small format CCD camera recommendations please

[fwm891]

Hi, I’m thinking about getting a second CCD camera. I have an Atik 383L+ mono which I use with Baader filters on my AA CF RC8 and AA 115 APO refractor with 1x and 0.79x reducer.

These combo’s produce a reasonable range of image scales for the more common objects but I would like to image smaller galaxy’s, planetary nebs etc and find that if I use the 383 then I end up cropping the majority of the image areas. I know I can set a crop frame with the 383 but would prefer not to (just me!).

So looking for a smaller chip size, mono or OSC being considered, but probably a mono in preference.

Any thoughts / recommendations gratefully received.

[Earl]

A 314L+ will serve you well

http://www.12dstring.me.uk/fovcalc.php

check your rc on m106 i think it will be what you want.

[perfrej]

314L+ has smaller chip but bigger pixels than the 8300 so resolution will be less... Not a good choice. If you want to improve resolution you need to get smaller pixels, regardless of the chip size. I'd say something Sony. Your 383 has 5.4um pixels, the 314 has 6.45 and the Sonys run from 4.5 down to about 3.5. Sony. Just make sure you don't lose to much real estate...

/per

[Chris]

Maybe an Atik 420L then? 

I've not ran it through 12Dstring with your setup, but its a reasonably small chipped camera with fine pixels.

[swag72]

How about the new QSI6120? That has small pixels as has the QSI690.

[ollypenrice]

Per's analysis is on the money. For galaxies at modest focal length you need resolution, so small pixels. If you don't need real estate then you have the luxury of  Sony available to you. Grab it with both hands. Depending on the seeing you are likely to encounter I'd go for the Atik 460 or 490. I like Atik bang for buck and you wouldn't confuse yourself with new software. (This might not be an issue for you but I'm very easily confused!!)

Ollyu

[Steve 1962]

I can't recommend the Atik 460 ex mono too highly - it's a stunningly low noise camera with 4.54um pixels. If I wanted smaller pixels - I wouldn't hesitate in going for the 490.

[fwm891]

Many many thanks everyone.

Not looked at QSI Sara so will have a look at their site, already being an Atik user I'm looking at those too so thanks Olly (with a 'u'), Starfox and Steve for leaning in that direction.

[ollypenrice]

Many many thanks everyone.

Not looked at QSI Sara so will have a look at their site, already being an Atik user I'm looking at those too so thanks Olly (with a 'u'), Starfox and Steve for leaning in that direction.

Hehheh, normally I'm uless!

Olly

[Earl]

I have always been under the impression small pixels are for short F's and large pixels are for longer F's

Am I missing something here?

For myself im matching the following 6120 due in 5 weeks

FSQ106 with the QSI 6120 for all filters at F3.6

ZS71 with a Trius SX-694 for NB at F4.7

8" RC with a QSI 683 for all at F8 (might use a .7 reducer undecided) might even Bin it.

9.25 with SBIG 8XME for all at F10 or F6.3 This camera is the most maybe part of my project as I might just keep this for Luna / plantary

[NickK]

it's down to the airy disc and defraction limit. Normally the disk centre is kept so that it fits into the size of a pixel. This means that things appear pin point sharp (i.e. as sharp as the camera can perform). Next up is the additional atmosphere 'churn' that will show up with smaller pixels.

It's still possible to reduce the size of the sensor pixels. There will be a point where the pixel size is less than the resolution of the scope- at which the image will soften slowly. Additionally small pixels means a smaller well capacity.

The 383L/ 8300 is 5.45um and with a TV 2x gives me 0.88 arcsec/pixel compared to my 4" defraction limit of 1.10 (IIRC). This is not a problem, however it does soften the image slightly that can be undone with software. I've actually used a 5x and 2x stacked to give me f/64 6700mm which is well below the defraction limit and I still managed to get a solar image out of it!

I agree - Sony all the way.. and if I had the funds I'd be very tempted with a sub-4 um camera.. I like my 383, not because of the camera real-estate but the size of pixels..

[fwm891]

Thanks again folks. I run my 383 at 1x1 binning with the 1624mm fl RC, I should run at 2x2 but I find the results much better unbinned. What's annoying with Atik at the mo is they're ads saying what a good QE the cameras have but they won't give any numbers, unlike QSI and some other manufacturers - come on Atik,  I know it varies with wavelength but how about a QE / wavelength chart...

[opticalpath]

....... What's annoying with Atik at the mo is they're ads saying what a good QE the cameras have but they won't give any numbers, unlike QSI and some other manufacturers - come on Atik,  I know it varies with wavelength but how about a QE / wavelength chart...

