Digital SLR vs. dedicated Astro CCD Camera

Hello all,

I am thinking about getting into astro imaging, and
am wondering something...

The local club I belong to has an AISIG (Astro Imaging
Special Interest Group). The majority of the club members
that I have questioned about about getting started in
imaging use Digital SLRs, mostly acquired in the last
year or two. (In fact, only one member I have approached
has an ST-7, and that is a few years old).

These one-shot colour DSLRs are pretty popular, and
are obviously taking a *lot* of business away from
dedicated astro camera manufacturers such as SBIG
and Starlight-Xpress. The dedicated astro cameras
are *seemingly* much more expensive per newbie useful
chip unit area, even taking into account the greater
sensitivity of the dedicated CCD cameras, cooling
capabilities, etc.

Why should a newbie such as myself pay $6000 for a
dedicated 6-MegaPixel astro CCD, or $1300 for a
low-end 0.4-MegaPixel astro CCD, when $1400 can buy a
less sensitive, less specific, but perfectly adequate
large-format dual-purpose CMOS camera such as Hutech-
converted Digital Rebel?

Gregory

On Mon, 04 Jul 2005 20:27:37 GMT, Gregory
wrote:

These one-shot colour DSLRs are pretty popular, and
are obviously taking a *lot* of business away from
dedicated astro camera manufacturers such as SBIG
and Starlight-Xpress.

I doubt they are having much impact on these companies. Maybe on their
one-shot color cameras, but those are just for newbies anyway, and don't
represent a large part of their business (not for SBIG, anyway). In
fact, I wouldn't be surprised if DSLRs are helping astrocamera
manufacturers, since they represent a cheap way into the hobby, but
don't provide much room for growth (so some users graduate to more
serious cameras).


Why should a newbie such as myself pay $6000 for a
dedicated 6-MegaPixel astro CCD, or $1300 for a
low-end 0.4-MegaPixel astro CCD, when $1400 can buy a
less sensitive, less specific, but perfectly adequate
large-format dual-purpose CMOS camera such as Hutech-
converted Digital Rebel?

It depends what you are after. One of the biggest mistakes beginning
imagers make is thinking they need lots of pixels. The actual number of
pixels you need is determined by how large a field you want to image. If
you are interested in imaging typical DSOs through a moderate focal
length instrument (like an SCT) there is little need for megapixels.
Also, you shouldn't underrate the value of cooling. A dedicated
astrocamera has _much_ better noise characteristics than even the lowest
noise DSLRs (Canons), and more than anything else it is noise that
determines image quality. Finally, color sensors do not generally
produce good results compared with individual B&W exposures made through
color filters. In short, if "adequate" is what you are shooting for, a
DSLR may serve you well. If your goal is to keep pushing your skills and
produce high quality astroimages, however, you will rapidly outgrow the
DSLR.

_________________________________________________

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

Hello again Chris and all,

Chris L Peterson wrote:


...In fact, I wouldn't be surprised if DSLRs are helping astrocamera
manufacturers, since they represent a cheap way into the hobby, but
don't provide much room for growth (so some users graduate to more
serious cameras).

I see...


...One of the biggest mistakes beginning imagers make is thinking
they need lots of pixels...

Oh...

The actual number of pixels you need is determined by how large a
field you want to image. If you are interested in imaging typical
DSOs through a moderate focal length instrument (like an SCT)
there is little need for megapixels.

OK -- I will be imaging with an f/5 300mm Newtonian. I was just
looking at DSLR images such as the following:

http://aisig.sdaa.org/astroblog/astr....asp?imgID=342

and thinking that I would be quite happy to produce such an image,
spending less than $2000 on the camera equipment.

Also, you shouldn't underrate the value of cooling...more than
anything else it is noise that determines image quality.

I see, I didn't know that...

Finally, color sensors do not generally produce good results
compared with individual B&W exposures made through
color filters. In short, if "adequate" is what you are shooting
for, a DSLR may serve you well.

If the image referenced above is "adequate", I guess that would
be OK with me for now...

If your goal is to keep pushing your skills and
produce high quality astroimages, however, you will
rapidly outgrow the DSLR.

Well, that *is* a goal, within my limited budget :-)

Thanks, Chris!

Gregory

On Mon, 04 Jul 2005 21:18:22 GMT, Gregory
wrote:

If the image referenced above is "adequate", I guess that would
be OK with me for now...

If your goal is to keep pushing your skills and
produce high quality astroimages, however, you will
rapidly outgrow the DSLR.

Well, that *is* a goal, within my limited budget :-)

Budget is important. If a dedicated astrocamera is out of reach at the
moment, that is a very good reason to go with a DSLR (I'd recommend an
unmodified 300D). You will get pleasing results and learn a lot about
astroimaging. One thing to be aware of is that working with the images
from a DSLR is a lot harder than working with either B&W or individual
color frames from an astrocamera. Calibration, noise removal, and color
processing with a color camera involve a lot of work. If you move on to
a more advanced camera one day, you will be pleased with how much easier
the processing is.

_________________________________________________

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

Chris L Peterson wrote:
If
you are interested in imaging typical DSOs through a moderate focal
length instrument (like an SCT) there is little need for megapixels.


Hi, Chris.

I'd disagree there; I switched to a DSLR because the small chips in the
usual dedicated CCD imager are far too small for typical DSO work at
focal lengths above 2000mm. In a 12" SCT even the DSLR can barely
accomodate globulars and smaller galaxies. The dedicated cameras are
fine with shorter scopes or very small objects like planetary nebulae.

John

On Tue, 05 Jul 2005 21:04:53 GMT, John C wrote:

Hi, Chris.

