M9 Sensor Qualities

[NazgulKing]

CCD vs CMOS is a ridiculously old debate but it boils down to many factors.

1. Noise. CMOS has more circuitry and those circuits do add noise, but one can incorporate circuitry that can filter out some of that noise, while also digitising the output pixel wise. This has a tremendous benefit because analog circuits will simply pick up noise along the way.

2. BUT, CCD can potentially have more SNR and DR because of the less sophisticated design (and also simpler to manufacture) lends itself to potentially larger photosite area and one can apply a larger voltage bias and thus a larger voltage swing. But such designs tend to be in the province of non-commercial applications and it is highly unlikely you will ever see them for consumer and commercial imaging.

3. Because of improvements in processes, CMOS has caught up with CCD with regards to consumer and commercial imaging. It has NOT caught up in terms of scientific imaging. The best low light sensors are still CCD and their derivatives, such as EMCCD. Scientific CMOS is still somewhat behind EMCCD when it comes to the dark current.

4. The future is still CMOS because you can do more things with CMOS. The trend has always been to squeeze more into a smaller area and to do more. That trend will continue and we will likely see CCD phased out completely in the far distant future for most applications except the ones that need CCDs most, like a space telescope.

But we can all banter all we want, but it becomes an emotional question and thus any debate will be pointless no matter how many specifications are thrown out. I think if one plans to stake one's entire emotional need on an opinion, by all means do so. Just be prepared that not everyone will accept that opinion and you better have the thick skin to accept that opinion.

 

[Ranchu]

It's pretty clear to me the whole point of the thread is to create an opportunity to talk **** about leica and wind people up, I don't think you can blame people for taking him up on it. I appreciate the information you took the time to share, for sure.

 

[Lss]

I bought a Nikon D3100 two years ago to take on vacation for casual shooting. -- -- Honestly I doubt that you could tell the difference in 8x12 prints from my M9. You might even say the low light shots are better.

It's easy to believe there is very little difference in quality in a blind test. Even the DxO scores are similar, so they should be equal. 🙂

Like I said earlier in this thread, my M8 usually matches or exceeds the output from my RX1R in actual use. There is zero Leica mystique or marketing in it, it is merely the way the cameras perform in my use. For some users, the sensors are interchangeable and it is therefore the sensor quality that matters the most. For others, more factors enter the equation. It all really depends on the individual photographer's way of working.

The Sony RX1R is one of the best cameras/sensors according to DxO (and the RX1 is even slightly better at 5th), while the Leica M8 does not even make it to top 150 cameras/sensors. There is no argument which one is technically the more capable and advanced sensor or camera. I would be happy to have this Sony sensor in my M8 provided it worked with the lenses. I would not have bought the Sony for the same price if it had the M8 sensor.

 

[semilog]

And if they don't break it out on the spec sheets, how do you know it's better? Are you just assuming?

Because except on a test bench, you don't have access to the output before it hits the ADC. The digital data stream is the only thing you can use to make an image.

Here are two spec sheets for currently available, very good scientific cameras (one can, if one looks, also find data sheets for the sensors, but these are solid implementations that accurately reflect what the sensors are capable of).

http://www.andor.com/pdfs/specifications/Andor_Clara_Series_Specifications.pdf
http://www.andor.com/pdfs/specifications/Andor_Neo_sCMOS_Specifications.pdf

The first uses a top-line Sony CCD that has become an industry standard. Nearly every vendor sells a camera with that (very, very good) Sony chip in it. Early iterations of that Sony CCD almost completely displaced the slower, less sensitive, noisier Kodak CCDs from high-end microscopy applications.

The other, newer camera uses a CMOS sensor. Both cameras have similar pixel sizes, similar quantum efficiencies, and (very importantly for this thread) nearly identical spectral sensitivity curves.

But the CMOS camera offers faster frame rates, 4 times more pixels, much lower read noise, and 5X greater dynamic range.

