professionals.
In today’s professional market, photojournalists are a rare breed, and they are no longer as well paid as in previous generations. For a high-end brand that charges premium prices, this secular trend needed to be addressed."
http://www.forbes.com/sites/marcbabe...its-own-terms/
Leica feeds off the status buzz now rather than the real workmanship of making images. Those who were teethed on the M3 and M6's, Leica left you long, long ago.
"Stephan Schulz: Leica had a long history with professionals throughout the 20th century. Most of these professionals were photojournalists. But we realized that in the 21st century, the image of our brand was becoming weaker because Leica was no longer as strongly represented among professionals.
In today’s professional market, photojournalists are a rare breed, and they are no longer as well paid as in previous generations. For a high-end brand that charges premium prices, this secular trend needed to be addressed."
http://www.forbes.com/sites/marcbabe...its-own-terms/
__--
Well-known
Actually, in the in the article that you link, Schulz concludes the opposite on what you state about Leica not moving image quality forward with the Leica S.
My own experience is with the M-9 and the M-Monochrom. Forgive me for quoting what I have written elsewhere quite a few time but for me, it remains the best statement on the image quality of these cameras: n February 2103, In was pushed over the edge to buy a new M9-P, on which I got a promotional price that was only a little more than that of an M-E, by statement by Charles Peterson, a Seattle photographer who had showed me his personal color work that I found to be outstanding; he wrote, more articulately than I could:
I do think that the higher ISOs on the M9 are vastly underrated, and in general much prefer the image quality of the M9 to the M240. The M9 (and Monochrom by de facto) IMO are truly two of the most unique digital cameras out there when it comes to the quality of the image. Not the "best" on paper but they have a look, an "umami" as the Japanese might say, that no other 35mm digital camera, comes close to. And that is what I have found in using these two cameras — not that they are status symbols, as you state.
MITCH ALLAND/Potomac, MD
Download links for book project pdf files
Chiang Tung Days
Tristes Tropiques
Bangkok Hysteria
Paris au rythme de Basquiat and Other Poems
Aristophanes
Well-known
Sensors have completely quantifiable performance based entirely on their ADC gathering pure photovoltaic information from their wells. There is nothing special about any sensor. It simply registers an electronic signal from photons. If you want to wax on about Leica's algorithms...go right ahead. The M-9's sensor cannot in any quantifiable way gather or impart any more information than any other CCD sensor of its era. It's a 14-bit linear sensor. There's no "umami" in bits! That sensor's IO is absolutely dreadful which speaks to noise and suppression. That also explains the poor DR. Leica was scraping the bottom of a desperate Kodak barrel for that sensor.
No Leica, not even the S-series, has a truly up-to-date sensor. They are all at least 1/2 generation behind Sony's fabs right now.
http://www.nytimes.com/2012/09/20/fashion/leica-cameras-favored-by-celebrities.html?_r=0
Leica has an unabashed product placement system for music videos, fashion catwalks, New York gallery parties and the like. Leica is notorious for sponsoring news stories. It's no secret they subsidize and gift cameras to visible events and trendsetters; their own guerrilla marketing. That was the whole point of the Hermes investment; to make Leica a status symbol:
http://www.imaging-resource.com/new...hermes-leica-m9-p-sometimes-beauty-is-only-sk
Please note that it is Leica's marketing that repeatedly uses the word "status" when promoting their products. Leica has entered into dozens of deals to promote only luxury placement of their product:
http://investing.businessweek.com/research/stocks/private/snapshot.asp?privcapId=4509058
I've been to one of these presentations. They don't even make a pre tense of selling you a camera. Leica is entirely positioned as a luxury accessory.
The schmucks who purport to buy Leica's for their IQ, which is definitely not supported by the sensors...not even close (Ming Thein's S2 review is a killer)...are only giving street cred to the status brand effort ("See. The old school RF geeks still like them"). When the RFF crowd reach up to pay luxury prices, it only feeds the Leica marketing effort to give the stars who use Leica the impression they are continuing the workaday, photojournalist credibility that built the brand. It's the mystique and nostalgia that sells. You are helping them. They've got you doing their work.
As Mr. Schultz said last year, Leica left photojournalism behind. One could argue that Leica is working to leave the RFF crowd behind as well. Fuji's taking note.
The M3 was built tough because it had to be tough. It went to war! The T's images are pedestrian given the price. It goes to fashion shows. It is a fashion show!
