Go Back   IceInSpace > Equipment > Astrophotography and Imaging Equipment and Discussions

Reply
 
Thread Tools Rate Thread
  #1  
Old 27-10-2011, 07:41 PM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
Further request for CCD camera advice.

Hi. I've been researching CCD cameras, learning CCD theory, asking questions and generally reading as much as I can, in deciding whether to upgrade. From what I can tell an OSC is the best choice for my application.

Other than price, the basic criteria seems to be QE, noise (including electronics), cooling and ease of use. So far, FLI and Apogee are of interest with KAI 8050 sensors 5.6um pixels which is the correct pixel size at 200mm and 400mm with 2x2 binning.

EDIT: This is derived from an online pixel size calculator.

Another option is the KAI 4020, 4.2mp interline transfer sensor 7.4um. These seem to have a good reputation, although pixel size is not quite right for the focal length I'm using.

EDIT: 7.4um is somewhere in the middle between 200 and 400mm.

Is pixel size going to be an issue as far as a visible difference in an image taken with a 4.2mp vs 8.3mp at the short focal lengths I'm using. I think not, but that's just a guess.

Does anyone have experience with these criteria to say one way or the other.

Last edited by rcheshire; 28-10-2011 at 01:37 PM.
Reply With Quote
  #2  
Old 27-10-2011, 07:44 PM
Poita (Peter)
Registered User

Poita is offline
 
Join Date: Jun 2011
Location: NSW Country
Posts: 3,586
What scope is it going onto, and what are your likely targets?
Reply With Quote
  #3  
Old 27-10-2011, 08:48 PM
Hagar (Doug)
Registered User

Hagar is offline
 
Join Date: Oct 2006
Location: Australia
Posts: 4,646
Just one thing to be aware of. A OSC camera uses a bayer mask in front of the CCD to capture and distinquish colour. Using a OSC camera at 2 x 2 binning or anything higher than 1 X 1 makes the images just black and white.
There are plenty of options out there so the end choice is yours alone. b You might find a cheaper alternative is a camera with either a Mono or colour KAF8300. This CCD has proven itself as probably the most sucessful, best priced CCD available today.

Good luck.

Quote:
Originally Posted by rcheshire View Post
I've been researching CCD cameras, learning CCD theory, asking questions and generally reading as much as I can. For my use an OSC is the best choice. I want better performance than a DSLR. Alternatively, I could modify my 1000D with an Astrodon inside filter... but the noise overhead and dark frame requirements are painfully time consuming.

I think the choice, other than price, is coming down to QE, noise, cooling and ease of use. So far, FLI and Apogee are of interest with KAI 8050 sensors 5.6um pixels. Ideal for 200mm and 400mm with 2x2 binning.

Another option is the 4.2mp interline transfer sensor 7.4um. These seem like extremely good sensors, although pixel size is not quite right for the focal length I'm using.

Bottom line, low noise sensor, low noise electronics, high relative QE, effective cooling, not complex to use.

Which to choose?
Reply With Quote
  #4  
Old 27-10-2011, 09:23 PM
LightningNZ's Avatar
LightningNZ (Cam)
Registered User

LightningNZ is offline
 
Join Date: Oct 2011
Location: Canberra
Posts: 951
Just don't hope to bin with KAF8300, you won't gain anything much. If you have a long focal length then there are better chips out there. Why would you want to bin 5.6um pixels with a focal length of only 200 or 400? At short focal lengths small pixels work out better.

Cheers,
Cam
Reply With Quote
  #5  
Old 27-10-2011, 10:09 PM
cventer's Avatar
cventer
Registered User

cventer is offline
 
Join Date: Jul 2005
Location: Melbourne Australia
Posts: 957
Why should you not bin a KAF8300 ?
Reply With Quote
  #6  
Old 27-10-2011, 10:29 PM
marki's Avatar
marki
Waiting for next electron

marki is offline
 
Join Date: Mar 2008
Posts: 2,427
What scope has a 200 mm FL besides a finderscope???? I think you have made a mistake somewhere? What is the scope you plan to put it on?