The CMOS camera is a lot better than the Sony CCD. It is absurdly better than any of the Kodak CCDs that I have used in scientific applications. In fact, I still have a rather expensive camera with a Kodak CCD mounted on one of the research microscopes in my lab. I'm seriously thinking about replacing that one with a Sony A7S when it hits the market.

The practical experience of scientists like me who use solid-state image detectors in our research is in lockstep with the sensor measurements that, say, DxO is reporting for consumer imaging devices.

 

[semilog]

One last point.

Totally hilarious to see people who are ready to drop $3k, $5k, or more on Leica's absolutely lovely glass*, because of an incremental performance advantage over, say, a Cosina or Canon prime, then pivoting to argue that sensor performance is photographically irrelevant.

*Not joking about that. Not even slightly. Note that Leica publishes arguably the most detailed specifications for their lenses of any camera company. Only Schneider routinely provides more technical detail. That is because Leica and Schneider are really really proud of their lenses.

 

[Lss]

Totally hilarious to see people who are ready to drop $3k, $5k, or more on Leica's absolutely lovely glass*, because of an incremental performance advantage over, say, a Cosina or Canon prime

I guess I fall into that camp. The price difference has been too much for my budget so far, but I would like to replace my CV 28/1.9 by Leica 28/2 due to its better handling. Glad it makes someone happy. 🙂

 

[Richard G]

Nice touch with Ferris Bueller there, hardly believing the audience is still in the cinema. Still, I've learnt a lot. The M interface and the lenses make up for the deficiencies of the 'old technology.' The OP seems to contend that the whole Leica machine is now one massive confidence trick with no substance. I still maintain that the out of date sensor in the Monochrom was a serious, brave, bold, worthwhile, photographer-devoted venture. What advertising man with sway lets you release a matte black product with invisible branding? The M8 and the M9 prevented a complete defection from Leica by enthusiasts of their lenses and the M way of shooting once you want to embrace digital. I came to this as late as 2012. It is easy to take these things for granted. Before the M9-P my newest to me Leica was an M5 from the early '70s and this was my main camera. What do I care that there is something not quite up to date about the sensors in my current two main cameras? I'm just a fool because I don't have a Nikon/Sony with a really recent technology sensor? The game is called photography, not 'who's got the latest sensor'.

 

[Aristophanes]

One last point.

Totally hilarious to see people who are ready to drop $3k, $5k, or more on Leica's absolutely lovely glass*, because of an incremental performance advantage over, say, a Cosina or Canon prime, then pivoting to argue that sensor performance is photographically irrelevant.

*Not joking about that. Not even slightly. Note that Leica publishes arguably the most detailed specifications for their lenses of any camera company. Only Schneider routinely provides more technical detail. That is because Leica and Schneider are really really proud of their lenses.

Absolutely correct.

All the optical prowess of Leica glass cannot even be discerned on their own sensors. Their glass tests extraordinarily well, but their sensors test poorly, even in spatial resolution. The M240 has ~40% pixel pitch in its advantage as an FF CMOS sensor yet the new Sony 24MP APS-C CMOS have nearly identical profiles, even slightly exceeding in DR. Lesser glass on a Sony sensor equals or exceeds Leica glass on a Kodak/CMOSIS sensor.

Available light photography has moved in extraordinary directions with the new gen of sensors, leaving Leica behind because their chose Kodak and CMOSIS. Of course at low ISOs all take great photos. It's when you ramp up the ISO towards today's new available light opportunities that the Leica isn't making its way around the track. At that point consumer-grade sensors provide more image content to the end user even with lesser optics.

 

[bobbyrab]

As I don't have a digital M to compare actual examples with other digital cameras. So could you clarify for me, are you saying that all digital M cameras with Leica lenses will be outperformed by entry level Sony and Nikon cameras, and if yes, is this in all circumstances or just in low light?
Secondly, would you be able to see this quality difference in print or on screen, or are you only discussing technical data?
Lastly to give me some context with which to asses your judgments worth in what is after all an aesthetic art, have you a link to any good technically cutting edge imaging that you think shows output not achievable with the poorer sensors.