Leica was never a value brand save for their outstanding durability; a rugged, enduring photographic tool. The T looks like it was made out of aluminum block so they could make a marketing video showing it made out of an aluminum block. This is engineered marketing wrapped around electronics that are as average as a sub-$700 Nikon.
Quite the change:
http://petapixel.com/2014/01/08/think-camera-old-leica-m4-ad-brags-cameras-lack-technology/
Richard G
Veteran
I got the same two cameras as Mitch. Love them both. The Monochrom hardly fits the idea of image conscious product placement. Almost the opposite.
YYV_146
Well-known
Just to be fair, the M9 was at a nice place in the sensor world when it first came out, and so is the M240 if we are talking about the original release date (2012), not the time when most people actually got theirs.
Leica also seems to have a handle on Jpeg engines, something Sony distinctly lacks. I use Sony bodies for all my work, the jpegs are so terrible that I only use them to show people quick previews. The X cameras have good jpegs, so does the M9 - if AWB and exposure happens to be right, that is.
My experience with the S system is limited, but I don't like medium format back in general. CCD backs are basically dinosaurs, any decent full frame CMOS body will eat them for breakfast in terms of dynamic range and tonal range. I don't think Leica in particular has done anything more or less than the rest of the MFDB world - Hasselblad and Phase have bigger sensors, but the outdatedness is the same.
The new crop of CMOS medium format bodies is much more interesting, though, and Leica seems to have a new CMOS S in the works. Actually if I ever decide to step into digital medium format it would probably be with Leica, since I hate 4:3 with every single fiber in my body
Leica has its technological advantages, both in optics and electronics...
__--
Well-known
No one said the M9/M-Monochrom sensor had any mystical quality. Obviously there is more than the sensor itself that determines the image quality of a digital camera. One can also go further than you do and say that the innards of an M9 are wired together the way an experimenter would do if he put together a camera in his garage: i.e., virtually no system integration or modular construction, like that of Nikon or Canon, not to speak of Sony. However, the whole Rube Goldberg assembly (if you will) of the M9/M-Monochrom cameras, including the sensor, the AFC and everything else inside these cameras produces an image quality that is unique, "not the best on paper," as stated in my quote in post #145 above, "but better than any other digital camera comes close to."
You also refer to the noise and low DR of the M9 sensor but, using the technique described in this thread, because of the color rendition, some of us "schmucks who buy Leicas for their IQ" (as you so elegantly put it) also find the M9 unique and excellent as a camera for low-light night photography — but I guess we're mesmerized by the Leica advertising, or seeking status, and should be concerned that we like an "outdated camera."
MITCH ALLAND/Potomac, MD
Download links for book project pdf files
Chiang Tung Days
Tristes Tropiques
Bangkok Hysteria
Paris au rythme de Basquiat and Other Poems
YYV_146
Well-known
I don't think he's talking about the assembly. And no, there is no magic with any combination of electronics inside a digital camera - cameras are sensors, and everything else just helps the sensor get the correct data. You can use a processor to better read the data, but at the end of the day, the IQ of a body is both defined and limited by the sensor and its level of sophistication.
You may prefer the results of an M9 to another camera, just as I may prefer to drive a vintage car instead of buying the latest one off a production line. But the thing is, the new car is going to be "better" - better fuel efficiency, better for the environment, most likely more comfortable and practical.
DR doesn't mean high iso performance. The M9's high iso performance obviously is dismal by full frame standards - 1600 looks like 8000 on the A7, and any of the Canikon flagship bodies can easily outperform it by 4-5 stops. But the dynamic range of the M9 sensor is also pretty bad, a by-product of using 2007-2008 CCD tech. At base ISO you can't get more out of M9 files than the files of most APS-C cameras. I did a pretty comprehensive dynamic range test between the M9 and NEX-7 in HK two years ago, when I was contemplating buying one. The NEX-7 has a 1.5 stop advantage in the shadows from 100 to 1250, afterwards both cameras have unacceptable DR for critical work.
Of course, the M9's jpeg color rendition is nice, and compared to other digital rangefinders of its day IQ is far superior, so I don't see why people don't like it. But is the M9 better than any other production full frame camera? No. With all due respect, people who still prefer to work with the M9 today should spend a few weeks with the M240 and carefully evaluate the images between the two cameras. I myself have not noticed any difference between M9 jpegs and M240 jpegs - Leica obviously made a point to cook the files as closely as they could. And the M240's files, while having the high dynamic range and good iso performance of CMOS sensors, do have a non-CMOS look to them.
