CCD's .....Kodak KAF8300 v's Sony ICX814

[swag72]

How many times have I read on this forum about people wanting to buy a CCD camera and being stuck between these two choices......

A little bit of background if I may to say why I feel I have some ideas on this.

I've used  the Sony ICX814 for over a year (In a QSI690) - I bought it because I'd always used Sony chips and had heard about the noise etc with the Kodak chips and frankly it scared me. No darks needed on the Sony chip ....... more on that in a bit ........ and I felt comfortable using them.

Fast forward to a year later............

Even using a 330mm focal length, I found I was having to do mosaics for many of the DSO's - 2x2 pane's or even bigger. It was time consuming and at times soul destroying! The small pixels were supposedly a good match for my shorter focal length scope, but not so for longer focal lengths. Then I found out via the US Cloudy Nights forum that there is an issue with the earlier QSI690's of Sony Amp glow. This is where, when doing long exposures, you get what looks like amp glow around the edges of the frame. 

Easy to calibrate out with darks ............ hence why I began doing them and found that they really did decrease the noise even on the Sony chips, so I continued taking them. But that wasn't the point, and so QSI offered to take the camera back and fix the Sony Glow.

Fast forward to now ............. and I have my camera back!!! But it's changed into a QSI683 as the excellent folks at QSI agreed to swap my sensor for the Kodak KAF8300.

So now I'm in the position of having used the Sony chip for a year and I've just done my first light with the Kodak 8300.

Initial thoughts

There is nothing to be scared of......

• The fov on the KAF8300 is so much bigger than the Sony chip that it's suddenly a joy to work with it again.
• The noise etc calibrates out easy enough - I've changed nothing in my stacking and there's little by way of noise.
• The pixel size of the KAF8300 gives me a resolution of 3.39 arc seconds ................. It would appear that this is well outside of the recommended resolution, but everything is fine, there's no untoward issues as far as I can see. The detail on the sub looks good.
• The 1.25" filters work absolutely perfectly with this camera. There is little by way of vignette at f3.9 and flats have sorted it all out.
• Significantly less QE than the ICX814 doesn't show itself in my Ha data

I am really pleased with this change and I have put my mind at rest about a few of the perceived issues with the KAF8300 sensor. 

I've heard people say that it's a bit long in the tooth now - That may be so, but by goodness does it produce the goodies. 

I've attached my first light with the camera - This is still going to be a mosaic (albeit only 2 panes) but I know that the ICX814 would have taken 4 panes. 

To me, having used the Sony chip for over a year and now this Kodak chip for one session - It's a no brainer for me........ I don't understand why I ever bought the ICX814.......The advantages of the Kodak far outweigh any of those on the Sony chip.

These thoughts are only my opinion of course, having used both I'm sure that after a few images with the KAF8300 my views will remain the same

[MarsG76]

Good Report, I'm in the market for a CCD.. and was weigh up between the QSI683 and a Sony Chip.. my question is if both were the same size, would you go for the Sony or still the Kodak 8300?

[swag72]

Good Report, I'm in the market for a CCD.. and was weigh up between the QSI683 and a Sony Chip.. my question is if both were the same size, would you go for the Sony or still the Kodak 8300?

But they're not the same size by a margin and that's the whole point of this post ...... And I doubt that they will ever be in their present form

[MarsG76]

But they're not the same size by a margin and that's the whole point of this post ...... And I doubt that they will ever be in their present form

The ICX493 chip is bigger then the KAF8300, by about 5mm... It's actually slightly bigger then the APS-C sensor. So I'm wondering whether the Sony sensors have a advantage over the Kodak if they were same size in your opinion.

[ChrisLX200]

Where the Sony chip gains is the fantastic QE (including sensitivity @ Ha wavelength) compared to the 8300, so for faint narrow-band targets I'll still go with the 814. For larger targets I agree - bigger chip wins even if you have to use longer subs. I was pondering an upgrade for some time and decided to go for the 16803 sensor for wide field subjects despite the fact its QE is only 45% as opposed to the 814 at 75% (still a lot better than the 11002 at just 30%!). Only other drawback - buying the NB filters in 50x50mm size!

ChrisH

[Davey-T]

Thankyou for this report Sara, I'm about to drop around £5000.00 on a QSI683 and full set of filters to try and capture as much of the FOV of the WO Star71 45mm imaging circle as possible, it is reassuring to know it works

Dave

[swag72]

The ICX493 chip is bigger then the KAF8300, by about 5mm... It's actually slightly bigger then the APS-C sensor. So I'm wondering whether the Sony sensors have a advantage over the Kodak if they were same size in your opinion.

Isn't that just a colour sensor? I'm not aware that they do it in mono.....

Then we get onto another interesting debate!!!  .......................... And having used a friends OSC with a Sony sensor, I'd never touch one with a bargepole!!

[swag72]

Where the Sony chip gains is the fantastic QE (including sensitivity @ Ha wavelength) compared to the 8300, so for faint narrow-band targets I'll still go with the 814. For larger targets I agree - bigger chip wins even if you have to use longer subs. I was pondering an upgrade for some time and decided to go for the 16803 sensor for wide field subjects despite the fact its QE is only 45% as opposed to the 814 at 75% (still a lot better than the 11002 at just 30%!). Only other drawback - buying the NB filters in 50x50mm size!

ChrisH

I didn't think my Ha image posted was too shabby Chris. I certainly expected to be very underwhelmed compared to the Sony sensitivity, but that was not the case. As you know, I am a NB junkie primarily, and still know that the camera is the right choice.

