Astro CCDs still dragging their pixel feet

Compared to consumer digital cameras, astro CCD cams
are still WAY behind when it comes to the number of
pixels. The latest one is the Meade DSI Pro, an "improvement"
on the still new DSI, that boasts "4x" the resolution
of the DSI. Here is the spec;
Specifications
CCD Sensor: High Sensitvity Sony® ExView HAD™ Monochrome CCD Sensor
Pixels: 510 x 492 pixels (250,000 pixels)

That's about 1/160th the count of the average 4 megapixel entry-level
consumer digicam. In fact, the colour HAD sensor is used in consumer
video cams because they can't process 30fps (video) if the pixel count
is much over 1 million.

Meade mentions that it's capable of exposures of over 1 hour, but
they intend it to be used to make very short exposures then combine
them for the finished result because an hour on a non-cooled digital
camera would be a disaster. They do offer an RGB filter set with it
for true colour reproduction.

But I can't help wondering when a company will offer a 4 megapixel or
larger CCD or CMOS-based astro cam with cooling for a price far below
that of the SBIG and other high priced cameras now on the market.
When they can sell a consumer camera with 4 megapixels for $200, how
expensive can the CCD/CMOS element possibly be, even in small
quantities??
-Rich

"RichA" wrote in message
...
Compared to consumer digital cameras, astro CCD cams
are still WAY behind when it comes to the number of
pixels. The latest one is the Meade DSI Pro, an "improvement"
on the still new DSI, that boasts "4x" the resolution
of the DSI. Here is the spec;
Specifications
CCD Sensor: High Sensitvity Sony® ExView HADT Monochrome CCD Sensor
Pixels: 510 x 492 pixels (250,000 pixels)

That's about 1/160th the count of the average 4 megapixel entry-level
consumer digicam. In fact, the colour HAD sensor is used in consumer
video cams because they can't process 30fps (video) if the pixel count
is much over 1 million.

Meade mentions that it's capable of exposures of over 1 hour, but
they intend it to be used to make very short exposures then combine
them for the finished result because an hour on a non-cooled digital
camera would be a disaster. They do offer an RGB filter set with it
for true colour reproduction.

But I can't help wondering when a company will offer a 4 megapixel or
larger CCD or CMOS-based astro cam with cooling for a price far below
that of the SBIG and other high priced cameras now on the market.
When they can sell a consumer camera with 4 megapixels for $200, how
expensive can the CCD/CMOS element possibly be, even in small
quantities??
-Rich

I think the issue has more to do with the purpose of the CCD system more
than anything else. Consumer digital cameras really aren't designed for
astro use, despite many using them for that purpose these days. Therefore,
little measures are taken to suppress/ eliminate noise. With an astro CCD
camera, just adding a peltier cooler is a significant cost plus the fact
that you're taking exposures significantly longer than any consumer cam.
When the CCD must be low noise and high gain, even with long exposures,
you're talking $. I think sooner or later though, as more and more
astrophotographers are making the jump to consumer digicams instead of astro
CCD cams, the astro CCD firms are going to have to drop prices to stay in
business.

Able

RichA wrote:
Compared to consumer digital cameras, astro CCD cams
are still WAY behind when it comes to the number of
pixels. The latest one is the Meade DSI Pro, an "improvement"
on the still new DSI, that boasts "4x" the resolution
of the DSI. Here is the spec;
Specifications
CCD Sensor: High Sensitvity Sony=AE ExView HAD=99 Monochrome CCD
Sensor
Pixels: 510 x 492 pixels (250,000 pixels)

That's about 1/160th the count of the average 4 megapixel entry-level
consumer digicam. In fact, the colour HAD sensor is used in consumer
video cams because they can't process 30fps (video) if the pixel
count
is much over 1 million.

Meade mentions that it's capable of exposures of over 1 hour, but
they intend it to be used to make very short exposures then combine
them for the finished result because an hour on a non-cooled digital
camera would be a disaster. They do offer an RGB filter set with it
for true colour reproduction.

