Cypress *MONOCHROME* sensor... please please please...

nada
 
$50 APS, $500 FF so it'd mean an APS DSLR for $500 and FF for $950
or $1000 and $1500 respectively. Some folks overpaying :).
DALSA charge 2000 euro for a 24x24mm 4MP 30 fps sensor. So 2450 for a short DoF high ISO 'film' camera then :-) add $300 for a MJPEG and 80GB 2.5" HD implementation too ...

The primitive DALSA 'camera' using the sensor is 20k euro though... and slightly more for the 36x24mm 6MP 18fps camera.

They have 'slight' smaller volume then the DSLRs :-)

--
Henrik
 
Who says sensors are expensive????
you get what you pay for....
yep. All the people hping for good high ISO performance should have
a look on the last fillfactory sensors without microleses.

"...Our patented pixel architecture, proven in Kodak's DCS Pro SLR
line of cameras, is the only CMOS architecture capable of achieving
the noise and dark current targets demanded by high-end DSC
manufacturers..."

So you know what to expect.

Maybe this sensor will deliver good dynamic range and low CAs
maybe, but there are still issues like the "italian flag", lack of AA filter, ...

--
http://www.pbase.com/dgross (work in progress)
http://www.pbase.com/sigmasd9/dominic_gross_sd10

 
it also depends on the packaging. Large FF chips need more ground and voltage planes to avoid getting overwhelmed by ground noise due to the large silicon area. It also needs more pins, to give higher bandwidth. The base chip alone, in silicon area, wil usually triple the cost.

Yields affect FF much more severly. If the APS yield is 75%, than the FF yield will be 30%.

Volumes for FF are also lower, due to higher intrinsic costs, which also adds to the cost.
  • Shel
 
Only, I'd love to see it in a compact dSLR. Think about where cost savings could come from! No need for demosaicing (sp, I'm sure) in camera. Even better if they leave the hot mirror filter off of it. And a dSLR could be cheaper, because they wouldn't need to put a high end lens on it. A true 9mp B&W camera is about the same as a 36MP Bayer color camera, right?

If Canon or Nikon picked this up, I'd be way happy. Pentax, too. I'd consider buying a Pentax. R&D can't be very high, compared to current dSLR's, can it?
 
Moire shows up in B&W at well, not just color. It's obviously much more noticable in color, but it doesn't go away entirely.

OTOH, the AA filter could be much weaker than it would be otherwise.

I wonder how this sensor performs at ISO 3200.

This is exciting news, no doubt about it.
 
With about the same rate of defects per wafer, cost to camera makers might increase about five-fold or more when going to 24x36mm, which is 2.25 times larger than this Cypress sensor

For example, Canon reports 70% yields for its "APS-C" sensors, 25% for the 5D sensor up from 10% for the 1Ds Mk II sensor. The ratio from 70% to 25% and the fact that they get about 2.5 times as many of the smaller sensors per wafer means they get about 7 times as many working APS-C sensors per wafer, cutting that part of costs seven-fold. Costs tend to get doubled or tripled from manufacture to retail, so if a Cypress sells 24x36mm sensors for 5 to 7 times $90, or $450-630, it could flow through as about $1,300-1,900 impact on retail cost.

Still about a $1,000-1,500 premium for a bigger sensor, and still having to use bigger, longer lenses.
 
Without the CFA, sensitivity ought to go up. That may be part of why they think microlenses are unnecessary. Boy, this thing opens up possibilities.

Canon? Nikon? Are you listening?
 
Moire shows up in B&W at well, not just color. It's obviously much
more noticable in color, but it doesn't go away entirely.
no color => no bayer interpolation => no color moiree (see Foveon)
OTOH, the AA filter could be much weaker than it would be otherwise.
get rid of it on a monochrome camera.
I wonder how this sensor performs at ISO 3200.
who said, that it will provide ISO 3200 ?

best regards
 
A monochrome APS body could be interesting; say in a body also aimed as astronomical and IR, a bit like the EOS-20a.

But Kodak makes various digital cmera sensors in both color and monochrome versions, with the latter only getting used in scientific/engineering equipment, not retail cameras.
 
A monochrome sensor would need an AA filter just as much as a certain color sensor that starts with an F, that is, not a whole lot. Same with the Italian flag thingy, which could anyway be easily corrected in firmware in a fixed-lens camera, since the characteristics of the lens would be known. No more difficult than interpolating color on Foveon.

