PDA

View Full Version : Good CCD Camera for 10k-15k Euro


David Y.
February 26th 04, 11:30 AM
Hi!

I am looking for a CCD camera in the price range of 10.000 - 15.000 Euro.

I was told that SBIG produces good cameras.
But I am still searching for more options or other producers respectively.

Are there any cameras that you guys could recommend?

What do I have to watch out for? What are common problems of cameras in
this price range (if there still are any ;-) )?

I appreciate any serious answer or additional links.

Thanks!

David

Fred
February 26th 04, 11:56 AM
David Y. wrote:

> Hi!
>
> I am looking for a CCD camera in the price range of 10.000 - 15.000 Euro.
>
> I was told that SBIG produces good cameras.

What is good? I can sell you almost anything for 15K...

David Y.
February 26th 04, 12:08 PM
Fred wrote:

> What is good? I can sell you almost anything for 15K...
>

There are some camera types which have a guider camera included (which I
also would need).
But the disadvantage of those included guider cameras is that they
stress the CCD chip.

Is this disadvantage still given within my price range?

What software is included?

How is the image quality?

I don't want to buy the pig in a poke so I try to find out which cameras
are good or rather with which cameras people have made the best experiences.

What do I have to watch out for concerning the chip?

I hope this helps to understand more what I actually want ;-)

David

Roger Hamlett
February 26th 04, 12:10 PM
"David Y." > wrote in message
...
> Hi!
>
> I am looking for a CCD camera in the price range of 10.000 - 15.000 Euro.
>
> I was told that SBIG produces good cameras.
> But I am still searching for more options or other producers respectively.
>
> Are there any cameras that you guys could recommend?
>
> What do I have to watch out for? What are common problems of cameras in
> this price range (if there still are any ;-) )?
>
> I appreciate any serious answer or additional links.
>
> Thanks!
>
> David
What do you want from the camera?.
What focal length are you intending to use?.
What do you want to image?.
What mount do you have?.
The SBIG cameras, are excellent units. They have the advantage of a built in
OAG, which allows the guider to issue corrections for any changes in the
optical system inside the scope (hence mirror shift in SCT's). This also
allows them to be used with the simple 'first order' adaptive optic system
(the AO7). This really has more to do with rapid guide correction, than
correction for the atmosphere, but can allow good images to be generated on
mounts that are otherwise borderline for imaging.
StarLight do an excellent range of alternatives, with either their 'single
chip' guiding solution on the MX cameras, or a seperate guide camera (on the
SVX). They compete directly with the smaller SBIG models (and there is a new
large format colour unit coming, that might represent an excellent choice if
your target is 'pictorial imaging'). The advantage in Europe especially, is
that their prices are much more agressive (though the recent week dollar
should help SBIG here).
The seperate guide camera, is the 'better' solution, if you want to image
using very narrow bandpass filters, since the guider can be left
'unfiltered'.
Most of the SBIG cameras, have a mechanical shutter, that limits the minimum
exposure that can be used (but makes taking dark-frames, and absolute
'doddle').
Many mounts that would be considered adequate for visual work, will be
inadequate for imaging.
Large CCD's, need a large flat field from the scope (an SCT, has very
significant field curvature, and correction will be needed).
If your scope has a large flat field, and you guide with a seperate guider
(the SBIG STV, is probably the best 'standalone' guider made), then you can
consider cameras that have no guide ability.
You could look at 'Finger lakes' (http://www.fli-cam.com/), or Apogee
Instruments (http://www.optronis.com/docs/apogee.htm).
Remember that larger pixels collect more light, so having large numbers of
small pixels on a long focal length scope, is a waste (this is why
'professional' cameras, often have a quite small pixel count of larger
pixels). Also larger pixels are easier to make with a larger well-depth.
Hence camera selection, needs to be adjusted according to the scope to be
used.
Similarly, there are other camera 'features', that may be good/bad,
according to what you actually want to do. A camera with ABG, is far easier
to use for 'pictorial' imaging, where bright objects are close to the target
object, 'clipping off' the signal at the bright end. The 'downside', is that
for photometric measurements, this clipping introduces some non-linearity in
the top part of the cameras response (this is probably least on the Sony
chips used in the StarLight cameras).
Conversely, NABG cameras have better linearity, but then have their data
destroyed close to bright objects, by blooming.
Remember that there will probably be a significant number of 'extras' with
any camera you consider. Accurate focussing becomes critical with a CCD, and
a system to provide this, is an extra cost. There are also things like
colour wheels/filters, and perhaps focal compressors and/or field flatteners
needed, depending on the CCD chosen.