 

[hepcat]

No, I did not generalize at all like that. Nor did I name-call.

In fact, most of your posts have been generalizations painted with a VERY broad brush... and "schmucks" isn't name calling?

One last point.

Totally hilarious to see people who are ready to drop $3k, $5k, or more on Leica's absolutely lovely glass*, because of an incremental performance advantage over, say, a Cosina or Canon prime, then pivoting to argue that sensor performance is photographically irrelevant.

I have to agree with you here. I think that speaks pretty highly of the quality of modern Japanese glass in general, and CV glass in particular. My entire modern lens lineup is CV. The only Leitz lenses I own are '60s vintage.

Absolutely correct.

All the optical prowess of Leica glass cannot even be discerned on their own sensors. Their glass tests extraordinarily well, but their sensors test poorly, even in spatial resolution. The M240 has ~40% pixel pitch in its advantage as an FF CMOS sensor yet the new Sony 24MP APS-C CMOS have nearly identical profiles, even slightly exceeding in DR. Lesser glass on a Sony sensor equals or exceeds Leica glass on a Kodak/CMOSIS sensor.

You've been on and on about the engineering specs and test charts of all these sensors... and in all these pages you've not addressed how significant those differences are in the real world and how easily you can discern the differences in, say, 20x24" prints. So, other than gaining some high-iso performance, and perhaps some increase in DR, other than illustrating the minute differences in test charts, what are the real-world gains you are touting? I am genuinely curious because I ought to be seeing them in images being produced, and I'm having a tough time finding them "out there." If they're as significant as you claim, I'd think they'd be easy to discern. What am I missing?

There's an old saw about the definition of MilSpec (military specifications) that says the definition of MilSpec is "measured with a micrometer, marked with chalk, and cut with an axe." That's where we are with this discussion. The variables that lead to the end product hugely overshadow the incremental image quality input gains that you're touting so strongly.

Unlike the market of even five years ago when the gains were real and signficant for the consumer, the engineering of these devices is at a point today where, for the most part, the improvements are driven by the demands of folks like you who apparently are more concerned with test charts than making images, and the needs of the advertising folks for something new to advertise rather than a real and significant improvement that is useful in the daily use of most consumers/photographers. Are there incremental differences in the sensors? Sure. Are these huge improvements to be made that end-users are going to readily discern? No, not so much any more.

If a viewer can't differentiate the image from a new sensor from an image with an old inadequate M9 sensor, (or heaven forbid actually prefer an M9 image in a blind showing) then the premise of this entire thread is moot.

That said, though, and taking your posts in the thread in their entirety, I should think that the only thing Leica should fear today is Cosina-Voigtlander introducing a FF digital body using the Sony A7 sensor and processor based on the ZI long-base rangefinder body at a price point under $3k. I fear that would entirely deflate the entire used Leica M digital market, and except for the fashion and celebrity market, pretty much kill off the actual user market for new sales in the M-E and M240 as well.

 

[rogazilla]

"That said, though, and taking your posts in the thread in their entirety, I should think that the only thing Leica should fear today is Cosina-Voigtlander introducing a FF digital body using the Sony A7 sensor and processor based on the ZI long-base rangefinder body at a price point under $3k. I fear that would entirely deflate the entire used Leica M digital market, and except for the fashion and celebrity market, pretty much kill off the actual user market for new sales in the M-E and M240 as well."

That will be my dream come true a good range finder on top of the A7 that is 🙂

 

[MCTuomey]

That said, though, and taking your posts in the thread in their entirety, I should think that the only thing Leica should fear today is Cosina-Voigtlander introducing a FF digital body using the Sony A7 sensor and processor based on the ZI long-base rangefinder body at a price point under $3k. I fear that would entirely deflate the entire used Leica M digital market, and except for the fashion and celebrity market, pretty much kill off the actual user market for new sales in the M-E and M240 as well.

