85MP

u19925 said:

At 30 e noise, it is ruled out for professional astronomy.

Click to expand...

That's at room temp, I think. Chilling will bring that down drastically.

u19925 said:

Also, 85
mp is too damn big to make it of the quality required for space.

Click to expand...

I don't know what you mean. Do you mean that the can't make it work
in orbit? I would bet that that is where you would find the majority of
the production run right now.

u19925 said:

This means, military applications are more likely.

Click to expand...

I'm sure that this was developed for reconnaissance purposes, and
Fairchild is just trying to make another few bucks from it.

[snip]

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Dan Frederiksen said:

sounds like a cool place to "work"
type of place that should be begging for someone like me : )

what is your job description/qualification? comp sci? image
processing?

Click to expand...

My job description is "technical lead, voice technolgies". And I keep pushing for promotion to "technical fellow", or at least for the creation of a legitimate "psychophysics" position, since I do both speech and image processing work (and some graphics software

My masters is control systems, the emphasis was DSP and pattern recognition, and my postgrad work is bioelectronics, psychoacoustics, musical acoustics, image processing. I could never focus long enough to get a Ph.D. out of any of that. I also acquired a nonspecialized BFA, but have managed to purge most of it.

Dan Frederiksen said:

does any of the product category projects ever become real products?
I've seen IR as a novelty in a jag once but never really in mass
production.
are the rearview "mirrors" in the works?

Click to expand...

I can't comment on that.

--
Ciao!

Joe

http://www.swissarmyfork.com

Vtie said:

u19925 said:

Also they need lesser noise than 30e specified for this chip.

Click to expand...

They are cooled. That's how they achieve lower noise.

Click to expand...

Well, I am just quoting the spec. If the chip will perform better at lower temperature, then it needs to be specified.

BTW, cooling reduces dark current and only small amount of read noise. I thought the 30e noise mentioned was read noise.

Vtie said:

Vtie said:

most telescopes except
the survey ones don't have very large FOV, so they don't need
really huge CCD.

Click to expand...

I don't understand what you mean by that. The FOV may be quite small,
but the usable image thrown by the telescope is large, and this
is what determines the size of sensor you can use.

Click to expand...

The ground based telescopes are limited to close to 1 arc second angular resolution due to air turbulance (during best nights, they are good upto 0.2 arcseconds). This means, FOV and image sizes are related. No point having 85 mp sensor if your FOV is 1 arc minute.

You might be tempted to mention adaptive optics but that only works with very narrow FOV and requires even smaller CCD.

For information, Keck telescopes (the world's largest ones) uses 8 MP CCDs as the biggest size. They achieve 8000x8000 using an array of 8 CCD (each 2000x4000).

Vtie said:

Of course, the
actual size of the sensor depends on what you want to observe. If you
are looking at a distant small object, it may be small. However, for
many applications such as high redshift cosmology, the larger the
sensor
the better. In the old days of photographic plates, people routinely
used plates like 20cm x20cm and more for many observations.

Click to expand...

20x20 plates were used for surveys. I have seen 30x30 too. These are Schidmt telescopes specially designed for these kind of work. As I said, there are very few applications where these are used.

Again, nothing prevents them from using mosaic CCD array.

--dhiraj

Vtie said:

Vtie
http://www.pbase.com/vtie

Click to expand...

u19925 said:

Vtie said:

u19925 said:

Also they need lesser noise than 30e specified for this chip.

Click to expand...

They are cooled. That's how they achieve lower noise.

Click to expand...

Well, I am just quoting the spec. If the chip will perform better
at lower temperature, then it needs to be specified.

Click to expand...

Like where it says "Value shown is for 15C" on the data sheet?

Anyway, a bit of looking around finds the more complete data
sheets on the lesser arrays, e.g. the 16MP 6x6 cm back-lit array:

http://www.fairchildimaging.com/main/documents/ccd486.pdf

The "typical" dark current looks to be around 0.1 e- sec at -100C.

"Dark signal is a linear function of integration
time and an exponential function of the chip temperature."

u19925 said:

BTW, cooling reduces dark current and only small amount of read
noise. I thought the 30e noise mentioned was read noise.

Click to expand...

That's right. But they don't mention the readout rate. No, wait, they
do: "The nominal read noise of the output amplifiers is less than 25
electrons at 25MHz." The 16MP sensor is rated at less than 4 e- at 50K
pixels/sec; there's obviously a tradeoff involved.

u19925 said:

u19925 said:

I don't understand what you mean by that. The FOV may be quite small,
but the usable image thrown by the telescope is large, and this
is what determines the size of sensor you can use.

Click to expand...

The ground based telescopes are limited to close to 1 arc second
angular resolution due to air turbulance (during best nights, they
are good upto 0.2 arcseconds). This means, FOV and image sizes are
related. No point having 85 mp sensor if your FOV is 1 arc minute.

You might be tempted to mention adaptive optics but that only works
with very narrow FOV and requires even smaller CCD.

For information, Keck telescopes (the world's largest ones) uses 8
MP CCDs as the biggest size. They achieve 8000x8000 using an array
of 8 CCD (each 2000x4000).

Click to expand...

Right. So that's 64MP total.

[snip]

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

u19925 said:

Well, I am just quoting the spec. If the chip will perform better
at lower temperature, then it needs to be specified.

Click to expand...

It holds for every CCD. It's physics.

u19925 said:

The ground based telescopes are limited to close to 1 arc second
angular resolution due to air turbulance (during best nights, they
are good upto 0.2 arcseconds). This means, FOV and image sizes are
related. No point having 85 mp sensor if your FOV is 1 arc minute.

Click to expand...

