Atik 460 or Atik 4000?

[Martin-Devon]

I currently have an Atik 314L+, which is excellent, but I want to get wider field captures now. I have a 4" (102mm) refractor, and will probably stick with this size as my main imaging scope as I upgrade. I don't want to go up to the Atik 11000 CCD since that is huge, costs a fortune, and also requires 2" filters.

I've homed in on 2 possible CCDs that may fit my needs - the Atik 4000, or the new Atik 460 - I'd very much appreciate views & feedback on which of these may be best or whether there are alternatives as well that I should consider. I've passed on the Atik 383, it gets somewhat mixed feedback, can be noisy, and has been around a while as well. Just to add that I have an Atik EFW2 as well, and the Atik CCDs are totally compatible with this, which helps.

In terms of specs, the Atik 4000 uses the Kodak KA1 04022 chip, 2048 pixels x 2048 pixels, and pixel size is 7.4uM x 7.4uM whereas the Atik 460 uses the Sony ICX694 chip, 2750 x 2200 pixels, and pixel size is 4.54uM x 4.54uM. From the few reports so far on the Atik 460 it seems to have low noise, probably lower than with the Atik 4000. The 460 also has a larger chip size both in length & breadth, although the actual pixel sizes are smaller than with the 4000, I don't know what difference this makes though? The 460 is currently about £600-700 cheaper also than a 4000.

So my question - is the Atik 4000 still the superior camera (as you would expect with it's higher price) - it certainly has a good and proven track record. Or is the new Atik 460 better? Feedback appreciated....

Thanks,

Martin

[NickK]

I know which I'd go for - the 460.. but that's because I like small pixels and sub-dawes limit imaging with 0.88 arcsec/px resolution with my 383L+'s 5.4um pixels

[robbieince]

i don't think the 460ex is actually bigger than the 4000, its reported to be about 12.5x10mm whiich makes sense as it has 2750 x 0.0045mm as its long axis. The 4000 is2000,0.0074 = 14.8mm

The 460 has a higher QE and lower noise though which will offset the smaller pixels. That makes therm close.

I love my 4000, I personally think smaller pixels would be better for shorter focal length telescopes, but as i image at 1.96 arc seconds per pixel, I am around the seeing effects anyway.so wouldnt want the 4.54 micron pixels in my setup.

[Martin-Devon]

Would somebody please explain to me in simple terms the arc second/pixel referrals above - how do you calculate it, why is it important and what is the optimal value (my refractor has focal length of 715mm) - thanks

[NickK]

arc seconds are a measurement of angular resolution. If you stick your arms out up to the sky - it's the angle between them. In short the bigger the angle the bigger the area of the sky you see. Now imagine the same but for each pixel on your camera's sensor - then your arms are your scope (the focal length). The smaller the pixels, the smaller the angle with the same focal length.

Seeing is usually 2" or two arc seconds.. so going below means that the image will wobble.. then the next limit is the aperature of the scope which is something called dawes limit. that governs the smallest arc second resolution you can achieve with a scope.

A lot of DSOs are measured in degrees or arc minutes rather than arc seconds.. but having a pixel resolution allows you to see more detail and have the DSO appear larger on the image. Limits applying.

[NickK]

Would somebody please explain to me in simple terms the arc second/pixel referrals above - how do you calculate it, why is it important and what is the optimal value (my refractor has focal length of 715mm) - thanks

there's a great application called CCDCalc that allows you to input the details and it will scale some common DSO objects.

http://www.newastro.com/book_new/camera_app.php

[pete_l]

So far as arc-pixels per second goes, take the CCD pixel size in microns (i.e. 4.54) and multiply it by 200. That gives you an idea of the "optimal" focal length that the CCD will work with. In fact the accurate number is 206, but it doesn't really matter and the calculation tells you the FL that will give an image scale of 1 arc-second per pixel. There's no real hard and fast rule that makes this "best", though.

The theory goes that most people have an amount of atmospheric turbulence that means yer average star will subtend about 2 arc-seconds. What happens next is that gets a bit mixed up with Shannon's Theorem from electronics signal processing theory, and out pops a diktat that your telescope resolution should be half of the 2 arc-second number. Basically, it's as good a guideline as any other but it's not magic.

What you want to avoid is having such large pixels and so short a focal length that your stars look square - i.e. fully fit on a single pixel, instead of having some light "leak" across many pixels.

[JacobvonChorus]

I think the 4000 is the better option because the pixels match you telescope a bit better.

Jacob

[swag72]

Martin, while I can offer no advice or comments, this is something I am thinking about in my more lucid moments as opposed to thinking the impossible with my gear. Thank you for asking the question and I look forward to some more ideas.

[ollypenrice]

I'm afraid I pay far less attention than most to my arcseconds per pixel for two reasons. 1) I have a wide range of FLs available and connot afford to buy a separate camera for each one.

2) I don't find it matters! I post plenty of images from the Atik 4000 at 328mm focal length. I can't see much to moan about but the sampling rate is, in theory, way out. What would be better in this image if I'd used smaller pixels?

