First light with Atik 314L+... and a bit lost!

[AndyUK]

(Yes - That's what I was thinking... . I actually had two of them 3-4 months ago but sold them on... )

[Zakalwe]

This might be a daft thought, but what temperature had you set the camera too? Is there a possibility that it was over-cooled and some frost condensed on the glass front or on the chip itself?

[AndyUK]

Hi Zak - On these frames, I set the temp at 0C... I've since set it to -10C though. I did take some more test subs last night @ -10C with the Equinox 80. I've documented what I did, but haven't really had a chance to review the script with the frames...

The first test involved taking 2 min frames and rotating the camera, the 2nd test involved exactly the same but with the OTA cover on, the 3rd was the same with the camera covered with a small towel, and the 4th was OTA cover on and the whole scope and camera covered with a towel - The last test frame(s) looked fine until I used autostretch and then the gradient reappeared...

Before jumping to conclusions though, I want to see what happens if I stretch one of my "stock" darks (ie taken with lens cover on camera). I'll hopefully get the chance to look at them in more detail tonight (I'm not on the right machine at the moment...)

[Zakalwe]

The last test frame(s) looked fine until I used autostretch and then the gradient reappeared...

So the gradient appeared with the OTA cap on, and the whole kit and caboodle covered??

Doesn't sound right at all. Possible faulty camera?

[LunarLight]

gutted mate I hope you get this sorted soon, must be frustrating.

This may be an irrelevant point but I know you have mentioned a problem with street lights before and I once had similar light leakage in my scope/camera. I tracked it down the back of my scope being bathed in light from a street lamp and some of it sneaking through a gap between the tube and primary mirror. Of course this might not be the issue and you may have already looked into it plus having never seen the MN190 or your other scope it may be the case the your scopes dont have the same gap between tube/mirror as my explorer....still worth looking I guess

Mark

EDIT: just noticed this: "4th was OTA cover on and the whole scope and camera covered with a towel" I guess that rules out the possibility of leakage from the back ...hmmm

[AndyUK]

A busy week this week, so I haven't been able to get out or conduct any tests... until this morning. Admittedly only using a temp of 0c, I bumped up the exposure to 300s and took a straight dark with the Atik with the lens cap on (no OTA) - I got a gradient. So I wrapped it up... and still got a gradient - I think I might be getting somewhere....

[roundycat]

Are you sure that gradient is not Bias. Does the gradient vary with chip temperature? Take a Bias at the same temp and subtract it from the Dark.

Dennis

[Zakalwe]

Take the camera out of the scope. Cap the nosepiece. Then take a dark. You are removing a lot of variables if you do this.

[AndyUK]

Hi Dennis - If it IS bias, there's an awful lot of it... When I get home tonight I'll post something up for you to see...

Thanks for the suggestion Zak, but that's exactly what I did in the last 2 frames... Again, it will probably be easier to understand what I'm talking about if I post something up later.

(Cheers again)

[AndyUK]

Okay, I'm back home now - Nadeem kindly sent me one of his files taken with the same parameters and certainly there's no such "artifact" in his dark frame, but even so, perhaps I can ask others to pass judgement on these...?

I've put these two screenshots into a word doc (direct from Artemis capture) as they're too big to post as FITS and I didn't want to make any manual adjustments of my own at all in order to reduce them. Both are taken with no OTA and the lens cap on the Atik - The first was taken with the camera on the table (non auto-stretched), and the second was taken with the camera wrapped in a towel (auto-stretched).

Although the first frame isn't auto-stretched, you can still make out the gradient at the bottom, and there was actually no difference at all between the two frames (covered or uncovered)

Is this amount of gradient normal, and isn't this rather a lot of noise if it's bias? .

[Note: If anyone would like the raw fits files, please PM me...]

Atik Test 20110507 2003.doc

[swag72]

So Andy, let me get this straight in my head - These images above are taken as darks with absolutely no light getting to them at all?

[AndyUK]

Hi Sara - Yes - These are Darks taken with the lens cap fitted on the camera... and in the 2nd frame, the camera was also wrapped in a towel to prevent the possibility of ANY light getting in from ANYWHERE.

[roundycat]

I thought I would post this to show what Bias frames can look like and how much of that Bias component is present in the Darks.

The pictures are screen grabs from MaxIm and show the Bias frame from an STL 11000k camera at -20C. When warmer the bias gradient is much worse. The same gradient is present in the Dark but the Dark frame was exposed for five minutes so has a lot of thermal pixels in it that are not present in the Bias.

The pair of pictures shows just the vertical centre of the single Bias and the single Dark in order to save space.

The single picture shows the single dark after having had Bias subtracted from it. Not something you would normally do but it serves to show how much the Bias can show up in a dark, light or Flat frames. All the partial defective columns and, most importantly, the gradient have gone, all that is left are the thermal pixels which is what you would expect.

If you suspect a problem with the camera or if you suspect there is an external gradient problem it is a matter of moments to do this simple test to see where the problem lies.

The single picture has a much more enthusiastic stretch in order to show the evenness of the overall field.

Clearly there are no problems with my camera but if you have a bright area or gradient look to your Bias first before suspecting everything else.

Dennis

[AndyUK]

Thanks Dennis - I'll try that tonight - Will it matter that my bias frames are 0.001s? (I know you've mentioned that even 0.1s should be fine, but will it make any difference? I can always take another one...!)