[martin_h]

I have only used the Kodak chip and wouldn't change it. As to noise, yes its noisy but as posted above easy to deal with, in fact I don't use darks, but use a bad pixel map ( can you imagine doing a set of 30 x 30 mins darks!!)

[ChrisLX200]

I didn't think my Ha image posted was too shabby Chris. I certainly expected to be very underwhelmed compared to the Sony sensitivity, but that was not the case. As you know, I am a NB junkie primarily, and still know that the camera is the right choice.

It's excellent Sara, and the 8300 has more than twice the well depth offering better dynamic range.

ChrisH

[swag72]

....... the 8300 has more than twice the well depth offering better dynamic range.

So even as a NB junkie, there's not a big advantage to the increased QE of the Sony sensor when compared to the increased dynamic range......

[Earl]

have you looked into drizzle integration?

Yours arrives... mine goes....such is life.

[Earl]

oh my tip is defo use a cosmetic correction while calibrating too, works wonders.

[ChrisLX200]

So even as a NB junkie, there's not a big advantage to the increased QE of the Sony sensor when compared to the increased dynamic range......

apart from shorter subs - what you get in 15min with the Sony will take 20min for the 8300...

ChrisH

[Earl]

apart from shorter subs - what you get in 15min with the Sony will take 20min for the 8300...

ChrisH

but you need to double the amount of subs to cover the same area ish... its give and take

[ChrisLX200]

but you need to double the amount of subs to cover the same area ish... its give and take

Yes, unless the target doesn't need the bigger area - faint planetary neb halos for example. There's no 'one camera fits all' for us sadly.

ChrisH

[Earl]

Indeed, or is it one camera and specific scopes for different targets, i think that's a better plan with a touch of theory crafting too.

[pete_l]

• Significantly less QE [ pete_l: 50% at Hα for the 8300 chip, compared with 65% for the ix814 ]  than the ICX814 doesn't show itself in my Ha data

An interesting observation. Could you venture an explanation for why this would be?

[swag72]

An interesting observation. Could you venture an explanation for why this would be?

Here's a QE graph that is readily available on the net comparing a various number of chips. That's where my understanding stops I'm afraid

[davew]

A rather interesting report Sara, thank you.

As someone who has used the Kodak full frame chip for over four years now, I have to admit I'm glad I didn't have the choice of these Sony chips back then. It would have taken even longer to decide ! There was only the 285 chip and that's really small.

As you've just found, the 8300 chip is all but twice the size. That is a great incentive to buy the 8300. As so many do, I too thought it was a noise bucket when I first got it but that's because I insisted on using 5 or 10 minute subs for narrow-band for a while and then compounded the problem by only taking a few darks. You, on the other hand don't play about with time and do things properly. Same with darks  and flats.

If people want to see a video then go to YouTube and see Kevin Nelson of QSI describing these two chips. Although I think he's " Massaged " the data slightly he does make the case that each is as good as the other and it's just the sizes of chip and pixels that matters to the individual.

The filters don't have to be more expensive with the 8300 so long as you chose a camera that keeps the filters fairly close to the chip. QSI and Moravian do this very nicely. On a ridiculously fast system then the 31mm unmounted filters do come in handy. 

The QE and smaller pixels of the Sony could come in handy with very short focal length imaging. I'm thinking along the lines of camera lenses that would also enable shorter sub times if necessary with less accurate mounts.

I think you've made a reasonable case for not using noise as an excuse not to buy the larger chip. Buy the chip that suits your imaging style and interest. 

Dave.

[Uranium235]

(spits out his coffee...) Gawd blimey, Sara plumps for an 8300!! Welcome to the club

Nice having a lot more real estate isnt it? Your reasoning was pretty much the same as mine, although its nice to have a sensitive camera (I had two 314's running simultaneously which pulled data in at a daft rate), I still found myself doing mosaics due to the chip size. Well, I still do to some degree - but things get done faster overall with the 8300.

Dont worry about the resolution, you will be fine up to about 3.5" p/p. The great thing about that chip is that it is very flexible, so if you get a longer FL scope (ie: you get to use the RC again) then it will be very happy on that too - especially when binned.

[Ikonnikov]

It's excellent Sara, and the 8300 has more than twice the well depth offering better dynamic range.

ChrisH

I had assumed this too but the figures on the QSI website actually have the ICX814 with a higher dynamic range (73db) than the KAF8300 (70db) since the much lower read noise (3e^- versus 8e^-) more than compensates for reduced well size (dynamic range being calculated from full well capacity/read noise). Interestingly the Atik 383L+ (KAF8300) is quoted by the manufacturer as having a 7e^- read noise despite being considerably cheaper than the QSI. Alas the Atik ICX814 camera (which I own) is quoted at 5e^- giving it the lowest dynamic range of the lot. Still liking it for narrow-band though.

Paul

[Scott]

I have only used the Kodak chip and wouldn't change it. As to noise, yes its noisy but as posted above easy to deal with, in fact I don't use darks, but use a bad pixel map ( can you imagine doing a set of 30 x 30 mins darks!!)

yep. start capture, go to bed, get up and go to work, come home....darks done

[swag72]

(spits out his coffee...) Gawd blimey, Gina plumps for an 8300!! Welcome to the club

*spits out coffee* Gina? ........... Me? ............ I'm not Gina ............

[JamesF]

*spits out coffee* Gina? ........... Me? ............ I'm not Gina ............

I was a little confused there myself

James