The M9 is no doubt a "unique" and "sufficient" camera for night photography, but is it an excellent one, especially compared to an A7 or A7r? No. I checked the files in that post carefully, and I believe the technique only cut back on chroma noise. Overall performance still loses to the M240 by a stop, and frankly if I really made an effort, I can clean the A7's ISO 10000 to look as good as those. Nice pictures in the thread, though...
The A7's iso 5000, pushed over a stop to about ISO 12800, and processed to reduce overall graininess. The entire area was lit by ten dim streetlamps in a snowstorm. With an M9, the same shot could only have been made by pushing more than 2 stops over the 2500 maximum...
__--
Well-known
Thanks for the kind words on my night pictures with the M9. But, frankly, I find these car analogies for cameras tiresome and irrelevant...
On the M240, it does have less noise at high ISO and more dynamic range than the M9 but because of its superior color rendition, from what I've seen, I prefer night shots with the M9 to those of the M240 that I've seen. In fact, from what I've seen, I am not happy with the color rendition with the latter camera — and don't agree that everything can be fixed in post-processing, as many M240 owners claim. "fotografz" (Marc Williams), whose color judgment I trust (he's particularly concerned about skin tones), tried the M240 for two weeks and reached a negative conclusion on its color rendition. It looks like the M240 has issues with red-yellow rendition and I am not sure that further firmware fixes will improve its color rendition; but I don't want to start an M9-M240 war here.
On the Leica-T, which after all, is the subject of this thread: at this early stage it looks like it doesn't have the M240 color issues.
MITCH ALLAND/Potomac, MD
Download links for book project pdf files
Chiang Tung Days
Tristes Tropiques
Bangkok Hysteria
Paris au rythme de Basquiat and Other Poems
Aristophanes
Well-known
14 but liner read-out is 14 bit liner readout. It matters not which camera body or brand to which it is attached.
There can be nothing "unique" to it save for the ADC or in-camera algorithms. Kodak's capacity for the ADC is dismal and it shows in the DxO scores.
Simple point: The M9 gets less information to the in-camera processor than a much much cheaper APS-C. All those Leica optics transmit information that doesn't make it to the picture taker, especially in low-light. The physics of that sensor and its ADC are definitive: you get less.
There is no "colour rendition". There is bit-level dynamic range which is quite poor.
That same FF sensor could be hacked into any camera and produce identical results. It is a sub-standard sensor compared to the Sony fab output. The Leica M9 sensor is magnitudes worse than the Sony's and worse than top-end APS-C sensors of the same era. The T looks to also be lagging behind in sensor tech.
If you do not understand this you not understand digital cameras. They are just a sensor surrounded by a small computer that lets you see what a software programmer wants you to see.
Richard G
Veteran
Hmmm I must re-read 'The Clouds.'
__--
Well-known
mfunnell
Shaken, so blurred
And hugely more succinct than the po-faced literalist response I decided not to post. My hat's off to you, sir!
Sparrow
Veteran
mfunnell
Shaken, so blurred
Can you not at least allow it as a "gentleman's third"? After all, they seem quite popular in the political classes.... strangely, that came in a poor third too
...Mike
dcsang
Canadian & Not A Dentist
I would tend to agree here.
Digital files are digital files. The rendering from the lenses is what makes the image look the way it does in my opinion (I say opinion because I do not have scientific proof to prove that). Put that same lens on any sensor and it should, given similar structure of the sensor (i.e. if one sensor has micro lenses then the other sensor should as well etc. ), provide a similar image. This is why the "leica glow" from a 35mm Summilux pre-ASPH at f1.4 exists not just on film but also in digital regardless of which camera that lux is mounted on (M9 or A7 or any other full frame digital camera).
This is also why a digital image will look like a digital image until we manipulate it to "look" like film. Even film looks different b/w printing traditionally and scanning/post processing/printing.
At the end of the day, you end up with what YOU want - You can manipulate your files (or negs) to create art but the science in how that file (or neg) was created in the first place is limited to, as Aristophanes said, what a software programmer (or film maker) wants you to see...
Cheers,
Dave
Aristophanes
Well-known
Really?
They made a film entirely about it which was paid for entirely by Leica and submitted to film festivals.
At those festivals Leica had a number of illuminati wearing their Leica's prominently on show.
The entire Leica marketing machine was created by Saatchi and Saatchi, an ad agency.