Petteri
--
My flickr page: [ http://www.flickr.com/photos/primejunta/ ]
Me on photography: [ http://194.100.88.243/petteri/pont/ ]
Me on politics: [ http://p-on-p.blogspot.com/ ]
 
in the lab I work in the biggest draw back of digital cameras are the bayer sensors. you can buy scientific cameras but they're a terrible value (pixel/dollar). They need specialty software and offboard storage and training. Good software costs more than the camera, and you need to buy software for every camera, so if you bought ten camera's you have to spend an insane amount on multiple version of the same software.

When I first read the headline for the 20Da my heart raced, I thought 'please god tell me it's bw'. but it's not.

The day you can buy a dslr camera with a monochrome sensor for under 2000 I am in big time.
 
You might make fun of us, but as you can see, there's a fairly
enthustiastic group dream of just such a thing.
Indeed. If it had the quality I hope it would (read: "blow the color sensors out of the water" :), and if it is cheap to produce, perhaps Nikon and Canon could incorporate it into their existing camera designs and sell them at similar prices.

The only possible snag could be low light sensitivity (see later post by BJL). I would like to have ISO 800 or higher.

--
hobster
 
It could tap into
one of the last groups of photographers still using film - the the
Tri-X, HP5+, PanF, Neopan shooters out there.
BTW, it's been a long time since I've used any of those, but I can still recite chemical recipies for various developers. Yes, I used to put them together from scratch. Just the thought of big dark jars with chemicals brings sweet memories. :)

--
hobster
 
Darn, now I will wait with my buying decisions to PMA to see if there are any hints from any of the major manufacturers of possibly incorporating it into their cameras.

--
hobster
 
A monochrome sensor would need an AA filter just as much as a
certain color sensor that starts with an F, that is, not a whole
lot. Same with the Italian flag thingy, which could anyway be
easily corrected in firmware in a fixed-lens camera, since the
characteristics of the lens would be known. No more difficult than
interpolating color on Foveon.
well, certain people tend to complain about the lack of AA filter on a certain F-Sensor, If you assume that compensating for the "italian flag" for a fixed lens is easily implemented, why is it impossible to correct for known lenses used on the 14n / SLRn / SLR/c?

Chances that you will get a monochrome camera are very low anyway, and for a color Camera all these problems become very important.

--
http://www.pbase.com/dgross (work in progress)
http://www.pbase.com/sigmasd9/dominic_gross_sd10

 
I think the ideal body/lens combo for this idea would be patterned after the Leica Digilux 2/Panasonic DMC-LC1. Unbelievable feel to that camera. Too bad the sensor and digital camera features weren't anywhere near up to the rest of the camera.
 
--A big monochrome CCD using the tricolour process for landscapes would
kick every commercial camera's ass when it comes to quality.
-Rich
 
A monochrome sensor would need an AA filter just as much as a
certain color sensor that starts with an F, that is, not a whole
lot. Same with the Italian flag thingy, which could anyway be
easily corrected in firmware in a fixed-lens camera, since the
characteristics of the lens would be known. No more difficult than
interpolating color on Foveon.
well, certain people tend to complain about the lack of AA filter
on a certain F-Sensor, If you assume that compensating for the
"italian flag" for a fixed lens is easily implemented, why is it
impossible to correct for known lenses used on the 14n / SLRn /
SLR/c?
Probably because there are just too many of 'em to database properly.

Why does this complaint remind me of the Nikonians' complaints about Canon's "plastic skin" or the Canonians' complaints about Nikon's moiré or the Olympians' complaints about dust on every camera that they don't use and so on ad nauseam...?
Chances that you will get a monochrome camera are very low anyway,
and for a color Camera all these problems become very important.
I wouldn't be particularly interested in the color version of the sensor anyway. But I'm less picky about monochrome.

Whazzup on the Sigma front, by the way? I haven't been following stuff much, but wasn't something big supposed to be in the pipeline from there? Anything new about that?

Petteri
--
My flickr page: [ http://www.flickr.com/photos/primejunta/ ]
Me on photography: [ http://194.100.88.243/petteri/pont/ ]
Me on politics: [ http://p-on-p.blogspot.com/ ]
 

Keyboard shortcuts

Back
Top