Francis, you can probably find what you need direct from the sony datasheets for the sensors, or indirectly from other camera makes using the same sensor.  The ICX694 used in the Atik 460EX, for example (from QSI web site):

http://qsimaging.com/images/photos/ICX694-QE-380.jpg

Adrian

[NickK]

QE is really dependant on the sensor itself.. that's Kodak and Sony. Normally the cover glass is the only other intervening feature on QE and the sensor manufacturer normally gives a recommendation with QE on the response curves.

Then there's the handling ability of the scope over that wavelength range. Many scopes can cope with visible range but IR/UV block is required because the scopes don't focus that properly resulting in bloat/blurring. However that's another story

If you're doing proper colour calibration, other than measuring the known responses yourself, using the QE curve with the filter transmission curves is about the best way of doing it..

[carastro]

Clocking up all this info for future reference.  

I really like the Atik314L but looks like there are even better cameras out there, so when it's time for an upgrade.........

Carole

[Earl]

Im still under the impression that longer Focal length imaging benefits from big pixels, does anyone have a link for any information on this?

[Steve 1962]

Here you go Earl - reading this, I think you're right.

http://www.astro-imaging.com/Tutorial/MatchingCCD.html

Steve

[NickK]

Longer focal length normally means a small photon count from a smaller sky area. Aperture is the only way to increase, the perceived other way is to reduce resolution by increasing the pixel size. Capturing at infinity means you're only going add more photons by increasing the aperture or by reducing focal length that accuracy works to lower the resolution, thus 'binning' the photons.

If you look at the spot testing diagrams for RC scopes etc with 24" you'll note that the outer edge spot size is quite large.. often 7-12um.

In idealistically there is no such thing as the perfect pixel size - idealistically with infinite resolution you can cope with infinite resolution given perfect optical capture. Reality though is that the resolution from the scope and atmosphere limits the minimum pixel size.

[ollypenrice]

Im still under the impression that longer Focal length imaging benefits from big pixels, does anyone have a link for any information on this?

This is true. I've tried a wide range of scales between 0.6 arcsecs per pixel and 3.5 (or even more with camera lenses.) What you can really resolve depends on the seeing and the guiding. (Round stars don't prove that you have reached the real resolution of your system.) You can always bin when your pixels are too small - though not all cameras bin well. I've come across two SX cameras which wouldn't bin satisfactorily. In the real world there are lots of other variables. Software sharpening can be very powerful but only if you have sufficient data. If you don't the image just breaks down into noise. So, what it doesn't say on the tin is that there is a strong link between final resolution in the image and the depth of the data. (Note that all these observations are assuming DS imaging. High res in fast frame cameras is another game and not one I play very often!)

Olly

[NickK]

Yup, it's possible to drizzle larger pixels and recover more resolution. At the same time it's possible to drizzle small pixels and recover more resolution however gut feeling is the additional error of spot size needs attention (although the averaging helps reduce it I'd suspect).

[fwm891]

Thanks everyone - great info. No decision yet...

[Gina]

I'm a big supporter of the Atik 460EX mono CCD camera.  Very sensitive, very low noise, bins well   In fact I'm so impressed that I've bought three of them!   For a triple imaging rig to catch the few photons that manage to get through our UK night skies

[Horwig]

is the qsi 6120 available yet, does anybody have any hands on experience with it?

[Zakalwe]

What about the Atik 428Ex? 4.54 micron pixels. Same Sony technology as the 460, just with a smaller chip.

[MiguelMJr]

When picking a sensor it is generally recommended that the image scale be between 1 and 2 arcsecs/px. This has nothing to do with the resolution that your scope is able to produce, it has to do with seeing.

The formula to determine image scale is ... Resolution=206*(P/F) where P=pixel size in um and F=focal length in mm. For example....

A KAF8300 chip has a pixel size of 5.4um

My telescope has a focal length of 420mm

5.4/420=0.1285...

0.1285*206=2.64

2.64 is the number of arcsec that cover one pixel. The KAF8300 chip has pixels that produce a larger image scale than the recommended 1-2 arcsecs if used with my scope. I need smaller pixels or a longer scope.

If you know your focal length and you want to know which pixel size falls within the 1-2arcsec/px then use this formula...

(F*arcsec/px)/206=image scale. Example using the same numbers as above......

420*1=420

420/206=2.03...

2.03 is the smallest recommended pixel size for my scope... now...

420*2=840

840/206=4.07...

4.07 is the largest recommended pixel size for my scope (you can also take the lowest recommended pixel size (2.03 in this case) and just multiply by 2 to get the largest recommended pixel size, it's easier.

So for my scope I should have a pixel size between 2.03um and 4.06um.

PS: if using a focal reducer or extender you need to use the "adjusted" focal length in these equations.

Hope this helps

Miguel