My interpretation of "image size" is the physical dimension of the
sensor. In that case, I fail to understand why the seeing limitations
or FOV have any implication on the image size. However, it may be that
by "image size", you mean the resolution (i.e. number of pixels) of
the CCD. Is this what you mean? In that case, you are right that the
atmospheric turbulence, together with the FOV puts some limitations on
the resolution that makes sense. A quick calculation for the VLT:
FOV for the Nasmyth focus: 30'. With a seeing of 0.2" under ideal
circumstances, this gives roughly an equivalent of 9000x9000 pixels,
or 81Mpixels.

u19925 said:

For information, Keck telescopes (the world's largest ones) uses 8
MP CCDs as the biggest size. They achieve 8000x8000 using an array
of 8 CCD (each 2000x4000).

Click to expand...

That's 64Mpixels, isn't it?

u19925 said:

20x20 plates were used for surveys. I have seen 30x30 too. These
are Schidmt telescopes specially designed for these kind of work.
As I said, there are very few applications where these are used.

Click to expand...

Including some of the hottest topics in astronomy: large redshift
cosmology, near Earth asteroid detection...

Vtie
http://www.pbase.com/vtie

Dan Frederiksen said:

the fact that they acknowledge this to the public means they likely
have better. probably a 12" full wafer sensor ~ 800MP.
first rule of government spending: why not

Click to expand...

What Adobe says:

Photoshop supports documents up to 300,000 pixels in either dimension and offers three file formats for saving documents with images greater than 30,000 pixels in either dimension. Keep in mind that most other applications, including older versions of Photoshop, can only support files up to 2 GB in size or images up to 30,000 pixels in either dimension.

Large Document Format (PSB)

Supports documents of any pixel size and any file size. All Photoshop features are preserved in PSB files. Currently, PSB files are only supported by Photoshop CS. In Preferences under File Handling, the Enable Large Document Format option must be selected before you can save files in PSB format.

I'm interested because, using PS 7, I had to build my largest panorama (34,412 X 1996) in two distinct parts...
Georges Lagarde

even though it will support it, photoshop CS can only use 1.777Gb of physical memory (in Windows). which means processing this will be a nightmare ...

jim

GURL said:

Dan Frederiksen said:

the fact that they acknowledge this to the public means they likely
have better. probably a 12" full wafer sensor ~ 800MP.
first rule of government spending: why not

Click to expand...

What Adobe says:
Photoshop supports documents up to 300,000 pixels in either
dimension and offers three file formats for saving documents with
images greater than 30,000 pixels in either dimension. Keep in mind
that most other applications, including older versions of
Photoshop, can only support files up to 2 GB in size or images up
to 30,000 pixels in either dimension.

Large Document Format (PSB)

Supports documents of any pixel size and any file size. All
Photoshop features are preserved in PSB files. Currently, PSB files
are only supported by Photoshop CS. In Preferences under File
Handling, the Enable Large Document Format option must be selected
before you can save files in PSB format.

I'm interested because, using PS 7, I had to build my largest
panorama (34,412 X 1996) in two distinct parts...
Georges Lagarde

Click to expand...

Bob Howland said:

I wouldn't be surprised if it topped $50,000.

Click to expand...

170000 Euros if I am not too mistaken. Some months ago I actually held one in my hand for a few precious seconds...

Peter

lrgfmt said:

even though it will support it, photoshop CS can only use 1.777Gb
of physical memory (in Windows). which means processing this will
be a nightmare ...

Click to expand...

True, but us panorama stitchers will sure be glad to not have a 30,000 pixel limit in one direction.

--
Ciao!

Joe

http://www.swissarmyfork.com

Um, where do you get 900 megapixels from? (300,000 squared would give 90 gigapixels. Obviously other limitations will come into play before you can ever begin to get close to that number.)

-Z-

GURL said:

Dan Frederiksen said:

the fact that they acknowledge this to the public means they likely
have better. probably a 12" full wafer sensor ~ 800MP.
first rule of government spending: why not

Click to expand...

What Adobe says:
Photoshop supports documents up to 300,000 pixels in either
dimension and offers three file formats for saving documents with
images greater than 30,000 pixels in either dimension. Keep in mind
that most other applications, including older versions of
Photoshop, can only support files up to 2 GB in size or images up
to 30,000 pixels in either dimension.

Large Document Format (PSB)

Supports documents of any pixel size and any file size. All
Photoshop features are preserved in PSB files. Currently, PSB files
are only supported by Photoshop CS. In Preferences under File
Handling, the Enable Large Document Format option must be selected
before you can save files in PSB format.

I'm interested because, using PS 7, I had to build my largest
panorama (34,412 X 1996) in two distinct parts...
Georges Lagarde

Click to expand...

Military products eventually make it into civilian world. I can't wait until we have hand holdable cameras with image quality that is as good as large format. 10 - 15 years maybe?

heavyweather said:

There are 22MP digital backs around for some time.

A raw file from a 85MP sensor would be like 70MB...big deal?
digital backs dump up to 500MB pictures on Notebooks.

In 5 years such a beast will hit MF...in 7 years it's in you 35mm.

BJL said:

Bill Dewey said:

Think of how much new business this would bring to Cray, for
post-processing. Or, if you believe the ads, the new Apple
"super-computer", the G5.

Click to expand...

Yes, in fact the new Virginia Tech 9,600 gigaflop supercomputer
consisting of 1100 dual G5's might be ideal for post-processing
with this sensor.

Click to expand...

--
******************************************
http://www.fotocommunity.de/pc/pc/mypics/40625

Click to expand...

Paul Stimac said:

Military products eventually make it into civilian world. I can't
wait until we have hand holdable cameras with image quality that is
as good as large format. 10 - 15 years maybe?

Click to expand...

If that is the case I can't wait until they give us the UFOs. Interstellar travel and unlimited energy. Of course a good camera to take pictures of the funny people will be nice : )