The M10 and its guts..

[jaapv]

Originally Posted by bfffer

Foveon X3 for sure.

For sure- until one has a look at at the limited acceptance angle which makes it useless for short regisyer cameras like rangefinders....

[kbg32]

One has to remember that the GXR and the NEX do not contain full frame sensors. The problems inherent with a full frame sensor rangefinder, do not exist for "cropped sensor" cameras, with, or without micro lenses.

[animefx]

I would like to see a Kodak CCD once in the Leica M10. I think they are more than capable of giving us a solid 2 stop improvement in ISO noise while giving us even more megapixels to take advantage of our wonderful glass, or some cropping power to make up for the minimum focusing distance in the Leica rangefinders.

I'm mostly in support of CCD over CMOS mostly because I've been spoiled by the dynamic range in my M8, especially shadow details. Forget all the of tests, in real world use I get better usable dynamic range than my 5D2 does. I know the M9 is slightly better, so hopefully the M10 takes this a step further, maybe a small improvement in the shadows and the highlights over the M9 would be nice.

Anyway, If we could get a 2 stop improvement in ISO noise vs. the M9, then I think Leica should limit the M10's maximum to ISO 5000, which would hopefully look like iso 1250 on the M9.

The Sony NEX focus peak option looks nice, but I'm not sure I want that in my rangefinder cameras, especially if it means sacarficing CCD for CMOS. The better option is the simply buy a Nex 7 or 5n as a secondary camera.

[jaapv]

If the Kodak CCD could do that, the M9 would have it, as it has a Kodak CCD... More megapixels=more noise so this is rather contradictory. And what would you need more pixels for? According to Zeiss, leica and Erwin Puts, anything over 40 Lp/mm is meaningless for image quality, so 18 Mp on a 24x36 sensor is the sweet spot.

[animefx]

The Kodak CCD in the M9 is the best they could do at the time, the sensor is now over 2 years old. Advancements in sensor technology make it credible that the M10's sensor could be an improvement in several areas including ISO noise, dynamic range, and even megapixels.

More megapixels does not always mean more noise... For example, look at the original Canon 5D vs. Canon 5D mk II... The mk II has 9 more megapixels and 2 stops better ISO noise. Another example is the Canon 1D mk4 vs. mk3, same story, the mk4 has more megapixels and lower noise.

I'm not saying 18 mp is bad, I'm just saying there is nothing wrong with having more if anything for cropping power. With prime lenses the extra megapixels are often a god send. Heck I prefer using my M8 over my 5D2, but that doesn't mean it wouldn't be useful having extra megapixels to play with.

Originally Posted by jaapv

If the Kodak CCD could do that, the M9 would have it, as it has a Kodak CCD... More megapixels=more noise so this is rather contradictory. And what would you need more pixels for? According to Zeiss, leica and Erwin Puts, anything over 40 Lp/mm is meaningless for image quality, so 18 Mp on a 24x36 sensor is the sweet spot.

[hxpham]

Originally Posted by animefx

I'm mostly in support of CCD over CMOS mostly because I've been spoiled by the dynamic range in my M8, especially shadow details. Forget all the of tests, in real world use I get better usable dynamic range than my 5D2 does. I know the M9 is slightly better, so hopefully the M10 takes this a step further, maybe a small improvement in the shadows and the highlights over the M9 would be nice.

Have you seen the DR of the new Sony sensor used in the K5/D7000? Better than the M9

[rogerzilla]

The software is getting better at reducing noise, however horrid such an idea seems to we film users. The M9 software is often reported as being poor when it comes to making JPG files and there is huge scope for improvement there.

Then they need to put in a manual shutter cocking mechanism (for quiet operation) and offer some VF magnification choices. They could even (sacrilege!) offer a zoom viewfinder for the different focal lengths and banish all those difficult .85, .72 or .58 choices completely.

[jaapv]

Originally Posted by animefx

The Kodak CCD in the M9 is the best they could do at the time, the sensor is now over 2 years old. Advancements in sensor technology make it credible that the M10's sensor could be an improvement in several areas including ISO noise, dynamic range, and even megapixels.