This is how Leica promotes its products:
http://www.dailymail.co.uk/tvshowbi...amera-store-opening-months-away-children.html
Leica throws exclusive launch parties to promote its products in the luxury segments. They are nothing like an Apple event which is open to app programmers and anyone who can score a ticket. Leica is all about controlling the image of the purchaser as being from an elite. Apple caters mostly on volume (see their latest financials, esp. unit sales) of a middle and upper middle class consumer. Leica is positioned totally at the 1%.
That is where your money is going. Leica sponsors fashionista users to get people from RFF to buy into the aura.
I think Leica make outstanding lenses. I also think that everyone else is making outstanding lenses. I think the touch screen on the T is a great idea. I think it is not exclusive to Leica and will be the norm in a very short time. Leica emphasizes extreme handicraft manufacturing, but inside the electronics are Radio Shack and the sensor is a $400 Nikon DSLR equivalent.The T is a design showpiece only because that is the last place Leica can compete; that and getting Seal to sing at their events.
willie_901
Veteran
Let's consider a sensor has two roles.
1. convert light energy into electrical charge
2. model the spatial and frequency aspects of the light exiting the lens rear element.
The electronics are photodiodes controlled by transistors. Their transistors role is to make sure things happen in the proper order and to provide a low-level of amplification. The photodiodes convert the light waves' electrical amplitudes into electrical charge. There is no fundimental difference between the photodiodes in a CCD sensor and a CMOS sensor. The electrical charge in the photodiodes is just energy stored in a semiconductor potential well. Instead the difference between CCD and CMOS performance has to do with the manufacturing processes between the CCD and CMOS integrated circuits.
The manufacturing and design process does affect the devices' efficiency. CMOS circuits are more efficient. By any objective measure the analog signal-to-noise ratio and dynamic range of the Kodak technology used in the M8/M9 series is inferior to CMOS technology available in a parallel time line. This is not Leica hating, it is not inflammatory. It is simply a state of nature based on competent, honest measurement. I don't make the news here, I just report it.
Which brings us to the the sensor's second role.
The evaluation of the raw-data SNR and DR is trivial because it is objective. The aethestic value of the rendered image, is another story altogether. On one hand, aesthetic judgements are subjective. Can arguments about CCD vs CMOS image quality can ever be resolved? In this case we have to rely on the observations of skilled photographers with integrity. I personally evaluate subjective evaluations based on the photographers' motivations. In general, human nature intervenes, so I pay more attention to the reports on subjective aesthetics from those who only have a stake in the quality of their work. Mitch is one example, but there are many, many others. After the M8 IR contamination debacle I ignore photographers with any personal ties to or commercial dependencies on Leica.
In the case of the M8/M9 there are numerous photographers who only care about the image quality in the scope of their artistic activities. These photographers assert the M8/M9 rendered images are aesthetically superior to other platforms.
This appears to create a dilemma. The CCD technology is inferior when measured objectively, but the rendered result is considered superior by many.
This dilemma does not exist because the components in-between the lens and the photodiodes play an important role.
The advantages and disadvantages of anti-aliasing filters are no longer controversial. AA filters are not required for aesthetically acceptable image rendering. I choose to ignore the IR filter.
This leaves the color-filter, micro-lens array. The CFA is just as important as the sensor signal-to-noise ratio and dynamic range. Unfortunately the characteristics of the CFA are much more difficult to measure and evaluate.
Consider two very different CFAs. One is made to minimize design, material and production costs. The other is made to provide spatial and frequency information that directly corresponds to the Bayer-reconstruction mathematical model. Is it unreasonable to assume a design where each micro-lens collects as much light as possible (regardless of it's physical location in the array) and filters out all frequencies except the frequency the Bayer model, costs more to design, source and manufacture than a design with less ambitious goals?
The value-aded aspect of the M8/M9 design is not the photodiode-array portion of the sensor. Instead it is found in the CFA optics.
Of course other brands can employ high-quality CFA optics along efficient CMOS electronics. In fact, these brands exist. But in some cases other brands are tempted to sacrifice frequency-filtering specificity for signal level. A green micro-lens that passes 15% more light from non-green frequencies generates more charge (signal). The SNR and DR are increased. However the color rendering may be inferior because the green-frequency information the Bayer model requires is contaminated. The data is not what the Bayer model expects. The rendering is compromised. The rendering is aesthetically superior.
1. convert light energy into electrical charge
2. model the spatial and frequency aspects of the light exiting the lens rear element.
