The ultimate which ATIK ccd thread.

[ncjunk]

Hi,

I´ve seen a few threads of "which ccd ATIK XXX or ATIK YYY" but I wanted to start a thread going through each of the CCDs and seeing if we could actually determine some Pros and Cons of one vs the other. I have included the 420, 314E, 314L,428, 450, 460 but thinking about it perhaps I should add the 383 at some point.

The field of views and pixel sizes are calculated based on an ed80 (600mm fl), ed80 with reducer (400mm) and a C11 with reducer (1867mm) because they are the scopes I have (I´m looking at buying one of these ccds hence the thread)

As you can see, hopefully, from the attached picture the 314L, 428 and 450 have pretty much the same field of view so you are looking at QE and pixel size as the determining factor. The QE is significantly better (judging from the little imformation available on the web) on the 428 and 450 cameras than the 314. Also the pixel size of the 428 seems to fit better with both scopes than the other two cameras, fewer arc seconds per pixel on the ED80 and even on the C11 where it is 0.5 you could try 0.5 or bin to take you up to 1.

The 420 seems to have a good pixel size, similar to the 428, but you are losing field of view and possibly slightly lower QE (not sure if that QE is correct though)

The other thing that crops up a lot is people extolling the vertues of the 460, which is a fine camera, but is it worth the extra money? It has exactly the same spec as the 428 but a larger ccd size. So, in my case, is it worth the extra 748€? I´m not conviced it is and I think if money is tight, relatively speaking with these prices, then perhaps the 428 would be the better choice.

Then we have the 450L, why don´t more people use this camera? Is it because of the small pixel size? does size really matter? if I have a lower well depth per pixel but more pixels per mm don´t the two cancel out to some extent and if that is the case are we guilty of interpreting the figures incorrectly?

I think that the general interpretation by people of some of these figures may not be correct and that we are unfairly discriminateing against one camera or prefering another due to urban myths or ignorance (that is too strong a statement but I can´t think of a less controversial way of putting it at the moment)

So, I am thinking of getting the 428 but would love to throw out all these ideas to see what I get hit with by people. It´s a shame I haven´t won the lottery or I would buy all of these cameras disappear to chile and do a thorough comparison.

Appologies for all the spelling mistakes!

NC

[Gina]

That looks very useful I've worked out fields of view for my scopes and lenses too and I'll post that shortly but a couple of the figures look wrong so I need to recheck first. I have the Evostar ED80 but with the 0.85x standard SkyWatcher reducer/flattener giving a focal length of 510mm. Which FR are you using - a shorter FL and corresponding lower focal ratio would be good (as long as the quality is still good). I have worked out the figures for the two cameras I have viz. Atik 460EX and 314L+.

[AlistairHowie]

There's no right answer to this, it all depends on what value you place on FoV versus resolution versus QE versus etc etc.

I would suggest that your first port of call is surely to decide what sort of targets you want to image, and what sort of FoV you need to image "most" of those without have to do multiple mosaics (unless you're really into that). New Astronomy CCDCal is very useful for that.

That will narrow you down the most, then you can look at arcsec/pix to ensure you're in the general oft quoted region of 1-3. If you want most flexibility, consider putting the 9 mega pixel 490EX into the mix since you can bin that to your heart's content and still keep the pixel number relatively high.

Then all other things being equal, choose the CCD with highest QE and which fits your budget.

Now .... I suspect you know all of that anyway, so this post may not be that helpful to you. At the end of the day, only you can decide whether you're a widefield junkie like Olly ... in which case you camera choice probably just comes down to how much real estate you can afford. The arcsec/pixel number is less important for widefield, some great pics have been taken way above 3. Sure the stars are blocky if you zoom in far enough, but that's not the point of a widefield image, is it? If you're just going to zoom and crop, then go for a smaller chip in the first place.

If you're not THAT much into widefield, then you've got a wider selection of cameras to choose from, in which case your main criteria will probably be arcsec/pix then QE and/or price.