But I can't help wondering when a company will offer a 4 megapixel or
larger CCD or CMOS-based astro cam with cooling for a price far below
that of the SBIG and other high priced cameras now on the market.
When they can sell a consumer camera with 4 megapixels for $200, how
expensive can the CCD/CMOS element possibly be, even in small
quantities??
-Rich

SAC imaging (http://www.sac-imaging.com/) their newest camera seems to
be what you are looking for.
Consumer cameras don't make good astro cameras because don't need
to pay as much attention to things like noise, contrast and efficiency
given that they are used in what we astronomers would consider well-lit
conditions. As you shrink the pixels you need to sacrifice these. It is
quite reasonable to make a large CCD sensor but you should remember the
price of a chip is based almost exclusively on how much silicon it
uses.

Simply adapting a consumer digital camera is possible, and if properly
cooled could produce some respectable images, but is wont be in
anywhere near the same league as SBIG, FLI, or Apogee.

As for the Mead DSI line, I think they provide a good entry for
potential imagers who don't know if they want to jump into a $3000
purchase. The choice not to install a TEC was because the cooling
system would cost more than camera itself and require an external power
supply. Either that or meade is looking to patent convective cooling.

Ian Anderson
www.customopticalsystems.com

Not entirely sure that more pixels is such a good thing. The PC would
likely become the new "weak link" as processing multi-megapixel images
takes a LOT more beef than that required for processing a "mere"
megapixel or so.

When memory limits are being reached, I find it faster to process four
images and mosaic them than to process a single larger image in one go
(for things like deconvolution in particular). Image download times are
a factor too...

Clearly, more memory, more CPU and more comms bandwidth can cure this.
But for the "standard equipment" available to amateurs, I reckon a
megapixel (SXV-H9 in my case) balances the various tradeoffs pretty
well.

Cheers
Beats

"RichA" wrote in message
...
Compared to consumer digital cameras, astro CCD cams
are still WAY behind when it comes to the number of
pixels. The latest one is the Meade DSI Pro, an "improvement"
on the still new DSI, that boasts "4x" the resolution
of the DSI. Here is the spec;
Specifications
CCD Sensor: High Sensitvity Sony® ExView HADT Monochrome CCD Sensor
Pixels: 510 x 492 pixels (250,000 pixels)

That's about 1/160th the count of the average 4 megapixel entry-level
consumer digicam. In fact, the colour HAD sensor is used in consumer
video cams because they can't process 30fps (video) if the pixel count
is much over 1 million.

Meade mentions that it's capable of exposures of over 1 hour, but
they intend it to be used to make very short exposures then combine
them for the finished result because an hour on a non-cooled digital
camera would be a disaster. They do offer an RGB filter set with it
for true colour reproduction.

But I can't help wondering when a company will offer a 4 megapixel or
larger CCD or CMOS-based astro cam with cooling for a price far below
that of the SBIG and other high priced cameras now on the market.
When they can sell a consumer camera with 4 megapixels for $200, how
expensive can the CCD/CMOS element possibly be, even in small
quantities??
-Rich
Smaller pixels = less photons gathered per pixel...
This is not a 'good thing' for deep sky imaging. Ideally, you want pixels
that are as large as possible, within what is sensible to get good spatial
'sampling' of the incoming image.
The big problem though in cost, is the difference in technology. The '4
megapixels for $200' cameras, are using CMOS sensors (noisier than CCD's),
with the ADC built into the chip, and usually lmited to perhaps 10bit
resolution, with only about 7 bits really useable, given the noise levels
on the chip. A CCD sensor, is less noisy, but more expensive. Add a 16bit
ADC, and price again rises (though the Meade DSI, has a 16bit ADC, the
actual circuit noise, limits the useable range to only perhaps 11 to
12bits - still an improvement over the webcams. Unfortunately, each extra
useable bit involves not only improvements in the ADC itself, but
reduction in the noise from every source (thermal, electrical in the
camera etc..), and bigger 'wells' for the electrons too.
For a small pixel camera, if you look at the Starlight M8C, this is a
typical consumer CCD, improved with Peltier cooling, and a better ADC. The
useable 'real' ADC resolution, is limited to perhaps between perhaps about
12, and 13bits, and this from a chip that still costs nearly 3* your
'$200' price. This will be a perfect 'planetary' camera, where the small
pixels will still gather plenty of light, and will also suit short focal
length camera lenses, for relatively widefield imaging.
If you want the same resolution, as you can get from cheap cameras, then
simply attach one to the scope. However if you want better signal to noise
performance, then this comes at a cost, and the cost rises significantly
for each extra 'bit' of useable ADC range.
With 'mono' cameras, small pixels are less of a problem, since the pixels
can be 'binned' in the camera, to behave like larger pixels.
Unfortunately, this cannot be done with the 'one shot' colour cameras, so
you are stuck with them...