The electronics are photodiodes controlled by transistors. Their transistors role is to make sure things happen in the proper order and to provide a low-level of amplification. The photodiodes convert the light waves' electrical amplitudes into electrical charge. There is no fundimental difference between the photodiodes in a CCD sensor and a CMOS sensor. The electrical charge in the photodiodes is just energy stored in a semiconductor potential well. Instead the difference between CCD and CMOS performance has to do with the manufacturing processes between the CCD and CMOS integrated circuits.
The manufacturing and design process does affect the devices' efficiency. CMOS circuits are more efficient. By any objective measure the analog signal-to-noise ratio and dynamic range of the Kodak technology used in the M8/M9 series is inferior to CMOS technology available in a parallel time line. This is not Leica hating, it is not inflammatory. It is simply a state of nature based on competent, honest measurement. I don't make the news here, I just report it.
Which brings us to the the sensor's second role.
The evaluation of the raw-data SNR and DR is trivial because it is objective. The aethestic value of the rendered image, is another story altogether. On one hand, aesthetic judgements are subjective. Can arguments about CCD vs CMOS image quality can ever be resolved? In this case we have to rely on the observations of skilled photographers with integrity. I personally evaluate subjective evaluations based on the photographers' motivations. In general, human nature intervenes, so I pay more attention to the reports on subjective aesthetics from those who only have a stake in the quality of their work. Mitch is one example, but there are many, many others. After the M8 IR contamination debacle I ignore photographers with any personal ties to or commercial dependencies on Leica.
In the case of the M8/M9 there are numerous photographers who only care about the image quality in the scope of their artistic activities. These photographers assert the M8/M9 rendered images are aesthetically superior to other platforms.
This appears to create a dilemma. The CCD technology is inferior when measured objectively, but the rendered result is considered superior by many.
This dilemma does not exist because the components in-between the lens and the photodiodes play an important role.
The advantages and disadvantages of anti-aliasing filters are no longer controversial. AA filters are not required for aesthetically acceptable image rendering. I choose to ignore the IR filter.
This leaves the color-filter, micro-lens array. The CFA is just as important as the sensor signal-to-noise ratio and dynamic range. Unfortunately the characteristics of the CFA are much more difficult to measure and evaluate.
Consider two very different CFAs. One is made to minimize design, material and production costs. The other is made to provide spatial and frequency information that directly corresponds to the Bayer-reconstruction mathematical model. Is it unreasonable to assume a design where each micro-lens collects as much light as possible (regardless of it's physical location in the array) and filters out all frequencies except the frequency the Bayer model, costs more to design, source and manufacture than a design with less ambitious goals?
The value-aded aspect of the M8/M9 design is not the photodiode-array portion of the sensor. Instead it is found in the CFA optics.
Of course other brands can employ high-quality CFA optics along efficient CMOS electronics. In fact, these brands exist. But in some cases other brands are tempted to sacrifice frequency-filtering specificity for signal level. A green micro-lens that passes 15% more light from non-green frequencies generates more charge (signal). The SNR and DR are increased. However the color rendering may be inferior because the green-frequency information the Bayer model requires is contaminated. The data is not what the Bayer model expects. The rendering is compromised. The rendering is aesthetically superior.
aizan
Veteran
the perception of color is not absolute, so the proposition that the m9 has a special "umami" is moot. it both does and does not.
GaryLH
Veteran
The only Leica that I have ever seriously thought of buying was the monochrome. While the sensor may not have been the best at the time it was released.. To me, the idea of a monochrome only sensor maybe boohoo'd by most and only a small portion of the public would ever appreciate it.. Leica to me pushed the envelope. Though not the tech one, but the going outside the norm one. The components maybe exactly the same as a regular m9 (except no Bayer array sensor) but they had to put money into firmware for this to happen.
The Sigma foveon based sensors in the dp Merrill's are what I use for monochrome work now. I hope Sigma will come out w/ a csc version one day. Until then, as if said in other threads... Fuji or Leica T or whoever would just make a monochrome csc, I would be very happy.
Gary
The Sigma foveon based sensors in the dp Merrill's are what I use for monochrome work now. I hope Sigma will come out w/ a csc version one day. Until then, as if said in other threads... Fuji or Leica T or whoever would just make a monochrome csc, I would be very happy.
Gary
Aristophanes
Well-known
Hah!
Nevertheless, the signal is discrete and finite. It cannot be more or less than 14-bits. The transmission of the signal is 100% measurable and complete, and can be exactly compared to similar signals from other sensors.
There is a distinct red dot filter over some people's eyes.