I'd disagree there; I switched to a DSLR because the small chips in the
usual dedicated CCD imager are far too small for typical DSO work at
focal lengths above 2000mm. In a 12" SCT even the DSLR can barely
accomodate globulars and smaller galaxies. The dedicated cameras are
fine with shorter scopes or very small objects like planetary nebulae.

That makes no sense to me. You match your pixel size to your optimal
resolution. With a long focal length instrument you generally need a
large sensor, but you don't need lots of pixels. If your sensor has
small pixels (say 7um in a 300D) , the longest focal length that makes
sense to use is around 1500mm (beyond that, you are probably
oversampling). What you should be doing is using a focal reducer to
increase your FOV; you probably won't be losing any resolution.

With a short focal length instrument, your pixel scale is large. If you
want a large FOV (which is a common goal with short focal length scopes)
you need lots of small pixels to avoid severe undersampling. That's why
megapixel cameras are useful here.

I use an ST8i for imaging. The sensor is 1K x 1.5K pixels, 9um. That
sensor covers a 14x20 arcsecond patch of sky when I use my 12" SCT at
2280mm focal length. That is plenty large enough for the vast majority
of DSOs. I'm oversampled at 0.8"/pixel, so I almost always bin the
sensor 2x2 and produce 512x768 pixel images. That's the same as saying
that I could be using a little ST7 on the same scope at an even shorter
focal length and still capturing all the available resolution for most
DSOs.

_________________________________________________

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

In article ,
Chris L Peterson wrote:
On Tue, 05 Jul 2005 21:04:53 GMT, John C wrote:

Hi, Chris.

I'd disagree there; I switched to a DSLR because the small chips in the
usual dedicated CCD imager are far too small for typical DSO work at
focal lengths above 2000mm. In a 12" SCT even the DSLR can barely
accomodate globulars and smaller galaxies. The dedicated cameras are
fine with shorter scopes or very small objects like planetary nebulae.

That makes no sense to me. You match your pixel size to your optimal
resolution. With a long focal length instrument you generally need a
large sensor, but you don't need lots of pixels. If your sensor has
small pixels (say 7um in a 300D) , the longest focal length that makes
sense to use is around 1500mm (beyond that, you are probably
oversampling). What you should be doing is using a focal reducer to
increase your FOV; you probably won't be losing any resolution.

With a short focal length instrument, your pixel scale is large. If you
want a large FOV (which is a common goal with short focal length scopes)
you need lots of small pixels to avoid severe undersampling. That's why
megapixel cameras are useful here.

I use an ST8i for imaging. The sensor is 1K x 1.5K pixels, 9um. That
sensor covers a 14x20 arcsecond patch of sky when I use my 12" SCT at
2280mm focal length. That is plenty large enough for the vast majority
of DSOs. I'm oversampled at 0.8"/pixel, so I almost always bin the
sensor 2x2 and produce 512x768 pixel images. That's the same as saying
that I could be using a little ST7 on the same scope at an even shorter
focal length and still capturing all the available resolution for most
DSOs.

I would think that you would want to match the Airy disk to the pixel
size. The size of the Airy Disk at the focal plane is based on the
focal ratio of the optical system, not the focal length. Using a
wavelength of 550nm, the diameter of the Airy Disk is 1342nm times the
focal ratio. For the 7.38 micron pixel in the 300D, I get an optimal
focal ratio of about 5.5. This seems to work with my 102mm f/5 and
80mm f/6, both seem to do well with my Rebel. The 80mm being a semi-
apo does a bit better; even though the theoretical Airy disk is smaller
in the 102, the blue wavelengths are focused better in the 80.

Rob Johnson
take out the trash before replying

On Wed, 06 Jul 2005 02:21:00 GMT, (Rob Johnson)
wrote:

I would think that you would want to match the Airy disk to the pixel
size. The size of the Airy Disk at the focal plane is based on the
focal ratio of the optical system, not the focal length. Using a
wavelength of 550nm, the diameter of the Airy Disk is 1342nm times the
focal ratio. For the 7.38 micron pixel in the 300D, I get an optimal
focal ratio of about 5.5. This seems to work with my 102mm f/5 and
80mm f/6, both seem to do well with my Rebel. The 80mm being a semi-
apo does a bit better; even though the theoretical Airy disk is smaller
in the 102, the blue wavelengths are focused better in the 80.

Maybe if you are in space, but on the Earth it is rare indeed for your
performance to be limited by diffraction. Seeing is almost always a much
more significant effect, and that is what you should normally base your
image scale on. Basically, you make an image of stars, measure the FWHM,
and then calculate your optimal scale from that. In most cases, the FWHM
of a star will be considerably larger than the Airy disk.

_________________________________________________

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

Chris L Peterson wrote:
On Tue, 05 Jul 2005 21:04:53 GMT, John C wrote:


.


That makes no sense to me. You match your pixel size to your optimal
resolution. With a long focal length instrument you generally need a
large sensor, but you don't need lots of pixels.
Chris, if I purchased a camera for my C11 for DSOs to use without a
focal reducer, then would my best option be the ST9 by SBIG ? Thanks.

On Thu, 07 Jul 2005 17:18:13 GMT, "David G. Fitzgerald"
wrote:

Chris, if I purchased a camera for my C11 for DSOs to use without a
focal reducer, then would my best option be the ST9 by SBIG ? Thanks.

The ST9 is a good match to those optics. But I'd probably look instead
at an ST8, since it has a somewhat larger sensor (so larger FOV). With
the ST8 you can bin 2x2 to get effective 18um pixels, about the same as
the ST9's 20um pixels, but you still have the ability to use the smaller
pixels if you run into a session with extremely good seeing, or if you
want to use the same camera on another scope of shorter focal length.
The ST8 is 30% more expensive than the ST9, but gives you a lot more
versatility.

_________________________________________________

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