Best Wishes

justbeats wrote:
Not entirely sure that more pixels is such a good thing. The PC would
likely become the new "weak link" as processing multi-megapixel
images
takes a LOT more beef than that required for processing a "mere"
megapixel or so.

When memory limits are being reached, I find it faster to process
four
images and mosaic them than to process a single larger image in one
go
(for things like deconvolution in particular). Image download times
are
a factor too...

Clearly, more memory, more CPU and more comms bandwidth can cure
this.
But for the "standard equipment" available to amateurs, I reckon a
megapixel (SXV-H9 in my case) balances the various tradeoffs pretty
well.

Cheers
Beats

Thanks to the STL 1100 it has been proven possible to work with 11
megapixel photos. CPU power is not the limiting factor in photo
processing so long as you are willing to wait for it to execute a
command (which admittedly could take awhile). RAM is the most important
factor followed by hard disk speed and volume. It is now within reason
to have a computer with 3 gigabytes of ram and over a terabyte of hard
drive space for less than the cost of a good camera.

Ian Anderson
www.customopticalsystems.com

It seems to me that the questioner, no matter what his open or hidden intentions
were, is operating under the paradigm that if the heart of the technology is
similar (CCDs, film, et al), then what applies to one particular use should
apply to another.

Given that technology even in the film era of Astronomy got somewhat complex,
leaving a lot of laymen clueless as to why certain things were done the way they
were, digital imaging should be a bit worse.

Remember film? There were dozens (at least) out there just from one
manufacturer alone, since no one film can handle any and all situations. There
were even emulsions developed by Kodak just for use by astronomers. And then
there was the use of gases to hypersensitize the response, as well as cold
cameras.

There is a similar divide between CCD/CMOS imagers for regular everyday use, and
those for the special conditions for astronomical imaging. As one gets more
into this specific area of interest, the more one gets educated as to why things
are done the way they are, and why it's different from everyday imaging.

There are a lot of people working with CCDs, as part of companies, universities
and colleges, and as individuals, in the astronomy community. I'm sure they've
looked into, and are looking into, whether the newer CCDs and CMOS imagers can
be used. If and when they determine that something is useful, the rest of us
will know about it in a few months time.

--- Dave
--
----------------------------------------------------------------------
Pinprick holes in a colorless sky
Let inspired figures of light pass by
The Mighty Light of ten thousand suns
Challenges infinity, and is soon gone




"Roger Hamlett" wrote in message
...

"RichA" wrote in message
...
Compared to consumer digital cameras, astro CCD cams
are still WAY behind when it comes to the number of
pixels. The latest one is the Meade DSI Pro, an "improvement"
on the still new DSI, that boasts "4x" the resolution
of the DSI. Here is the spec;
Specifications
CCD Sensor: High Sensitvity Sony® ExView HADT Monochrome CCD Sensor
Pixels: 510 x 492 pixels (250,000 pixels)

That's about 1/160th the count of the average 4 megapixel entry-level
consumer digicam. In fact, the colour HAD sensor is used in consumer
video cams because they can't process 30fps (video) if the pixel count
is much over 1 million.

Meade mentions that it's capable of exposures of over 1 hour, but
they intend it to be used to make very short exposures then combine
them for the finished result because an hour on a non-cooled digital
camera would be a disaster. They do offer an RGB filter set with it
for true colour reproduction.