More megapixels does not always mean more noise... For example, look at the original Canon 5D vs. Canon 5D mk II... The mk II has 9 more megapixels and 2 stops better ISO noise. Another example is the Canon 1D mk4 vs. mk3, same story, the mk4 has more megapixels and lower noise.

I'm not saying 18 mp is bad, I'm just saying there is nothing wrong with having more if anything for cropping power. With prime lenses the extra megapixels are often a god send. Heck I prefer using my M8 over my 5D2, but that doesn't mean it wouldn't be useful having extra megapixels to play with.

Even cropping won't change things like the Nyquist frequency and the relationship between lens resolution and sensor resolution. I think that in the M9 and S2 sensor the CCD technology has reached its pinnacle. CMos is moving quite slowly nowadays as well, the 5Dii sensor is not that different from the 10D. The main progress is in the incamera processing, Nikon especially has it down to a fine art. The world is waiting for new technology. Both CCD and CMos have reached the stage of diminishing returns. And technology runs up against the biological barrier. The human eye has it limitations too, and though imaging may have the potential to become much better with new technology - if you cannot see it it makes no sense to implement.

[jaapv]

Originally Posted by rogerzilla

The software is getting better at reducing noise, however horrid such an idea seems to we film users. The M9 software is often reported as being poor when it comes to making JPG files and there is huge scope for improvement there.

Then they need to put in a manual shutter cocking mechanism (for quiet operation) and offer some VF magnification choices. They could even (sacrilege!) offer a zoom viewfinder for the different focal lengths and banish all those difficult .85, .72 or .58 choices completely.

Please no. I really miss the mechanical advance when using film Ms... And JPG? Real men shoot raw...

[jaapv]

Originally Posted by hxpham

Have you seen the DR of the new Sony sensor used in the K5/D7000? Better than the M9

Numbers? What is your source?

[FrozenInTime]

Originally Posted by jaapv

Numbers? What is your source?

http://www.dxomark.com/index.php/Cam...(brand3)/Leica

Then hit measurements|Dynamic Range

Has anyone done real world side by side comparisons to back this up ?

[animefx]

hxpham make a good point... take a look at the insane amount of dynamic range on the k5 here...

http://photography-on-the.net/forum/...d.php?t=955995

if you did this kind of thing with the sensor in the 5d2 the resulting photo would look awful beyond reason.

this just goes to show you that we aren't even close to the end of sensor technology advancements. in addition, I just learned kodak has a 29 megapixel ccd as well in the works.

nikon certainly massages the raw files with in camera software and maybe canon does as well, but there is a significant difference between the canon 10d sensor and 5d2 sensor, and this is over the span of what 5 or 6 years in technology?

one last thing to look at quickly if you have the time... is the dxo test on the k5 sensor, which is a crop sensor by the way. it has more dynamic range than nikon's $9,000 flagship camera. as a side note I hate dxo test and the like because they are often deceiving especially for any kind of "real world use" the leica m8 has a more usable range than the 5d2 in my opinion, but by only looking at their tests many people wouldn't know that.

http://www.dxomark.com/index.php/Pub...Range-K5-rules!

Originally Posted by jaapv

Numbers? What is your source?

[gavinlg]

DXO is rubbish. I wouldn't trust 1 line from DXO's website.

Have a look at the lens reviews. They rate the Zeiss 21mm distagon for EOS the lowest resolving lens they tested for canon DSLRs (it's actually one of the very best lenses available). They also rate the other Zeiss for canon lenses badly. Now I've personally owned the Zeiss ZE lenses and they're superb - you can cross check that with other review sites across the web.

There's so many things that they list that are false the whole dxomark thing is just ridiculous. I wouldn't be basing my opinions on cameras off their data.

[Eddie90723]

I vote Foveon on this one. Can you imagine the elitism and sheer snobbery that would come along with that combo? "oh, you use cameras whose images can be processed by standard photo programs? That's nice, mine can only be processed by a single program and the colors are better than your colors and the pictures can't be anything but amazing because it's a Leica." lol. It would be a fun if not incredibly frustrating experience from time to time. I actually had purchased the SD1 with the idea on popping a Leica R adapter onto it and using R glass in hopes of the glass being able to keep up with the sensor, but I sent it back in less than a week because it just did not get the job done at all. Oh Lordy I wanted to cry, it was just frustrating! So I got an M9 instead. Worked out so much better. I think once Sigma works out its issues with Foveon sensors, it would match M glass really happily.