[roundycat]

Short answer - no. I would still go for a longer time such as 0.1s.

Dennis

[AndyUK]

Okay Dennis - Here's a word doc with screenshots of an unstretched and stretched Bias frame taken at 0c, 0.1s... Unfortunately though DSS doesn't allow me to just subtract the Bias from the dark so I can't do that test you suggested , but going from these bias shots, can anything be deduced? (Maybe I can email a dark and a bias to you, if you don't mind?

Aside from the bands(?!) I can still see the light area at the bottom of the frame as well... (?)

PS I'm sorry to be a pain, but I'm close to tearing my hair out here...

Bias frames.doc

[mark7331]

Andy I've just created a bias frame on my 314L+ using exactly the settings you did. The outcome is attached below. I can't see the light that you're getting. I've also included a zipped version of the FITS bias frame itself so you can have a better look at the data.

Hope this helps.

Mark

Bias-0c-0100ms.fit.zip

Bias.doc

[roundycat]

Andy, by all means send me a bias and a dark. I'll do the necessary and send the result back.

If you have Photoshop you can subtract the bias from the dark using Image - Apply Image and Subtract. It will only do it in fifteen bit so may not be totally accurate.

Dennis

[AndyUK]

Thanks for that Mark - (I'd forgotten about zip files... DOH!)

It's very helpful to see an identical frame from another camera (thank you!). I know two camera's can't be 100% the same, but I would assume that they should be within a tolerance limit (I just don't know what that is...)

Dennis - Here's jpg versions of a PS bias subtraction, one straight subtraction and the other SERIOUSLY stretched - Before I email you the raw fits files, does this tell us anything? Again, both of these were taken with no OTA, purely the camera with its lens cap on...

[roundycat]

I'm afraid they don't tell me much. The stretching is important in some respects as going beyond what you would ever do to a 'real' picture is somewhat pointless. If this is a bias problem I would expect the bias subtracted frame to be a flat grey at the same stretch as the original non bias subtracted frame.

Dennis

[AndyUK]

Ah... So if I compare the straight dark with the bias subtracted dark, I should have a flat grey... I think I do mostly, except for a slightly brighter smudge bleeding into the frame about 1/4 across from the left at the top and (barely visible but still there) another smudge in the right hand corner...

[AndyUK]

Hi Dennis - The fits files zip down to just below the limit - Could I ask you (please?) to subtract and compare these and give me a prognosis? If there's apparently nothing wrong, then I think I'll be going back to my DSLR, but if there is an issue, I'll ask if it can be swapped out and I'll try one more time...

AS BIAS.zip

AS Dark.zip

[roundycat]

Ho hum! The plot thickens.

I opened both frames and displayed them side by side with a Range stretch in MaxIm. You can read off the pixel values, OK so far.

Next, I Bias subtracted the Dark and obtained a funny looking result. MaxIm would not let me subtract a Bias and as it reads the FITS header I had a look as well. The Atik FITS header is sparse to say the least but it doesn't say if the frame is a dark or a bias. Initially it subtracted the dark from itself which gave the funny looking result. Putting the Bias into its own folder and then re-trying gave the result below. A very smooth looking dark showing a range of 0-33, pretty dark!

The Bias does not show the bright patches but it seems to get rid of them from the dark. However, it does show some banding; could this be due to something else using the USB connection when the bias is being downloaded? It may be a camera characteristic in which case the banding should be present in the same place in every bias frame.

I have compared bias from the STL with the Atik. Although the STL bias is a master that will not make much difference to the pixel values. Just look at the level of the bias in the Atik. With that level of bias you must be careful to ensure proper calibration.

If you apply a very small blur to a single bias it will show you roughly what a master will look like. The banding is much more prominent

In summary, it seems as if the bright patches are a bias issue. With the high level of bias in this camera it is important to make sure bias is taken out of the lights, either by dark subtraction or bias subtraction if not using darks. The same caution applies to flats as well.

Dennis

[AndyUK]

Thanks Dennis - Your efforts are very much appreciated...!

Re: USB connection, no, there wasn't anything else attached to the laptop, and I was careful to ensure I kept the power lead from the USB cable. I guess it could be an "dodgy" USB cable though (surely not another problem on top of everything else?!).

As you say, this camera does seem to have a lot of bias signal and comparing to Mark and Nadeem's respective bias and dark frames, I have to say that I'm not happy that mine seem to be so different... I keep coming back to the fact that I've never experienced any of these types of issues with my 40D, and although the noise was getting to be a problem, it cost significantly less...

[roundycat]

If you can get single Bias frames from other, identical cameras and they are quite different it sounds like it is time to send the evidence to Atik and ask their view. Be careful to ensure the other Bias frames are taken at the same exposure time and temperature and are displayed and measured using the same software. (MaxIm display set to Range, screen grab and then jpeg it in Photoshop).

It is worth checking with another cable just to be sure, I assume the Atik is USB2?

That will not be the cause of the high bias level if it is high for that camera but it may cause the banding. Remember that the Bias is also present in the darks and subtracts out perfectly. By itself it is not a problem and may not have anything to do with the bright patches.

I would not consider for a moment going back to a DSLR. You will sacrifice a huge amount of sensitivity and flexibility.

Dennis