Hope this helps.

[ncjunk]

Great replies thanks!

Yes you're right in all you say but to me there's not a lot beteen these cameras. The only one that would make a difference for widefield is the 460, which is probably why a lot have gone for this. I'm not a widefield person which means looking at these figures the 314l is a bit big in the pixel size and the 428/460 is actually much closer to a perfect all rounder.

Hence the other question i posed, related to you mentioning the 490 which has little baby pixels!, which is do small pixels really matter? What if the 4 pixels that are taking up the same space as one 314l pixel have a greater combined well depth and qe response? Or do you lose a little bit of resolution? So that it's not just as simple as well depth multiplied by 4....it could be 80% of well depth multiplied by4 for instance.

Apart from that the new generation of cameras have really changed my mind about the 314l, its no longer the top dog and best allrouder and i'd even argue that it would be better paying a little bit more for the 428ex.

Keep the opinions coming! Especially about pixel size! If i bin a small pixel camera to the same equivelent larger pixel size is it win win or are there loses?

Sent from my GT-P5110 using Tapatalk 2

[RAC]

What about the 383L. It has a much larger sensor than all the above cameras and is a bargin in my view. I agree with others about the colour verson not being the one to go for but there is nothing wrong with the mono version. It has taken many stunning images.

I love mine and don't for one second wish i spent another $1000 on the 460.

[ncjunk]

Yes! The atik 383l it has pixel size of 5, i think, but the qe is lower peaking at 56.

An interesting line of thought, is the increased sensor size at the expense of qe worth it or is a smaller sensor with higher qe more important?

Sensor size or qe?

Has anyone done any tests to show the differences between two cameras with differing qe values?

Sent from my GT-P5110 using Tapatalk 2

[ollypenrice]

I like Alistair's post.

I also get less excited about pixel resolution than some people but that may, indeed, be a result of my passion for widefield. It is also connected with the fact that if you get more signal you can do more software sharpening.

If I were going to buy an 8300 chip I might go for more expensive electronics and cooling or choose a different chip.

If I weren't fussed about chip size (which I am, big time!!) I'd just go for a 460. I've used two and they are as crisp as you like. Remember, if you have a bigger chip you can use a bigger scope. My old Atik4000/TEC140 gave a nice result but it could not compete with the same(ish) FOV captured in Yves' ODK14/Full frame. How could it?

No mention of the 4000?

In the end, how much does it matter? In my view, less than you might think... I think!!

Olly

[RAC]

In the end, how much does it matter? In my view, less than you might think... I think!!

Olly

Yes i agree. Integration time and good guiding/tracking is right up there on my list of important things.

[Gina]

I like the square format of the 4000 but the S/N of Kodak chips is poorer than Sony ones.

[RAC]

But if there is a few hours at least spent on the image could you really tell the difference between sony and Kodak chips? If you were doing single sub images as a final picture then yes maybe it would make a difference.

[RAC]

Here's one photo done with one of these poor old noisy Kodak chips in a cheap 383L. No darks, biases or flats and only 8 subs. http://www.flickr.com/photos/46302893@N02/8485079462/sizes/o/in/photostream/

[perfrej]

And here's another with a noisy Kodak KAF-8300, no darks, no BIAS, just flats. They suck, don't they

On a serious note, I do believe that the differenses of s/n ratio between the chips sort of disappear at a certain number of subs. Let's just assume that in order to obtain the same result with an 8300-based camera you hace to take 20 subs instead of 17. Is it then something to worry about? Hot pixels disappear in the stacking anyway so they can be left out of the discussion.

A thing to do leave in the discussion is something that Olly pointed out: read electronics quality. One of my 8300-based cameras has almost twice the noise of the other, and it ain't the chip!

That said, I would say that the parameters for camera choice comes down to field of view, "/px and perhaps anit-blooming properties.

/per

[ncjunk]

Those 8300 chips are awful, in both pictures you've got really large grey smudges blocking the stars!