Mark

Last edited by marki; 28-10-2011 at 12:57 PM.
Reply With Quote
  #7  
Old 27-10-2011, 10:32 PM
marki's Avatar
marki
Waiting for next electron

marki is offline
 
Join Date: Mar 2008
Posts: 2,427
Quote:
Originally Posted by cventer View Post
Why should you not bin a KAF8300 ?
No reason if it is a mono, only affects OSC as Doug stated in post # 3. 2 x 2 bin is good for collecting colour quickly with the mono chip as you dont need resolution just signal. The stars saturate quickly and the end product is only 2MP but does the job. I have never tried to make up for long FL by binning the KAF8300, better to buy a camera with bigger pixels and deeper wells.

Mark
Reply With Quote
  #8  
Old 27-10-2011, 10:50 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by cventer View Post
Why should you not bin a KAF8300 ?
I have heard a couple of people suggest that binning a KAF8300 doesn't appear to give a significant increase in effective well depth. I haven't ever tried binning on my SX H18, but I might have a play with a light box on the weekend (if it's cloudy) and see if I can test this hypothesis.

Cheers,
Rick.
Reply With Quote
  #9  
Old 28-10-2011, 12:05 PM
Tandum's Avatar
Tandum (Robin)
Registered User

Tandum is offline
 
Join Date: Apr 2008
Location: Wynnum West, Brisbane.
Posts: 4,161
Quote:
Originally Posted by RickS View Post
I have heard a couple of people suggest that binning a KAF8300 doesn't appear to give a significant increase in effective well depth.
It uses non active standard pixels as a buffer to readout the image from the sensor so at 2x2 it sums 4 image pixels into one for readout.
Reply With Quote
  #10  
Old 28-10-2011, 02:04 PM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
I've tried to be more specific. Maybe what I'm asking is not clear, perhaps the answer is obvious, either way I'm using Canon lenses - no plans for a scope. Yep, I'm imaging with a finder scope and want to attach an expensive CCD to it. Is anyone else using camera lenses?

Thanks Doug, the binning issue is not something I understand as far as OSC's are concerned. I'm working on the assumption, pretty much, that there's not a lot of difference between 5.6 and 7.4um, except that the KAI 4020 is 7.4 has a great reputation, even at 4.2mp.

I'm not sure which way to go - the bigger pixel or more pixels, I think it's coming down to that, as a compromise across the two focal lengths. I'm not sure I have a complete grasp on the science yet to make an informed decision. It's not like I'm selecting football team.
Reply With Quote
  #11  
Old 28-10-2011, 11:08 PM
marki's Avatar
marki
Waiting for next electron

marki is offline
 
Join Date: Mar 2008
Posts: 2,427
On a OSC chip it is best to think in groups of 4 pixels in a square. What the manufactures do is use dyes (RGB) which they carefully place over the individual pixels in a set pattern (eg RGGB) to act as a filter for each colour similar to the way we do on mono cameras. When you download an image in 1 x 1 bin the ADU count for each pixel is stored in the file. With astro cameras these are often 16 bit images in the FITS format. These images are grey scale and need to be debayered in a suitable program (eg MaximDL, DSS or simlar) to retrieve the colour information. When you bin 2 x 2 that information is lost and the image cannot be debayered hence Dougs comment about grey scale images.

In your original post you did not mention you were using a camera lens hence the confusion. The KAF8300 would give you higher resolution at both focal lengths as the pixels are smaller but there would be little difference between the two chips you have suggested. I would go for the one that gives you the greatest FOV as you obviously want to do widefield AP. The KAF8300 is used by many at focal lengths much greater then 400mm so I doubt you would have any problems with either chip in the range you are contemplating. The major consideration would be well depth and sensitivity as you may saturate the stars before getting the fine detail.

Mark

Last edited by marki; 28-10-2011 at 11:22 PM.
Reply With Quote
  #12  
Old 29-10-2011, 03:41 AM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
Thanks Mark. I understand the Bayer matrix, and now that you explain it the effect of binning is really very obvious. There is no way around that as far as I can tell.

I am happy to forego FOV for image quality if it comes to that. As I remember the KAI 4020 has greater well depth and is more sensitive than its 8mp counterparts.

Given that 3 - 4 minutes is optimal exposure time with a DSLR CMOS under my skies, this would be within any practical limit of an OSC from what I understand, avoiding blooming. Would this be true with anti blooming switched off.

