Problem with Flats

One other question - is the 383L NABG?

dmcdona wrote: One other question - is the 383L NABG?

I do not see anything about the anti blooming gate in the spec, however it is the KAF-8300 chip which is supposed to be an Anti-Blooming Chip

Simon - it looks like ABG alright. I suspect that, given your data, it makes little, if any, difference.

One you thing you may ne pleased to hear is that it appears that the KAF-8300 has an overscan region. Kodak call them "Dark Reference Pixels" in the spec sheet. But also note that there are other "dummy" pixels which Kodak point out should *not* be used as dark reference pixels - there are three types on the 8300 chip - Dark Dummy Pixels, Dummy Pixels and Virtual Dummy Columns.

What you need to do now is figure out how to capture that region - this region will give you a accurate signal offset data for the signal offset column. Once you have figured out how to capture the region, you need to then figure out if it already contained within the images you took (I doubt it). In that case, you need to redo all the images - darks and lights and light-darks - in overscan mode (NOTE - do NOT do a 7200 dark again). You will then have all the data you need to accurately characterise the CCD. (I ednded up doing all my data twice but did actually get better results the second time because I had a better idea of what I was doing).

The final hurdle for now is getting the overscan region. You should see if your acquisition (or driver) software allows it (perhaps as an option) or checking the manual or contacting the manufacturer. I have heard that some CCD manufacturers do not supply any means to access the overscan region. If so, that's a deal breaker for accurate characterisation unfortunately. See what they say anyhow - they may have some internal software they *might* share with you.

You've a lot to do...

Dave

Added the info into the spreadsheet and emailed it over to you Dave

Many thanks for your help on this, much appreciated

I didn't do the additional three frames yet

Simon

I've asked the question

www.atik-cameras.com/forum/index.php?topic=1869.0

Simon

Some good stuff posted in this thread.

I think Dave's basic thrust (bearing in mind that your original posted query is about an odd-looking flats image) is go back to basics in case using special software features of software packages might be confusing the situation - and the more you know your CCD the better - Dave can correct me if I've mus-interpreted.

One does not actually HAVE to do CCD characterisation - doing bias, darks, and flats should suffice - and - as Dave pointed out - bias frames are really only needed if one is doing dark libraries - for the purpose of scaling darks to match different temperatures - or image frames of differing integration / exposure.

Doing CCD characterisation - if nothing else - is good to do because you get to know things about your CCD specifically rather than what the spec says.

You can determine how many electrons per ADU (its gain) - what full well capacity is - where is the detector linear in response to light - and where does it become unreliable for accurate measurements of brightness.

But it could be that this is unnecessary for your initial query - if you want to do it I should be able to offer another voice (probably saying same thing - different words) to Dave - sometimes seeing something expressed a little differently can provide that A-HA! moment.
The more you know (and one can determine a great deal about a CCD's characteristics with all sorts of clever tests using a faint light-box) about the CCD the better in being able to have confidence in the quality of measurements.

It may be that I've missed the main thread here - I'll read more thoroughly and follow-up.

if I am on thread - problem is flats through one filter - stands out as being particuarly un-flat?
If that's the case then maybe a stand-back and re-think is in order - I will help if I can.

Regards

Mark C.