Sent from my GT-P5110 using Tapatalk 2

[ncjunk]

And here's another with a noisy Kodak KAF-8300, no darks, no BIAS, just flats. They suck, don't they

That said, I would say that the parameters for camera choice comes down to field of view, "/px and perhaps anit-blooming properties.

I'm begining to think you're right. Although i would argue that field of view may not be that importand if you're interested in smaller galaxies. All these cameras have a large enough fov for any galaxies you might want to image....apart from m31 which is too fat and needs to go on a diet.

Also, how would it be operating a 8300 based on a c11 at 1867 mm? I assume a large part of the exterior would suffer from elongated stars? But then on my ed80 i could image an olympic size swimming pool!

So why would you chose a 460 if a 383 is larger and cheaper, is there anything that a 460 could do that a 383 couldn't?

Sent from my GT-P5110 using Tapatalk 2

[NickK]

It really does come down to what you want to image, the scope you're using (dawes limit, colour correction etc), the seeing and finally the camera implementation.

Olly is a self confessed widefield junkie, I'm the opposite as I like the camera to see into detail I can't. Now before people mention the seeing.. it's possible to get more out of the seeing than 2 arcseconds that people seem to bang on about which is why I use a camera with 5.4um that gives me 1.67 arcsec/pixel down to 0.88 arc sec/pixel. Yes I process them heavily to recover detail too. However to me seeing more than I can with the naked eye really is the goal.

Don't get me wrong... I'd love todo some more beauty shots over multiple nights too (I may have that facility in future *horray*) but for now.. it's all about the detail and usually in mono with 2 hours

(the end galaxy is something like 22 arc seconds across).

Now my experience of "beauty" shots is rather limited.. as you can see my first and only multi-night image:

Loads of noise because I'm so used to ignoring it seriously probably because I should really get around to processing it properly..

[ollypenrice]

I'm sure the 2 arcsecs business is dead in the water. Lots of people have demonstrated that it isn't true and Nick is right, above.

Of course, I'd like a huge chip and small pixels but that isn't an option for me. I'm very 'Atik' so I wish they did a camera with the SXVH36 chip and 7 micron pixels, but they don't.

I might just go and buy an 8300, though, to capture finer detail for overlaying onto widefields. I can't see it being as fast as the 11000, which guzzles enough light for two to three hours to look good. I can't run to a QSI but I suspect that this is the 8300 of choice. I strongly suspect that this chip likes a top camera. The other option would be to buy the exhorbitant flattener for the TEC and use that with the 11 meg.

Neil, good question about noise and whether or not it matters. I've gone for noisy Kodaks in my last three cameras and can't say the noise bugs me, though calibration routines do vary in efficiency and I can't say I'm convinced that one routine fits all. The 11000 thrives on a BPM/bias and really isn't that noisy anyway. Mind you, at 3.5 arcsecs per pixel it's like working in Bin3! I dare say beinners, intimidated by the perceived complexity of bias and darks, are very much drawn to the Sony chips. It is nice not to have noise but does it matter much? Perhaps not.

Bad news about Andromeda. It's getting fatter!

My latest widefield fix will be coming up soon. The Veil Nebula is bigger than I thought!

Olly

[ncjunk]

Nick - are all those pics with the 383? I think i prefer the same kind of targets as you. I want to do a few lesser known smaller ones.

Olly - you're starting to get into my head....fov...fov...look at the fov on that! Cor.

We live in a consumer world where marketing dictates bigger numbers = better but we all enjoy a hobby where we know in reality bigger isn't always better and theres always a trade off (sometimes its cost). But there's a nagging doubt in my head that we are all getting too preoccupied with the numbers and camera specs. I'm getting less and less inclined to say x camera is the one to get for beginners as theres a y and z thats equally capable. Also a lot of people are recommending the 460 to people when perhaps a cheaper, larger, slightly less sensitive 383 camera could suit them better, well its cheaper 400 pounds ish in the pocket and a little extra processing...does that really make the 460 better?

