Vignetting Problem with Optical Train

[Dave Watson]

Hi Mark,

My setup is Tak FSQ85 / EFW2 with Baader 2" filters / 383L+, for these images I didn't use my reducer.

When I take Flats I always use a Electroluminescent light source and add sheets of paper to get the required exposure of not less than 5s, I found that anything less than 5s gave shutter smearing.

I always take Biases,  Flats and Darks at the same temperature as I take the Lights at, in this case -20C, so all where taken at -20C.

See below for answers in RED.


Dave


1. Bias:  Your bias master looks different than mine.  Yours is more normal (even), with brightening along each vertical side.  Mine has moderate to heavy darkening towards the bottom 1/3rd, and streaks.   My bias subs were taken using Artemis at 0.001s and no temperature control.  However, I also have bias taken with IP at 0.20s and -10C.  My IP bias master still shows darkening in the lower 1/3rd, but much smoother - overall similar to yours.
a) What was your exposure? (0.20s ?)0.1s
b) What was your temp setting? (mine was OFF)-20C, ALWAYS same as Lights
c) How many bias subs were used? (20 ?) (mine was 30)35, 30 should be OK

2. Dark:  Your dark master does show darkening in the lower 1/3rd , just like my Bias master.  My Dark master is uniform throughout.  Sure you didn’t accidently switch the two?Have checked, no they are correct
a) What was the exposure? (300s ?) (mine was 600s)Same exposure as the Lights, in this case 300s
b) What was the temp setting? (mine was -10C).-20C
c) How many subs were used? (20 ?) (mine was 30)35

3. Flat:  Your flat master looks good.  I can selectively stretch in PI and see that there is no sharp edge to the vignetting.  Whereas mine does show a sharp edge (see my original post).
a) What was the exposure? (3-4s ?) (mine was 5s)6s for the L
b) What was the individual sub ADU value? (25K-35K ?) (mine was about 30k)I always check the corners and the centre of the Flat and make sure they both are between 23K and 33K, generally they are 25K-29K
c) What was the temp setting? (mine was -10C).-20C
d) How many subs? (20 ?) (mine was 30)35, but in actual fact because there is a lot more signal 12 would have done

4. M31:  Damn, that looks good!  That wasn’t taken with a large scope, no?No I used my FSQ85 at FL 430mm @ f5.6

[bwa]

Hi Peter,

1. I have reviewed each of the flat subs.  All look good.
2.   4sec appear to be the safe minimum period for flats.  3sec or shorter tend to run into the shutter issues.  My flats were shot at 5sec.
3.  Here are the Master files along with one Light file.

http://www.sendspace.com/file/jkk6dz
http://www.sendspace.com/file/l7v5bn
http://www.sendspace.com/file/4l3k4x
http://www.sendspace.com/file/42cmuu

Mark

Mark,

Is there a way to download these files w/o installing yet another piece of junk software on my PC?

bwa

[Dave Watson]

Hi Peter,

1. I have reviewed each of the flat subs.  All look good.
2.   4sec appear to be the safe minimum period for flats.  3sec or shorter tend to run into the shutter issues.  My flats were shot at 5sec.
3.  Here are the Master files along with one Light file.

http://www.sendspace.com/file/jkk6dz
http://www.sendspace.com/file/l7v5bn
http://www.sendspace.com/file/4l3k4x
http://www.sendspace.com/file/42cmuu

Mark

Mark,

Is there a way to download these files w/o installing yet another piece of junk software on my PC?

bwa

Just click on the link "   Click here to start download from sendspace", no software needed.

Dave

[niteman1946]

One question, have you created your masters in accordance with the tutorial http://pixinsight.com/tutorials/master-frames/en.html ?

Dave

Hi Dave,

FWIW, I went back and followed Vincent's tutorial to a tee.  No luck.

Mark

[bwa]

Hi Peter,

3.  Here are the Master files along with one Light file.

http://www.sendspace.com/file/jkk6dz

Mark

Mark,

Was the cooling working on your CCD?  I downloaded your Light.  It is probably one of the noisest images I've ever seen!  Just curious if this is typical of the 383L camera?  If it is, I'm wondering how Darks are working to clean it up?

bwa

[bwa]

I did take a fast pass at your Light image. 

I used GradientXTerminator to reduce the vignetting.  With a bit more effort I could probably eliminate the remaining "shadow".  The vignetting is significant but not unlike vignetting I've seen with similar combinations of scope, reducer and camera (http://tinyurl.com/bwaReducerVignetting).  To further reduce the vignetting I'd suggest reducing the camera to reducer distance; however, this will increase the f ratio.

[niteman1946]

Mark

Was the cooling working on your CCD?  I downloaded your Light.  It is probably one of the noisest images I've ever seen!  Just curious if this is typical of the 383L camera?  If it is, I'm wondering how Darks are working to clean it up?

bwa

Hi bwa,
I use Images Plus Camera Capture and had the cooling turned on to -10C.  When I look at the Image Properties window in IP, it reports a set and CCD temp of -10.04C.

Would you mind describing to me what you see that suggests noise/no cooling?

Thanks,

Mark

[bwa]

Would you mind describing to me what you see that suggests noise/no cooling?

Mark

The attached 100% crop out of your Light sub is just full of specks/hot pixels/etc.  I consider this a very noisy image!

bwa

[niteman1946]

I did take a fast pass at your Light image. 