Best Wishes

Roger Hamlett
February 26th 04, 12:22 PM
"David Y." > wrote in message
...
> Fred wrote:
>
> > What is good? I can sell you almost anything for 15K...
> >
>
> There are some camera types which have a guider camera included (which I
> also would need).
> But the disadvantage of those included guider cameras is that they
> stress the CCD chip.
'Stress the chip'!...
The ones with a seperate guide CCD, do nothing at all to the main chip. Ones
using 'single chip' guiding (the Starlight MX models), do not 'stress' the
chip (the life is massive), but do generate a tiny amount of heat on the
chip, which results in an increase in background noise when using this
feature.
Seriously, a CCD, on a video camera, is being read every 25th/30th second.
How often do you hear of these failing?. Electronic parts if properly
designed, will not suffer from being used as they are meant to be used.

> Is this disadvantage still given within my price range?
There are dozens of cameras in your 'price range', offering just about every
feature imaginable. The question is which one will actually suit your scope.

> What software is included?
Depends on the cameras. Most have basic imaging solutions. SBIG, have a
fairly good 'suite', that covers most basic imaging. In most cases, extra
software does allow more powerful operations.

> How is the image quality?
Depends on your scope...

> I don't want to buy the pig in a poke so I try to find out which cameras
> are good or rather with which cameras people have made the best
experiences.
>
> What do I have to watch out for concerning the chip?
I'd really suggest perhaps buying Ron Wodawski's book on CCD imaging,
_before_ looking any further. You need to understand how the cameras
features interact with your scope, what large chips, guiding etc.,
involve/entail.

> I hope this helps to understand more what I actually want ;-)
Not really. I think you need to learn more before making a decision.
In your 'price range', with a bit of careful purchasing, you could (for
instance), buy an SBIG ST10XME, with a CFW8, and an AO7. On something like a
700-800mm apochromatic refractor, this probably represents many peoples idea
of an 'ideal' camera. However it still would not suit everybody, and there
may well be cheaper solutions, that would suit you better.

Best Wishes

David Yasli
February 26th 04, 09:26 PM
Roger Hamlett wrote:
> "David Y." > wrote in message
> ...
>
>>Fred wrote:
>>
>>
>>>What is good? I can sell you almost anything for 15K...
>>>
>>
>>There are some camera types which have a guider camera included (which I
>>also would need).
>>But the disadvantage of those included guider cameras is that they
>>stress the CCD chip.
>
> 'Stress the chip'!...
> The ones with a seperate guide CCD, do nothing at all to the main chip. Ones
> using 'single chip' guiding (the Starlight MX models), do not 'stress' the
> chip (the life is massive), but do generate a tiny amount of heat on the
> chip, which results in an increase in background noise when using this
> feature.
> Seriously, a CCD, on a video camera, is being read every 25th/30th second.
> How often do you hear of these failing?. Electronic parts if properly
> designed, will not suffer from being used as they are meant to be used.
>
>
>>Is this disadvantage still given within my price range?
>
> There are dozens of cameras in your 'price range', offering just about every
> feature imaginable. The question is which one will actually suit your scope.
>
>
>>What software is included?
>
> Depends on the cameras. Most have basic imaging solutions. SBIG, have a
> fairly good 'suite', that covers most basic imaging. In most cases, extra
> software does allow more powerful operations.
>
>
>>How is the image quality?
>
> Depends on your scope...
>
>
>>I don't want to buy the pig in a poke so I try to find out which cameras
>>are good or rather with which cameras people have made the best
>
> experiences.
>
>>What do I have to watch out for concerning the chip?
>
> I'd really suggest perhaps buying Ron Wodawski's book on CCD imaging,
> _before_ looking any further. You need to understand how the cameras
> features interact with your scope, what large chips, guiding etc.,
> involve/entail.
>
>
>>I hope this helps to understand more what I actually want ;-)
>
> Not really. I think you need to learn more before making a decision.
> In your 'price range', with a bit of careful purchasing, you could (for
> instance), buy an SBIG ST10XME, with a CFW8, and an AO7. On something like a
> 700-800mm apochromatic refractor, this probably represents many peoples idea
> of an 'ideal' camera. However it still would not suit everybody, and there
> may well be cheaper solutions, that would suit you better.
>
> Best Wishes
>
>
Hey,