Sent from my GT-P5110 using Tapatalk 2

[Gina]

With the weather we get here I was looking for a camera with the best ability to grab photons while the going is good so noise and imaging time IS important. OTOH I don't want the pixels too big as I want the detail for galaxies and smaller nebulae without going to a longer FL than my MN190. For me the Atik 460EX would seem to fit the bill better than any other camera I know of. Now if Sony did their ExView HAD II sensors in larger sizes then they would scoop the pool... Anyway, so far I'm delighted with my 460 - it beats the 314L+ and that's excellent!

[ncjunk]

With the weather we get here I was looking for a camera with the best ability to grab photons while the going is good so noise and imaging time IS important. OTOH I don't want the pixels too big as I want the detail for galaxies and smaller nebulae without going to a longer FL than my MN190. For me the Atik 460EX would seem to fit the bill better than any other camera I know of. Now if Sony did their ExView HAD II sensors in larger sizes then they would scoop the pool... Anyway, so far I'm delighted with my 460 - it beats the 314L+ and that's excellent!

How much imaging time would be gained in reality? Again does it justify the costs? Is the gain really THAT significant between a 460 and a 383? If i do 4 hours ha on each is there a big difference in resolution/quality? Again, i suspect we may be over stating the benifits and in really need a ferrari over a ford? If i only get there slightly quicker (consider traffic as light polution!)

I think back to that great mit physics lecturer who has videos on youtube. "A measurement is useless unless you have quantified the possible error inherent in the test"

What happens on a lot of discussions is we state ccd x is better than ccd y without any measurement and proof. Without knowing how much better x is than y it becomes difficult to determine whether one is more suited to your pocket than another. If the 460, for instance, gathers 10% more light resulting in 3.5 hours imaging time for the same level of detail as a 4 hour 383 image then it may be worth it. It it was only 10 minute the it might not. If light polution is such in your area that smaller signals are swamped then the extra sensitivity may not be the advantage it is if you live on a chilean mountain top.

I demand a team of scientists to look into my mad rantings NOW! I want answers damit! And a tablet that doesn't keep hitting the wrong letters.

I fully understand your reasons gina and if i had the money i'd do the same, i'm considering the 428 as its the same spec as the 460 and i don't really need the extra fov. (Although i have a little Olly now living in my head with pompoms and a cheer leaders outfit going FOV! FOV!)

Sent from my GT-P5110 using Tapatalk 2

[ollypenrice]

How much imaging time would be gained in reality? Again does it justify the costs? Is the gain really THAT significant between a 460 and a 383? If i do 4 hours ha on each is there a big difference in resolution/quality? Again, i suspect we may be over stating the benifits and in really need a ferrari over a ford? If i only get there slightly quicker (consider traffic as light polution!)

I think back to that great mit physics lecturer who has videos on youtube. "A measurement is useless unless you have quantified the possible error inherent in the test"

What happens on a lot of discussions is we state ccd x is better than ccd y without any measurement and proof. Without knowing how much better x is than y it becomes difficult to determine whether one is more suited to your pocket than another. If the 460, for instance, gathers 10% more light resulting in 3.5 hours imaging time for the same level of detail as a 4 hour 383 image then it may be worth it. It it was only 10 minute the it might not. If light polution is such in your area that smaller signals are swamped then the extra sensitivity may not be the advantage it is if you live on a chilean mountain top.

I demand a team of scientists to look into my mad rantings NOW! I want answers damit! And a tablet that doesn't keep hitting the wrong letters.

I fully understand your reasons gina and if i had the money i'd do the same, i'm considering the 428 as its the same spec as the 460 and i don't really need the extra fov. (Although i have a little Olly now living in my head with pompoms and a cheer leaders outfit going FOV! FOV!)

Sent from my GT-P5110 using Tapatalk 2

Neil, you might be seriously disappointed if you saw me in a cheer leader uniform... In fact you might feel as you would after swilling a bottle of Fundador to slake your thirst on a hot afternoon andthen wolfing a bag of churros.