Having cleared up these grey areas, I think it comes down to a sensible decision balancing needs and camera attributes/limitations. Given that the difference in relative QE between the chips mentioned is only a few % - I'm not sure whether that is significant in practical terms - the remaining considerations seem to be resolution and noise.
Reply With Quote
  #13  
Old 29-10-2011, 12:28 PM
marki's Avatar
marki
Waiting for next electron

marki is offline
 
Join Date: Mar 2008
Posts: 2,427
Another major difference is electronic shutter (KAI 4020) vs Mechanical shutter (KAF 8300). The electronic shutters are easier to use when taking flats etc, the mechanical shutter needed for the full frame chips can leave artifacts if you dont have your timing down pat. The electronic shutter is also better for very short exposures like the moon and planets (I would not even try with the KAF 8300) but full fame sensors are better for AP even with the extra complexities introduced by a mechanical shutter. This makes a good read on the differences.

http://en.wikipedia.org/wiki/Charge-coupled_device


The KAF 8300 does have 1000x blooming protection whilst the KAI 4020 has 300x but you would expect this when well depth is 25000e vs 40000e respectively. I have not used the KAI 4020 but the KAF 8300 is pretty clean noise wise at temps below -20 C. The KAF 8300 is cleaner with 16e Vs 25e total noise for the KAI 4020. In short the noise performance is going to depend on the camera you buy. The deeper the cooling the cleaner the result. Good cooling and darks go along way towards minimising noise.

Mark
Reply With Quote
  #14  
Old 29-10-2011, 02:32 PM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
Thanks Mark. There's lots to learn. And I wouldn't be happy making a purchase without a substantial level of knowledge. I think I'm at the stage where I need to see the results of combinations of gear, similar to my set up.
Reply With Quote
  #15  
Old 29-10-2011, 10:40 PM
Tandum's Avatar
Tandum (Robin)
Registered User

Tandum is offline
 
Join Date: Apr 2008
Location: Wynnum West, Brisbane.
Posts: 4,161
new astronomy ccd press

Rowland, google search for the above, that tool lets you combine ccd spec and telescope spec to get a fov and arc/sec/pix result. I can't tell you you which is best. It's up to you to decide which one you like.

If I run that code here with a 50mm F4 lens and a qhy10 I get 6.28 arc/sec/pix and 271 arc min FOV.
Reply With Quote
  #16  
Old 30-10-2011, 11:06 AM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
Thanks Robin. I'll have a look.
Reply With Quote
  #17  
Old 30-10-2011, 11:20 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Quote:
Originally Posted by RickS View Post
I have heard a couple of people suggest that binning a KAF8300 doesn't appear to give a significant increase in effective well depth. I haven't ever tried binning on my SX H18, but I might have a play with a light box on the weekend (if it's cloudy) and see if I can test this hypothesis.

Cheers,
Rick.
Here's a link that explains it if you can understand it. The writer hasn't bothered to define the unusual words but you get the idea that the way the CCD is read affects the result when binning and it isn't a straight quadrupling by binning 2x2 but there is an increase.

7.4 micron would not be in the middle of 200 and 400mm focal length but much longer.

It comes down to sampling theory.

My understanding of this is the Nyquist theory of sampling which states a good sample requires a minimum of 2X to get a good sample.

In other words you want your pixels to be of a size that will gather enough data to get decent resolution but not too many that simply reduces down too much the sensitivity ie several pixels trying to get the same amount of light - its spread too thin. You don't want it spread too thick (undersampling) and you don't want it spread to thin (oversampling).

One way to work it out on a simple easy to follow rule is 1 arc sec/pixel.

So if your seeing is 3 arc seconds you plug in the focal length, aperture and pixel size of the camera into the Wodaski CCD calculator and you get an arc seconds per pixel number.

I shoot for .66 arc seconds per pixel but what you will find with super short focal lengths is that number will get large even with small pixelled cameras.

Having said that this is just a guide. There are numerous awesome FSQ106 images using 16803 chips with 3.6 arc seconds per pixel.

One way of looking at that is it makes the system less prone to the effects of seeing.

These calculations I think are more important for long focal length scopes where small pixels will wreck an image. Larger pixels in shorter focal lengths will still be good but not the other way round.