[Brian Sweeney]

I doubt that a Foveon style sensor could be made to work with a rangefinder camera because of the short "film to flange" distance. And- there is no full-frame version of the Foveon sensor.

I'd rather see a Monochrome sensor offered as an option.

[Eddie90723]

A full frame Foveon would be awesome, but how expensive would that be? 7k for the ApsC size isn't very friendly. Monochrome sensor would be nice. I shoot my M in Raw + Jpg b&w and convert all my pics to mono. I get a little disturbed when I see my images in color honestly, makes me think I didn't get the settings right.

[Brian Sweeney]

A Monochrome sensor picks up twice as much light as any color sensor. And no dye layer to worry about.

[jaapv]

Originally Posted by fdigital

DXO is rubbish. I wouldn't trust 1 line from DXO's website.

Have a look at the lens reviews. They rate the Zeiss 21mm distagon for EOS the lowest resolving lens they tested for canon DSLRs (it's actually one of the very best lenses available). They also rate the other Zeiss for canon lenses badly. Now I've personally owned the Zeiss ZE lenses and they're superb - you can cross check that with other review sites across the web.

There's so many things that they list that are false the whole dxomark thing is just ridiculous. I wouldn't be basing my opinions on cameras off their data.

And they do not test sensor output as they claim, they test camera output, including all firmware corrections to the file....

[_mark__]

They call it the MD, it has the cloth m shutter, an upgraded kodak sensor with better iso and dr, can't see it happening for a few years yet though; M9's good enough. Look forward more to the X2.

[bwcolor]

For those suggesting an updated Kodak CCD, how is Kodak's financial health? There were rumors that Leica might look for a new sensor partner due to concerns over Kodak's ability to deliver into the future.

[hxpham]

Okay, fdigital ,how about this? http://translate.google.com/translate?u=http%3A%2F%2Fwww.guillermoluijk.com%2F article%2Fsnr%2Findex.htm&langpair=es|en&hl=EN&ie= UTF-8
Too long; didn't read -- The K5/D7000 sensor is practically ISO-less. You get almost the same results shooting a severely underexposed ISO 100 image and brightening it in RAW processing as you do shooting a well exposed ISO 1600 image.

[gavinlg]

Originally Posted by hxpham

Okay, fdigital ,how about this? http://translate.google.com/translate?u=http%3A%2F%2Fwww.guillermoluijk.com%2F article%2Fsnr%2Findex.htm&langpair=es|en&hl=EN&ie= UTF-8
Too long; didn't read -- The K5/D7000 sensor is practically ISO-less. You get almost the same results shooting a severely underexposed ISO 100 image and brightening it in RAW processing as you do shooting a well exposed ISO 1600 image.

Yeah damn cool, pity Pentax K5 colors are wack... I'd like to see what they can do with a new full frame sensor.

[htimsdj]

I'll take the M10 with something similar to the Nikon D3 sensor, with no Bayer filter. I just want black and white with high iso capability.

I tried the M9, and its not for me. I'll take film grain and high contrast for my low light shots at this point.

[rxmd]

[Faintandfuzzy]

I'd say either a Kodak sensor or Canikon. There really is no better color fidelity obtained when using a CCD. As to the Foveon, I still chuckle when I hear about color fidelity from Sigmas. Let's get one thing clear....the Foveon has the WORST color going. Period. Check an iso 800 or 1600 image from the SD1 and come back and tell us about color. It's a joke. If Leica went Foveon (and I don't think they're that stupid), I wouldn't touch anything from Leica again.

And Animefx, the M8 does not have great dynamic range. In fact, it's rather mediocre. Today's CMOS sensor wipe the floor with these old CCD chips that are ridulled with noise and weak DR. Cameras like the Pentax K5, Nikon D700, Nikon D7000, Nikon D3x, etc, etc, leave the M8 way behind in noise and DR.

We don't need CCD. We definitely don't need the poor color and horrendous aliasing and oversharpend look of Foveon.