I don't like big chips because they're bigger, I like them because I like the sky and more of it is better! I also feel less motivated by the quest for resolution because none of us can compete with the professionals there, but in widefield you can make new discoveries, at least new to yourself. For me the connections between objects, the large scale structures, Barnard's etc, are thrilling. When I first used my Atik 4000 I felt like say, 'Aaaahhh, I can breathe!'

I do miss adding high res from the TEC, though, to key bits. I suppose I could put my Parcel Farce refund towards it. Yes, why not? You have your revenge!

Olly http://www.google.fr/imgres?imgurl=http://www.eastportlandblog.com/wp-content/uploads/2013/02/male-cheerleaders.jpg&imgrefurl=http://www.eastportlandblog.com/2013/02/21/the-feminist-sports-fan-on-cheerleaders-by-cory-davis/&h=514&w=754&sz=99&tbnid=F8BUWrgrZCGekM:&tbnh=93&tbnw=137&zoom=1&usg=__cJckqV__vJFHCA9QCknhPlMA35M=&docid=5y5c2bpFLoouOM&sa=X&ei=8vfXUc3dHunG7AbNmIG4Bw&ved=0CDsQ9QEwAg&dur=127

[swag72]

Neil I am probably about as un-techie as you can get. I like a pretty picture and don't care how I get there - Even to the point of cloning!!! So take my thoughts with a pinch of salt compared to the others on here that have a good understanding of the more technical side of things. Despite that, I thought I'd post even if I look like a luddite!!

At the moment you are looking at the 428 and I can only compare it to a 460 as that's the camera I have used. Perhaps a 8300 chip comparison will give you a totally different conclusion - but here's how I understand it on a pure pretty picture and practicability level.

Both the 428 and the 460 have the same pixel size and the 460 a larger chip at 2750x2200 pixels. So if you take a picture with the 460 and crop it down to an image size of 1932x1452 (which is the same as the 428) then you have exactly the same image, with the same detail, just with less space around it.

My question then is what do you think youe will gain with a 428 chip? If you buy a 460 with this example only, then you get the advantage of a wider available field as well as being able to crop to exactly the same size as the 428 without any disadvantage at all. Isn;t that the best of both worlds in this instance. You can not go wider on a 428, but you can crop in from a 460 image with no discernable difference.

That's my understanding, explained in very simple un technical terms. If I've got it wrong I shall look a total nugget, but will of course stand corrected.

Now comparing a 428 and 8300 chip in the same way may give you a different advantage - But I wouldn't even know where to start with it!

[ncjunk]

You could have put a warning on that photo i've got orange juice everywhere now.

Sent from my GT-I9003 using Tapatalk 2

[ncjunk]

Sara,

I gain 750 euros!

The two cameras, 428 and 460, have effectively the same chip its just ones bigger.

So if they are both technically the same and the galaxies fit on both i could equally ask you why would anyone need a 460?

So we come to fov.

The bottom line for me is i dont have another 750 euros so why struggle to find the money when the 428 is basically the same?

B.T.W i would get the 460 if i had the money.

Hence all my other questions. If i want field of view perhaps i should get a 383? Its mono has a pixel size of 5 and is "only" 1200£ second hand...theres a saving right there but its not as sensitive....but what is "not as sensitive" in real world hard cold figures?

I don't think theres a right answer so i'm just playing devils advocate.

Sent from my GT-I9003 using Tapatalk 2

[perfrej]

I think you are looking at less sensitivity and s/n that you easily match by using just a few more subs with the 8300. I love both of my 8300s, the QSI 683 slightly more than the SBIG ST-8300.

I have tried to find relevant specs for the ICX694, but it appears that Sony doesn't really want to give out the numbers. For instance, how good is the anti-blooming protection?

/per

[Gina]

I'm afraid I'm not a linguist - is that Quantum Efficiency in the y-axis?