I used GradientXTerminator to reduce the vignetting.  With a bit more effort I could probably eliminate the remaining "shadow".  The vignetting is significant but not unlike vignetting I've seen with similar combinations of scope, reducer and camera (http://tinyurl.com/bwaReducerVignetting).  To further reduce the vignetting I'd suggest reducing the camera to reducer distance; however, this will increase the f ratio.

Thanks Brian,

Pity there's not a GradientXTerminator plugin for PixInsight.
I appreciate what you are saying about modifying the setup.  I can move the camera closer to the focal reducer by about 7.5mm.  Might be worth a try.
I also suspect that my Feathertouch focuser is problematic due to its length.  But would hate to loose it.

Mark

[niteman1946]

The attached 100% crop out of your Light sub is just full of specks/hot pixels/etc.  I consider this a very noisy image!
bwa

Thanks Brian,
That's what I suspected.  Just wanted to be sure.
I'm running this directly by tech support for their thoughts.  Will followup when I hear back from them.
Mark

[Dave Watson]

Hi Brain & Mark,

Yes I agree with Brian all the images, darks, flats and lights, seem very noisy to me...this is not normal with the 383 camera....I don't have a problem with this camera at either -20C or -10C(see the example links I posted).

Mark, don't forget that you can't just start changing the distance between the reducer/flattener and the camera sensor ad lib, you have to abide by the metal backfocus of the reducer/flattener as specified by the manufacturer.

Dave

[niteman1946]

Hi Dave, Brian, Peter et al,

Looks like there have been two separate issues (or maybe related) at work here.

Low Voltage:  The Atik recommended 12v power source was producing maybe 12vdc, or just under 12v.  I moved to the Pyramid 13.8v RPS and the "noise" problem went away (so far, and I think).

Reverse Vignetting:  This one is a little more weird.  This was chronic both before and after adding the EFW2 filter wheel (so no fault there).  It happened most of the time, although some images were fine.
  This latest go 'round had me shooting Ha of a portion of NGC2237.  The reverse vignetting was there in my calibrated subs and there in the finished product.  However, once I redid my Darks, Bias and Flats using the Pyramid 13.8v RPS, the Ha image was fine.  So "low voltage"?  Not sure.
  I then followed the same routine for the OIII image.  Used the same new Bias and Dark as Ha, but of course a new OIII master flat.  The reverse vignetting was back.  Reflecting on my life and the fact that tech support and I were running out of ideas, I decided to try the Ha master flat with the OIII image.  And the reverse vignetting was gone.  I did the same using the SII master flat with the OIII image and got good results.  So, the problem (at least in this case) was in the OIII flats.
  There were two noticeable differences between the Ha flats and the OIII flats.  The Ha exposure was at 8s, while the OIII was at 20s.  Both were shot using the t-shirt method with an attempt to achieve 25kADU value.  I used Images Plus (my capture software) to review the subs' statistics.
In both cases (Ha and OIII) the average ADU was around 25k.  But while the Ha max value was around 49k, the OIII max value was at 65,535, which I believe is the maximum. 
I reshot the OIII flats at 8s and monitored the images so that the average ADU was 25k and observed that the maximum stayed around 49k.  The reverse vignetting for the OIII image was gone.
  It appears the high maximum ADU had been the problem.  But, as with the voltage issue, time will tell.  I had not read anything suggesting that this could happen, and when it did would be a problem.  It was more dumb luck that I tried the Ha flat with the OIII image and found success.

Appreciate any thoughts you guys might have on this.

Thanks,

Mark

[niteman1946]

Sorry to persist with this thread, but one more point on the "reverse vignetting" phenomenon.  That is where the corners of the image over-brighten following calibration.
1.  It occurs if the flat sub's ADU value has reached 65k, even if the average is around 25k.  See my previous post.
2.  It occurs if the Light sub's temperature is higher than the Flats/Bias/Darks temperature.  My experience was with the Lights around +10C (forgot to turn the cooler on) and the F/B/D temps at -10C.  In this case I do not know where the tipping point is.  So far as long as all the temps are set at the same value this problem does not manifest itself.

Anyone seen anything like this with the 383L+mono?  I would like to hear that this is a recognized behavior and manageable.

Mark

[NickK]

Hi,
The 383L has a 16bit analogue to digital conversion which results in pixel values of 0 to 65535 (2^16).

Are you normalising with the PI calibration? If the PI capture is scaling then it could be over brightening during the calibration. I understand the maths behind the calibration, but I found the calibrate checkbox behaving different to what I would expect.. I normally do calibrate without a checkbox calibration of flats etc.

I usually run with about ~20-40K values at 3 seconds which is just long enough to prevent the shutter causing a problem for my 383L mono. Note that you must NOT perform any form of histographic manipulation on flats - the need to be linear.

I would run a 3D plot on the image (sliced through and scale the Z axis to 100%, and use 2 elevation) and you should see the shape of the flat. So here's a flat integration slice (Lum filter) used with the M42 on the Pentax... the centre is ~0.46 the outter edges are ~0.40 to show the range:

[niteman1946]

Thanks for your thoughts, Nick.
First of all I'm out of town, so I can't follow-up with some of your points. But I will when I get home.

I started this post in January but as said, the problem has been a problem since the beginning. It has just taken me this long to discern "some" ways to make it occur (examples from my last post).  And now how to avoid it (so far).

I originally thought PI was a contributor. But I ran one of the problem sessions in DSS, and got the same results.

So my concern at this point is whether or not I should have Atik look at the camera. So far no one has said they've seen the same thing and not to worry.

Mark