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/22435624_WLMPTM#!i=1793668870&k=R4v6bFN&lb=1&s=O

Tom's recent APOD of the Rho Ophiuchus region was shot on 9 micron pixels at 530mm FL. It looks OK to me...

http://apod.nasa.gov/apod/ap120828.html

The 460 has a much nicer chip and the 4000 has a much bigger chip. What else can one say? Why won't SOny make big ones???

Olly

[swag72]

Will Sony ever make a bigger chip? Olly, I presume that you have asked these sorts of questions before to people in Atik, what is the type of response? Maybe there's one around the corner?!!!

[ollypenrice]

I have never asked anyone personally but on the forums I've read that Sony say they are 'not in that part of the market.' Remember that astro cameras make up a trivial proportion of thier sales.

Olly

[robbieince]

Agreed Olly, its best not to get hung up on the theoretical numbers - they are only ever a guide and of course planetary imagers routinely expose at fractions of an arc second per pixel, but then they are only using the moments of best, static seeing and very short exposures.

I think that being undersampled such as at 3 or more arc seconds per pixel is really no big deal, it maximises sensitivity for distributed sources and can in fact makes pictures look sharper as stars fall onto a very few pixels. Its very easy to soften an image in processing if it looks too pixelated.

My point is that being oversampled such as 1 or less arc seconds per pixel whilst making nice smooth pictures with resolution as high as seeing allows puts a really high strain on mounts and guiding systems. It also means that the camera electronics need to have high gain and low noise to offset the smeared signal intensity. If you have kit that can satisfy this, ie very accurate tracking and ability to do long exposures with little or no tracking errors, then small pixels / small arc seconds per pixel are the way to go.

Sadly most of us are limited by seeing and our mounts tracking errors, so larger numbers of arc seconds per pixel will give us the best return in real picture quality.

[ollypenrice]

Ah, it's good to find someone who agrees since I don't often voice my opinion on this.

Olly

[swag72]

I value your opinions greatly Olly - I am glad that you voice them and are so willing to help others. I for one have always found your real life advice and examples invaluable. Things are not always as they seem on paper and in theoretical models and we are sometimes in danger of sticking to these instead of experimenting and trying things out that appear on paper to be 'wrong'.

[Martin-Devon]

Thanks to everybody for your feedback. As emphasized both CCDs are very nice and offer slightly different plus & minus points, there is not a clear-cut winner, I would be happy with either one as a step-up from my Atik 314L+

I had a long discussion also with Ian King yesterday and although the 460 has a slightly smaller chip size, it's lower background & higher sensitivity swung it for me, especially since I do virtually all narrowband work. It's also nearly £700 cheaper.

[ollypenrice]

I think that for NB the better Sony chip does make sense though I'd call the 460 chip significantly smaller, perhaps because I'm such a widefield junkie! You'll find that it's a lovely looking camera when you get your hands on it. Very convincing.

Good luck with first light.

Olly

[swag72]

Glad that you made up your mind Martin.

Now Olly, you have given me a quandry as I was all set on the 4000, but I will probably doing more NB than RGB - So does that muddy the waters and change my plan?

[ollypenrice]

Glad that you made up your mind Martin.

Now Olly, you have given me a quandry as I was all set on the 4000, but I will probably doing more NB than RGB - So does that muddy the waters and change my plan?

Well, you put wide field down as a priority and there is quite a difference in chip size. I'd model them both on the sky charts or CCD Calc or whatever and see whether your likely targets are covered by the smaller chip. If they are, then there is a lot going for the 460. I should be trying one in the next ten days. It is here and itching to get started. I just have to fit it in with whoever is doing what in which scope!

Olly

[swag72]

I want both Olly, as wide a field as possible and for narrowband predominantly. Faced with equal importance .......... now what? Have you known people to do NB with the 4000?

[Spikey]

I've been looking at something to replace my DSLR but the price puts me off. The 460 looks like an excellent compromise on chip size though. I'll probably be keeping my 16HR mono as it's so good. A big plus (for me anyway) is the fact that you can still use 1.25" filters with the 460 if you already have them

[swag72]

Ah yes, but I can still use 1.25" filters with the 4000 and that field of view and square chip is lovely!!

[martin_h]

Ah yes, but I can still use 1.25" filters with the 4000 and that field of view and square chip is lovely!!

Taken from Atik -

"The Atik 4000 features a 4 million pixel Kodak KAI 04022 sensor with a 15mmx15mm array and is available in both mono and one shot colour versions. Normal 1.25” filters are used to their fullest with this camera and offer a considerable saving compared to cameras requiring 2” filters."

[swag72]

Sensitivity (for NB)...460 ... over size ...4000... - I think that's what it boils down to ............. And I can't decide!

[robbieince]

My 2p - got for the 460 if narrowband is important as the extra sensitivity and Lower noise will make your exposures have a higher signal to noise ratio.

Regards

Rob