thanks for the book advise. I will try and get it somewhere.

And the CCD cam types will help me as well.

Cu!

David

Thomas Juerges
February 27th 04, 08:24 AM
Dear Roger, dear David,

On Thu, 26 Feb 2004 12:10:39 +0000, Roger Hamlett wrote:
> your target is 'pictorial imaging'). The advantage in Europe especially,
> is that their prices are much more agressive (though the recent week
> dollar should help SBIG here).

As for Germany I can tell you that the "weak" dollar does nothing to the
exorbitant prices one has to pay for the SBIG or Finger Lakes
Instrumentation CCD cameras if you buy from an official dealer here. By
the way: Our department considered to buy another ST9E from the
Netherlands at a price of 1000 Euros(!!!) less than the price given here
at Baader company which is the only official SBIG dealer in Germany.

> Most of the SBIG cameras, have a mechanical shutter, that limits the
> minimum exposure that can be used (but makes taking dark-frames, and
> absolute 'doddle').

Acknowledge: The SBIG cameras with mechanical shutter use a "thing" which
has two blades rotating on an axis of which one closes the window at a
time (like a propeller). The resulting minimum exposure time until you see
the shutter in your images is >0.1 seconds. The Finger Lakes MaxCam with
Kodak KAF0401 CCD (forgot the exact model number, sorry) uses a shutter as
known from photographic camera objektives (radial lamellas) and has a
minimum exposure time >0.01 seconds. But, you are right again, one has to
consider the pixel size (on the same telescope) when talking about minimum
and maximum exposure times. For our ST9E it is 20*20 m, for the FLI it
it 9 m.

I like the FLI shutter more due to the construction which has served in
millions objectives very well. But the FLI has a big disadvantage. The
ST9E comes at -20C with a readout noise of approximately 135 counts at
16 Bits. The FLI comes at the same temperature with approx. 4170(!) counts
at 16 Bits.

> Many mounts that would be considered adequate for visual work, will be
> inadequate for imaging.

Again, I must acknowledge. Our LX200 16" mount is a pain. It could never -
since it has been set up and has been repaired twice - track an
exposure for more than 30 seconds.

> Large CCD's, need a large flat field from the scope (an SCT, has very
> significant field curvature, and correction will be needed).

This problem could be solved by using dome flats.

> according to what you actually want to do. A camera with ABG, is far
> easier to use for 'pictorial' imaging, where bright objects are close to
> the target object, 'clipping off' the signal at the bright end. The
> 'downside', is that for photometric measurements, this clipping
> introduces some non-linearity in the top part of the cameras response
> (this is probably least on the Sony chips used in the StarLight
> cameras). Conversely, NABG cameras have better linearity, but then have
> their data destroyed close to bright objects, by blooming. Remember that

I consider the ABG feature of some cameras as a widget which is
not really usable. Think of usual photographic emulsions, I have to time
my exposure exactly to get the best out of them. The same is obviously
true for CCDs to use their advantage of linearity. ABG does not show up
with "some" non-linearity as you said, Roger, but it is completely
non-linear. Everytime the full well capacity for a pixel is reached, some
amount of it is dumped and the pixel is dimmed. The result is that
"overexposed" pixel cannot be compared in brightness to their not
overexposed neighbours, not to mention that one cannot compare these
exposures to exposures taken with other cameras.