That will be my dream come true a good range finder on top of the A7 that is 🙂

+1 to *hepcat* and *rogazilla* That's what's needed - real competition within the digital RF segment. None of what the OP and *semilog* argue for w/r/t sensors matters much if what the buyer wants is what only Leica provides today: digital capture with true rangefinder haptics and problem-free use of the full range of M-glass.

 

[__--]

rogazilla/Mike, sure, but Leica, in its niche market, didn't get much competion for the M6/M7/MP cameras. Zeiss stopped its ZI rangefinder initiative because they could not make a profit on it. Why would the situation be different with a digital rangefinder camera, considering that the Leica niche for this is also small?

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

 

[hepcat]

rogazilla/Mike, sure, but Leica, in its niche market, didn't get much competion for the M6/M7/MP cameras. Zeiss stopped its ZI rangefinder initiative because they could not make a profit on it. Why would the situation be different with a digital rangefinder camera, considering that the Leice niche for this is also small?

MITCH ALLAND/Potomac, MD

Mitch, I think that there are several things that could expand that market today over what it was just a few years ago; not the least of which is the explosion in mirrorless full frame cameras. Much hype has been made of the Leica rangefinders over the years, and there are now affordable lenses in the marketplace. The mirrorless "revolution" (if you will) has brought small, competent cameras back into vogue. I think that if the RD1 were reintroduced with a more recent, competent FF sensor and longer base rangefinder at a price point of 1/3 to 1/2 of the digi Ms, that the larger market share of digital rangefinders would really take off exponentially. Of course, my market research is just reading all the forums posts of folks who would buy them if they could afford them... so I don't know how representative that truly is? OTOH who'd think that Mr. K could still be selling Bessa film bodies?

 

[rogazilla]

I wonder if this niche market is a niche because a lot of would be buyers are priced out of the current offering? I think that should be surveyed among people who is currently buying mirrorless cameras. If they would be willing to spend 2500 on a A7 with a range finder on top. I picked 2500 because that's about what the market price for a 5d3 is.

The evf on the a7 is great and I don't mind using it except that I find in broad day light and I stop down my lenses the finder always seems a bit dim. it might be because my eyes are used to how bright the sun is and couldnt adjust to the brightness of the finder? I still prefer OVF.

One interesting device is the finder on the X100(s). I absolutely love it but I don't want to have to switch to EVF to mf... If anyone can come up with a clever solution for that hybrid view finder I think will be the answer that can sell the device to the old guard and the new generation.

One thing I thought about but probably cost tons to implement is to have another small sensor on the top right to take in image and display as a patch over lay in the OVF? a small sensor like the one in our mobile phone?

A lot of the new generation prefer the back LCD anyway. if you put a range finder on top of an A7, you probably can still sell it to the hipster or fashionista whatever you call that will still buy them 🙂 Market is an interesting thing and how you "market" your product is also sometime defy logic and reasons.

 

[__--]

Mike/rogazilla,

Interesting.

—Mitch

 

[rogazilla]

I just saw HTC and Apple are experimenting 2 image sensor on the next mobile phone...

but back to OP's point, Sony only in 2013 put the 24mp sensor in a mirrorless body and from my understanding, this is the only other FF mirrorless body out there. From a packaging point of view, could it be heat? or other aspect that before A7, leica is the only one doing so? Or is it simply that sony/canon/nikon don't believe there is a market for mirrorless FF?

There seems to be a lot of engineers who understand those spec sheets, were there any technical difficulties to package a FF sensor that is much superior than the leica ones in a mirrorless body? There seems to be more than just a plug and play when it comes to these sensors.

The reason I brought this up is because the sensor in A7R has a different shutter mechanism than the A7. The explanation from sony is that it has to be different because the sensor.

If all of these chips are plug and play... I would love to put a desktop version of the intel i7 into my laptop instead of the laptop version that is inferior...

 

[Aristophanes]

rogazilla/Mike, sure, but Leica, in its niche market, didn't get much competion for the M6/M7/MP cameras. Zeiss stopped its ZI rangefinder initiative because they could not make a profit on it. Why would the situation be different with a digital rangefinder camera, considering that the Leice niche for this is also small?