Take other factors into consideration as well like QE, well depth, read noise.

But recent Kodak chips largely have similar performance one to the next once you get past the 16803. The really large chips tend to have very poor performance characteristics for astronomy. Like the KAF39000, 40000 (one of the better large chips) and 50100. Really low QE, small wells, noise is fine.

KAI versus KAF. As pointed out there are 2 main advantages of the KAI chips - electronic shutter and no ghost images. KAF tend to have higher QE though. KAI only go up to 28.8mp, KAF go up to 50mp.

The KAI series seem to have more of the new True Sense one shot colour matrix with far better low light performance.

If you were doing one shot colour I would try out the KAI4050 before the tried and true KAF4021 one shot colour. Noone has posted an image with one but on paper the KAI4050 should be better as a one shot colour.

Most True Sense sensors are 5.5 microns which is quite small for Kodak and only the 8300 is smaller with 5.4 microns (insignificant difference in size).

I have seen for myself the difference with 8300 in fast refractors. Definitely the small pixels shine. The 9 micron pixels are a good compromise but the small ones take on an edge in APOs as they tend to be fairly short in focal length (given good seeing that is).

I hope this helps a bit.

Probably more important than the chip is the camera itself. A FLI or Apogee and QSI as well as Starlight Express seem to be the current leaders. SBIG has a good range and lots of accessories. Moravian are yet to be proven and perhaps still working out the little bugs.QHY are popular because of price and seem to be good bang for buck so is Atik. How they compare in other areas I don't know. Processing cleans up a lot of the little differences.

The main points to evaluate cameras are:

1. Price/value. No point in looking at a $6000 camera if your budget is $2500. Its a very competitive market these days and the cheaper cameras seem to make big strides forward.

2. Cooling. There is no substitute for powerful cooling and cooling solves almost all CCD problems and if anything no brand cools hard enough. FLI and Apogee have the most powerful cooling, Apogee takes 30 minutes plus to get there plus the firmware takes over so you can't utilise alll of the cooling power available. It is a flawed cooling system with incorrect engineering assumptions. FLI is best, QSI Series 6 is excellent, Apogee is great with small chips (if you're patient). SBIG ST8300 may have reasonable cooling but possibly a bit weak but good enough for 8300 chip.
Professional CCD cameras cool to -100C which is where some chips need to be to fully handle ghost images (more for back thinned chips).

As chips age they get more hot pixels or develop lines. Strong cooling fades these defects away. I noticed I have a vertical line in my Microline 8300. I don't notice it because its only there at warmer temps. At normal cooling it completely disappears. I run it -35C in summer and -40C in winter.

3. Sealed CCD chamber with inert gas installed. This prevents frosting. SBIG uses a desiccant plug and whilst this works well it is just something you have to service from time to time.

4. With the sealed chamber comes the option of no cover slip on the CCD which gains you an extra 2% QE and also no small halos on bright stars in fast systems from the cover slip (minor issue really but its something).

5. Download speed and instant connection. My main dislike of SBIG cameras (STX may not have this problem) is the download of the driver from the computer when it powers up. It makes it lag, sometimes it fails, if the power or any cable is disturbed it stops, and needs to have CCDsoft shut down, repower the camera and then reboot CCDsoft - a real drag. If you have a reliable SBIG camera and reliable cables and power supply this is probably no big deal. I have not experienced that and I estimate I have had to do this 300 times.

I love the instant connection of the FLI and Apogee cameras. They have a built in memory buffer and the driver is on board. Once turned on you can unplug them from the computer and they are still going once plugged back in. You can't do that with an SBIG camera except perhap STX (I assume). For me that is a deal breaker but for you maybe not.

6. Download speeds. FLI are fastest. I think the new QSI series 6 are now fast as well. Apogee is quite a bit slower but fast compared to other brands. My FLI cameras download 1x1 in 1 second. Great for dusk flats etc.

Also self guiding in SBIG cameras gets stopped during downloads. An STL11 takes 26 seconds to download a 1x1 image. That means your tracking has been building up errors and you need to program in at least a 30 second delay for the guiding to catch up before the next exposure.
Not so with STX. Also not with FLI or Apogee or QSI or others as they use a different guide chips.