As for the extras: I suggest you buy a camera which has a USB interface
for data transfer. Any chip size greater than 512*512 pixels makes it
awfully to wait until the complete exposure is tranferred via parallel
port into the computer. By the way: Does anyone know why SBIG introduced a
USB 1.1 interface with their semi-research grade cameras (STL series)? The
cameras come with up to 10 million pixels and it takes a while to tranfer
them via USB 1.1 to the computer. USB 2.0 should not be too expensive and
it is for sure not very complicated to implement a USB 2.0 but a 1.1
controller into the camera.

Another point I would like to mention: Our department develops the
camera software ourselves. Therefore it is necessary that the manufacturer
supports you with some kind of drivers or libraries to set up the
software. On the one hand it is nice to have some Win software which can
do everything the camera is capable of but if you are not going to do
consumer stuff, you have to rely on the manufacturer's support. The
example I can give is that SBIG and FLI provide driver sets or libraries
(ST9E/STL, MaxCam) which are usable with Linux. SBIG has added MacOS X
support some weeks ago, too.

David Yasli
February 27th 04, 12:42 PM
Hi Roger,

thank you for your long explanation.

> What do you want from the camera?.

The camera is planned to be installed on a reflector telescope with a
diameter of 50-60 cm.

> What do you want to image?.

The pictures which will be taken are going to be of the sun and light
objects at night.


> What mount do you have?.

Have not decided yet. It still is to be planned.


But I think I understand that in order to decide for the correct camera
I first need to think about the telescope which is being used.

David

David Yasli
February 27th 04, 12:43 PM
Oh, and sorry Thomas!

Of course thanks for your reply, too!!!

David

David Yasli wrote:
> Hi Roger,
>
> thank you for your long explanation.
>
> > What do you want from the camera?.
>
> The camera is planned to be installed on a reflector telescope with a
> diameter of 50-60 cm.
>
> > What do you want to image?.
>
> The pictures which will be taken are going to be of the sun and light
> objects at night.
>
>
> > What mount do you have?.
>
> Have not decided yet. It still is to be planned.
>
>
> But I think I understand that in order to decide for the correct camera
> I first need to think about the telescope which is being used.
>
> David

David Whysong
March 2nd 04, 08:11 PM
Thomas Juerges > wrote:

> I like the FLI shutter more due to the construction which has served in
> millions objectives very well. But the FLI has a big disadvantage. The
> ST9E comes at -20C with a readout noise of approximately 135 counts at
> 16 Bits. The FLI comes at the same temperature with approx. 4170(!) counts
> at 16 Bits.

That's not read noise, it must be the bias level. You can subtract it off
relatively cleanly.

> Another point I would like to mention: Our department develops the
> camera software ourselves. Therefore it is necessary that the manufacturer
> supports you with some kind of drivers or libraries to set up the
> software. On the one hand it is nice to have some Win software which can
> do everything the camera is capable of but if you are not going to do
> consumer stuff, you have to rely on the manufacturer's support. The
> example I can give is that SBIG and FLI provide driver sets or libraries
> (ST9E/STL, MaxCam) which are usable with Linux. SBIG has added MacOS X
> support some weeks ago, too.

If you're working on driving the newer SBIG cameras from Linux, I'd love to
hear about it. I'm thinking about doing that myself.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

Thomas Juerges
March 3rd 04, 12:58 PM
Dear David,

On Tue, 02 Mar 2004 20:11:34 +0000, David Whysong wrote:

> That's not read noise, it must be the bias level. You can subtract it off
> relatively cleanly.

Hmm, from the user's point of view one cannot determine if it is really
bias, i.e. due to too low pre biasing the pixel cells in order to remove
noise or if it is real read out noise. Please correct if I'm wrong.

Anyway, the performance is ok but the "bias/readout" noise. :)

> If you're working on driving the newer SBIG cameras from Linux, I'd love
> to hear about it. I'm thinking about doing that myself.