Fuji appears to have captured hundred of thousands of units with their X-series with similarities much closer to a traditional RF than a DSLR. They saw potential with tactile control and OVF and went for it. We still have not seen their best glass.

Leica has priced people out. Used digital is more of a gamble than used mechanical (M8 screens), so that curtails the buy-up starter market substantially. There is probably an Excel file somewhere at Leica HQ where they've plotted the used price thresholds of their brand against the functional depreciation and serviceability of the digital models (automakers do this). The Leica T appears to be the interim model to the M, but again there will be questions of value. They appear to be using V.3 of the 16MP Sony sensor which first appeared just under 4 years ago in the Nikon D7000 and Pentax models. And one has to wonder about a milled aluminum block which will last until our sun goes red giant compared to the serviceable life of the electronics. The teardown of the Sony A7 hinted that Sony's ability to service and replace the entire sensor module is possible. Hmmmm....

(Looking at it from a pure capital capacity I am not sure Fuji can step up to FF, not in the short to mid-term. Might be surprised. Their vertical integration between body, sensor, and lenses is well-executed. Weak link is software.)

There are obviously technical differences in adapting different sensors across brands and flange distances. Sony is doing it with their two mounts as has 43 to m43. The M-mount flange is almost exactly in between the 43 to m43 distance. Canon is doing similar by modestly altering their sensor for their new mirrorless.

 

[MCTuomey]

rogazilla/Mike, sure, but Leica, in its niche market, didn't get much competion for the M6/M7/MP cameras. Zeiss stopped its ZI rangefinder initiative because they could not make a profit on it. Why would the situation be different with a digital rangefinder camera, considering that the Leice niche for this is also small?

Mitch, I think that there are several things that could expand that market today over what it was just a few years ago; not the least of which is the explosion in mirrorless full frame cameras. Much hype has been made of the Leica rangefinders over the years, and there are now affordable lenses in the marketplace. The mirrorless "revolution" (if you will) has brought small, competent cameras back into vogue. I think that if the RD1 were reintroduced with a more recent, competent FF sensor and longer base rangefinder at a price point of 1/3 to 1/2 of the digi Ms, that the larger market share of digital rangefinders would really take off exponentially. Of course, my market research is just reading all the forums posts of folks who would buy them if they could afford them... so I don't know how representative that truly is? OTOH who'd think that Mr. K could still be selling Bessa film bodies?

Good questions/points. I'd jump on such a camera at that price point, that's all I know, for what I shoot.

Fuji appears to have captured hundred of thousands of units with their X-series with similarities much closer to a traditional RF than a DSLR. They saw potential with tactile control and OVF and went for it. We still have not seen their best glass.

(Looking at it from a pure capital capacity I am not sure Fuji can step up to FF, not in the short to mid-term. Might be surprised. Their vertical integration between body, sensor, and lenses is well-executed. Weak link is software.)

There are obviously technical differences in adapting different sensors across brands and flange distances. Sony is doing it with their two mounts as has 43 to m43. The M-mount flange is almost exactly in between the 43 to m43 distance. Canon is doing similar by modestly altering their sensor for their new mirrorless.

Interesting. I just purchased my first fuji x-trans camera - excited to see whether it helps me improve my club shooting, among other subjects. Judging from the web, it sure appears their glass is moving up the curve with the 14/2.8, 23/1.4, 56/1.2, even apparently their kit zoom. Trying not to think about equivalence, just going to compare output and haptics to my M-gear over a few months' shooting. Not sure what to do about processing software since I'm PC, not Mac, and the better x-trans RAW converters are Mac-only.

 

[mlu19]

I still believe modern Leica is just a premium/luxury brand. At least, that's how I felt when I decided to get a M9P in 2013, knowing there's tons of cameras out there that offer more image "quality".

A $10 Casio probably tells time better than a $10,000 Rolex. Which one would you go for?