Interesting new sensor tech announcement

T

Terao

Kiloran
Local time
11:07 AM
Joined
Aug 14, 2006
Messages
962
http://www.dpreview.com/news/0705/07051401panasonicrobustsensor.asp

Potentially a big deal for RF shooters as it promises uniform sensitivity across the sensor and seems to indicate that they've ditched conventional microlenses. That *could* mean the end of the crop and the end of (digital) vignetting, even taking into account the closeness of the rear element to the lens and the resulting high angle of incidence at the edges.

Given that its Panasonic tech perhaps we're likely to see it in an M9 :cool: . More importantly, I wonder if Zeiss is listening :D
 
Just something to point out I think...
There's no reason to think that current technology sensors won't last 20 years or more either.
The difference here is that this sensor is said to withstand constant exposure to light for a very long time. The CMOS sensor in my 300D is only exposed to light for very short periods of time. So even with 75,000 plus shots, it's in little danger of degrading due to exposure to light.
The new sensor's strength would be durability under different conditions than your normal camera application.
 
Agreed, even in Arizona its not going to be an issue for a DSLR (although the desert wind probably will be!)
The main application for this will no doubt be permanently exposed hardened sensors for traffic monitoring, security cameras, etc but a nice side-effect appears to be the death of the physical microlens. Probably a few years from market, and we can but hope that Panasonic don't use their Venus Engine...
 
They don't specifically say it, but is that last paragraph implying that you may be able to select the filtering ability?
 
There's an academic paper about this on the IEEE site if you have a login. The abstract states:

"This letter demonstrates a long-term reliable MOS image sensor equipped with a novel photonic crystal color filter (PC-CF). PCs are periodically structured dielectric media, generally possessing a photonic band gap. In the newly developed PC-CF, it is clarified that the desirable spectral property, i.e., peak wavelength and spectral passband, can be achieved by modifying the thicknesses of the defect layers, which act just like a "defect" in a PC. The spectral characteristics with the peak wavelengths at 450 (blue), at 530 (green), and at 610 nm (red) are realized in this letter. Moreover, the fabricated image sensor guarantees high reliability of longer than 200000 h and heat resistance of above 300 °C."

Doing some further reading it looks like these wee beasties are already used in lens coatings by both Zeiss & Nikon (perhaps the source of the Nano coating that Nikon have been marketing on their latest high-end glass?)

Looks like they think they're able to grow a crystal layer into a Bayer-type matrix such that each photosite is only receptive to one wavelength (be that red, green, or blue) and they can turn up or turn down the gain on each photosite according to its position relative to the optical axis.

Sooo, imagine the M9 proposition:

6-bit encoded lens tells the body what it is (or Leica let us do that manually)
Body amps the sensor appropriately
Rather than applying a post-process gain to those photosites the camera simply does it at capture and more importantly in RAW

Sounds cool to me :)

I only mention Leica because of their alliance with Panasonic, somehow I can't see Panasonic letting Zeiss/Sony play with this stuff...
 
It sounds more like it's built for video or digicam applications where there's no physical shutter.
 
Kin Lau said:
It sounds more like it's built for video or digicam applications where there's no physical shutter.
Click to expand...

Why does the M9 need a physical shutter? If they can produce a large sensor with electronic shutter, then isn't that the holy grail of un-obtrusive photography?
 
ywenz said:
Why does the M9 need a physical shutter? If they can produce a large sensor with electronic shutter, then isn't that the holy grail of un-obtrusive photography?
Click to expand...
Just imagine the sound that "no shutter" makes. Where's that Leica "snick"? I predict they will have a digital "snick" tone available for die-hard RF enthusiasts. :)

/T
 
toyotadesigner said:
No need for Zeiss because they are building their own sensor fabrication in Germany right now. Reason: their high res lenses need hi res sonsors featuring the same quality as the lenses or at least being able to transform into bits and bytes what their lenses deliver.

Makes a lot sense to me.
Click to expand...

Interesting, didn't know Zeiss were up that. Digital Ikon on the horizon then :cool:
 
ywenz said:
Why does the M9 need a physical shutter? If they can produce a large sensor with electronic shutter, then isn't that the holy grail of un-obtrusive photography?
Click to expand...

Sony R1, Nikon D70 and the original Canon 1D all have electronic shutters with APS-C or greater sensors. The R1 doesn't even have a physical shutter.

I was just remarking on the fact that it looks like a digicam or videocam sensor.
 
Tuolumne said:
Just imagine the sound that "no shutter" makes. Where's that Leica "snick"? I predict they will have a digital "snick" tone available for die-hard RF enthusiasts. :)
/T
Click to expand...
Downloadable soundeffects, to be used in a "My Leica" function?

Hmm...
 
Kin Lau said:
Sony R1, Nikon D70 and the original Canon 1D all have electronic shutters with APS-C or greater sensors. The R1 doesn't even have a physical shutter.

I was just remarking on the fact that it looks like a digicam or videocam sensor.
Click to expand...


The original 1D had a mechanical shutter but it was only to cap the shutter for protection and came into play only at slow shutter speeds for exposures. Don't remember the exact slow speed where it became a shutter but it was a great camera and very accurate shutter speeds.
 
x-ray said:
The original 1D had a mechanical shutter but it was only to cap the shutter for protection and came into play only at slow shutter speeds for exposures. Don't remember the exact slow speed where it became a shutter but it was a great camera and very accurate shutter speeds.
Click to expand...

I've read that it's for BULB only.

There's a reason why I only mentioned the R1 did not have a physical shutter:D .
 
Is this actually some kind of dichroic filter?

I still don't see why it could solve the problem of oblique incidence, since you would still have to vary the gain of the pixels at the edges according to different lenses used.
 
Yep, you'll have to vary the gain at the edges. Not an issue for digicams of course but an interesting challenge for uncoupled interchangeable RF lenses to overcome (6-bit coding a la Leica would help)
 
You must log in or register to reply here.
Back
Top Bottom