But I can't help wondering when a company will offer a 4 megapixel or
larger CCD or CMOS-based astro cam with cooling for a price far below
that of the SBIG and other high priced cameras now on the market.
When they can sell a consumer camera with 4 megapixels for $200, how
expensive can the CCD/CMOS element possibly be, even in small
quantities??
-Rich
Smaller pixels = less photons gathered per pixel...
This is not a 'good thing' for deep sky imaging. Ideally, you want pixels that
are as large as possible, within what is sensible to get good spatial
'sampling' of the incoming image.
The big problem though in cost, is the difference in technology. The '4
megapixels for $200' cameras, are using CMOS sensors (noisier than CCD's),
with the ADC built into the chip, and usually lmited to perhaps 10bit
resolution, with only about 7 bits really useable, given the noise levels on
the chip. A CCD sensor, is less noisy, but more expensive. Add a 16bit ADC,
and price again rises (though the Meade DSI, has a 16bit ADC, the actual
circuit noise, limits the useable range to only perhaps 11 to 12bits - still
an improvement over the webcams. Unfortunately, each extra useable bit
involves not only improvements in the ADC itself, but reduction in the noise
from every source (thermal, electrical in the camera etc..), and bigger
'wells' for the electrons too.
For a small pixel camera, if you look at the Starlight M8C, this is a typical
consumer CCD, improved with Peltier cooling, and a better ADC. The useable
'real' ADC resolution, is limited to perhaps between perhaps about 12, and
13bits, and this from a chip that still costs nearly 3* your '$200' price.
This will be a perfect 'planetary' camera, where the small pixels will still
gather plenty of light, and will also suit short focal length camera lenses,
for relatively widefield imaging.
If you want the same resolution, as you can get from cheap cameras, then
simply attach one to the scope. However if you want better signal to noise
performance, then this comes at a cost, and the cost rises significantly for
each extra 'bit' of useable ADC range.
With 'mono' cameras, small pixels are less of a problem, since the pixels can
be 'binned' in the camera, to behave like larger pixels. Unfortunately, this
cannot be done with the 'one shot' colour cameras, so you are stuck with
them...

Best Wishes

On Wed, 13 Apr 2005 21:24:16 GMT, "David Nakamoto"
wrote:

It seems to me that the questioner, no matter what his open or hidden intentions
were, is operating under the paradigm that if the heart of the technology is
similar (CCDs, film, et al), then what applies to one particular use should
apply to another.

Given that technology even in the film era of Astronomy got somewhat complex,
leaving a lot of laymen clueless as to why certain things were done the way they
were, digital imaging should be a bit worse.
"Hidden intentions?" Blah, blah, blah.
GO ask professionals WHY they use
hundred-megapixel cameras composed of banks
of large CCDs. I think the CFH scope has a 340 MEG CCD
array they used. They are TRYING to duplicate the
resolution of film plates while gaining the photon
efficiency of CCDs. When was the last time you saw
ANY observatory working in the visible spectrum using
a 250,000 element CCD except maybe for planets???
The people who used to shoot medium format film
in astrographs would benefit from a large, multi-pixel
CCD.
-Rich

On Thu, 14 Apr 2005 00:55:04 -0400, RichA wrote:

GO ask professionals WHY they use
hundred-megapixel cameras composed of banks
of large CCDs. I think the CFH scope has a 340 MEG CCD
array they used.

The main reason is because they use telescopes with very long focal
lengths, which means they need large sensors. In many cases they don't
need lots of pixels, but there are practical limits to the size you can
make a pixel. Most professional telescopes are oversampled. Outside of
surveys, most professionals are working with fairly small images- a few
hundred pixels square is very common (even if this was isolated from a
larger sensor).


They are TRYING to duplicate the
resolution of film plates while gaining the photon
efficiency of CCDs.

Virtually all CCDs already exceed the spatial resolution of film. Large
amateur CCD cameras (over about 5MP) exceed 35mm film both in resolution
and in potential image scale.

The size of detector you need is dictated by the FOV you are after. Most
wide field imagers want sensors with a physical size approaching 35mm
film and a pixel count of 5-15MP. Such cameras are available- not cheap,
but not out of range of many amateur imagers. DSLRs (especially Canons)
also do a very good job here, although with some penalties do to noise
and the color sensor.

People (like myself) who image individual objects rarely require more
than 1-2 MP- any more is simply wasted area.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

GO ask professionals WHY they use
hundred-megapixel cameras composed of banks
of large CCDs. I think the CFH scope has a 340 MEG CCD
array they used.

the CFHT megaprime is comprised of a mosaic of E2V CCD42-90 (8 mega
pixel each) sensors which is esentially 2 42-40 sensors available from
FLI or apogee.