No problem, David. Watch the following link:
http://almasw.hq.eso.org/almasw/bin/view/Main/WebHome
-> ALMASW.ACS
-> ACS For (non-)ALMA projects: ACS contributed code

After next week I'll upload the final source code release for our
telescope project. The software is integrated into the ACS (ALMA Common
Software, a kind of abstraction and expansion of ACE & TAO) but it should
be an easy job for a programer to grab the main things from it. I'm using
the latest SBIG developer libraries, so that the (hopefully soon arriving)
STL should run flawlessly. But... I have not integrated the filterwheel
yet. This is scheduled for 3rd quarter this year.

Best regards,
Thomas

David Whysong
March 3rd 04, 10:28 PM
Thomas Juerges > wrote:
> Dear David,

> On Tue, 02 Mar 2004 20:11:34 +0000, David Whysong wrote:

>> That's not read noise, it must be the bias level. You can subtract it off
>> relatively cleanly.

> Hmm, from the user's point of view one cannot determine if it is really
> bias, i.e. due to too low pre biasing the pixel cells in order to remove
> noise or if it is real read out noise. Please correct if I'm wrong.

> Anyway, the performance is ok but the "bias/readout" noise. :)

Ok, maybe I'm confused...

If you take two short (<< 1 second) exposures, then subtract one from the
other, what is the RMS of the resulting image?

It should be something of order 10; if it's more like 4000, then something
is very wrong with your camera.

> After next week I'll upload the final source code release for our
> telescope project. The software is integrated into the ACS (ALMA Common
> Software, a kind of abstraction and expansion of ACE & TAO) but it should
> be an easy job for a programer to grab the main things from it. I'm using
> the latest SBIG developer libraries, so that the (hopefully soon arriving)
> STL should run flawlessly. But... I have not integrated the filterwheel
> yet. This is scheduled for 3rd quarter this year.

Thanks for the link! I had a look at the SBIG library, the STL cameras have
a very high level filter wheel interface so programming that should be easy
compared to (for example) autoguiding. Anyway I'll definitely have a look
at your software when it's posted, thanks.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu

Thomas Juerges
March 4th 04, 12:06 PM
Dear David,

On Wed, 03 Mar 2004 22:28:31 +0000, David Whysong wrote:
> Ok, maybe I'm confused...
>
> If you take two short (<< 1 second) exposures, then subtract one from the
> other, what is the RMS of the resulting image?
>
> It should be something of order 10; if it's more like 4000, then something
> is very wrong with your camera.

See the personal mail I sent you for my current time schedule. I'll keep
the mailing list informed.

> Thanks for the link! I had a look at the SBIG library, the STL cameras
> have a very high level filter wheel interface so programming that should
> be easy compared to (for example) autoguiding. Anyway I'll definitely
> have a look at your software when it's posted, thanks.

Yes, you are right. There is only a small but: They have implemented not
as much as they should have done. Some features are still missing as
a function to be called for repetitive exposures or reading out the
ambient temperature sensor as it is possible for ccd temperature. Although
it is not a big problem to put the needed functionality into the library
code.

Regards,
Thomas

David Whysong
March 5th 04, 08:00 AM
Thomas Juerges > wrote:

> Yes, you are right. There is only a small but: They have implemented not
> as much as they should have done. Some features are still missing as
> a function to be called for repetitive exposures or reading out the
> ambient temperature sensor as it is possible for ccd temperature. Although
> it is not a big problem to put the needed functionality into the library
> code.

That's too bad. I must say that I'm a little annoyed that the driver source
code isn't available. They have a binary-only module, which limits the
architecture and kernel compatability of their cameras. Since my development
computer is a dual Opteron, I can't load the driver.

Starlight Xpress cameras are extremely well supported under Linux; I tested
the HX916 support and wrote my own camera control software in Perl/PDL. I'd
like to be able to do the same sort of thing with SBIG cameras.

Dave

David Whysong
DWhysong (at) physics (dot) ucsb (dot) edu