7. Accessories. SBIG has the best range. FLI has filter wheels and focuser but no guide cameras nor offaxis guiders. Apogee also. Starlight Express has the next best range. QSI also offers the filter wheel and guider in one package which is a fabulous idea and better than SBIG's self guiding as you can guide normally though narrowband filters.

8. Noise level of electronics. This should really be up higher on the list.
I am not sure of the others but FLI, Apogee and Starlight Express have super low noise electronics and you almost don't need darks.

9. After sales service and resale value. Speaks for itself.

10. Futureproof.
The STL design had a flaw in that they made the opening too small. So advances in CCD size meant they couldn't fit the next generation chips in the body. A big and costly mistake. FLI and Apogee have very large openings and will be able to handle any size chip for some time. FLI can put in any chip even the Kodak 50mp monster with no sweat.
QSI 6 series is limited to the KAF8300 at the largest and are working on another body for the larger chips. All the others seem to be able to accomodate up to 16803 at this time. FLI Microline can handle any sized chip. Apogee Ascent only KAI chips and limited to 29mp, Alta body anything. QSI the 8300. Starlight Express offer up to 16803, QHY and Atik I think both offer up to 16803.


11. Reliability. Obvious.

12. Weight:

Here is a loaded one. FLI Proline is arguably the best but its main problem is it is very heavy. I don't know what an SBIG STX weighs but it may well be the Proline is the heaviest of all these cameras. Apogee Alta is a few lbs lighter. But the good news is FLI Microline is lighter and more compact than a lot and can handle any chip. On a 16803 chip apparently the Microline cools to within 3 or 4C of the Proline. That may mean -30C instead of -35C on a 16803 chip. You would get -25C with a standard Alta Apogee or similar with the DO9 body (but that probably makes the Alta the same weight as the Proline).

Weight is a big issue with some scopes. Many scopes can not handle the heavy weight of a Proline, some can. A lens is not an issue as it will not be supporting the camera.

Phew - I did not mean it to be such a long post but there is a lot to take in on CCD cameras and its such a big investment you want to make sure you get it right.

If I were getting another 8300 camera I'd go QSI Series 6 as I like the idea of built in filter wheel with cheaper smaller filters and a built in offaixs guider. I love my FLI Microline though.

If I were getting another large chipped camera I would go with FLI Microline. Its a bit cheaper but mostly because it is way lighter and being lighter like heavy cooling, saves you a lot of problems.


Greg.
Reply With Quote
  #18  
Old 30-10-2011, 01:08 PM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
Wow! Thanks Greg. You have summed up a good many questions that have been revolving around my head. I keep coming back to FLI Microline, which is slightly more expensive than other makes as I recall. As you say, as an all rounder FLI seems to have good characteristics in several important areas.

I think given my skies (it has been raining for weeks now) and AP opportunities QE is important. I just need to feel satisfied that the pixel size is appropriate.

Thanks again. I'll keep chipping away at this.

EDIT: I see that the KAI 4050 is 5.5um.

EDIT: and available in Apogee - not FLI - which is my next choice behind FLI.

Rowland.

Last edited by rcheshire; 30-10-2011 at 01:22 PM.
Reply With Quote
  #19  
Old 30-10-2011, 01:12 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by gregbradley View Post
Here's a link that explains it if you can understand it. The writer hasn't bothered to define the unusual words but you get the idea that the way the CCD is read affects the result when binning and it isn't a straight quadrupling by binning 2x2 but there is an increase.
Would love to see that link about binning the KAF8300, Greg. Did you forget to include it?

Thanks,
Rick.
Reply With Quote
  #20  
Old 30-10-2011, 02:18 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by rcheshire View Post
Relative QE 4050 BGR ~ 31, 42, 43 - 4022 BGR - 45, 42, 35. In real terms is this significant?
Looking at the QE for Blue, the 4050 is (approximately) only two thirds as sensitive as the 4022 so, for example, you'd need a 9 minute exposure with the 4050 to capture the same amount of data as a 6 minute exposure with the 4022. That's the worst case. Green sensitivity is the same. Red is lower, but not by as much as Blue.

Only you can decide how significant this is relative to all the other pros and cons.

Cheers,
Rick.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 05:58 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement