Vignetting Problem with Optical Train

[niteman1946]

Hi group,
I (finally) have a full complement of components for CCD work.   

This shows my optical setup of Meade 12”LX200 Classic, Feathertouch focuser, Meade f/6.3 FF/FR, Mitsuboshi OAG5 off axis guider, Atik EFW2 (w/1 ¼” mounted filters) and Atik 383L+mono.  In between the major components are adapters, spacers, etc.
http://www.astrobin.com/full/30273/?mod=none
The distance from the focal reducer back shoulder to the focal plane of the CCD is 108.3mm.  Depending upon whom you ask, this value should be either 105mm or 95mm.  With 105mm the calculated focal ratio is f6.11.  And with 95mm the focal ratio is f5.53.  Based on the scale of the image, I’m inclined to believe that the 105mm value is correct.

But on to my issue,
This is an un-stretched image of the Master Flat taken through this setup and processed in PixInsight.
http://www.astrobin.com/full/30266/?mod=none

And this is the same Flat selectively stretched to highlight the sharp edge of the vignetting.
http://www.astrobin.com/full/30267/?mod=none

And this is an integrated image of M78 with an auto-stretch.
http://www.astrobin.com/full/30268/?mod=none

Even with the Flat (and Bias and Dark), I’ve had no luck processing the corner features out of the M78 image.  PixInsight has multiple tools, but none have provided an answer with my skill set. So far, my only solution has been to crop out the left and right sides of the image.
Any thoughts on a better or different approach?
Thanks,
Mark

[topboxman]

Hi Mark,

Perfectly normal. That's how telecompressor works.

I get the exact same thing with my slightly larger CCD of SXVR-M25C, Celestron F/6.3 FR and C-8.

To fix it is by either cropping the image or applying flats.

I love PixInsight and I use every possible tools EXCEPT calibration. I never have good luck with PixInsight's calibration. I use Nebulosity for calibration and then I use PixInsight for everything else.

If you still have problems, can you upload one light sub, stack of bias, dark and flat. Zipping all files would probably be a good idea.

I have the same Hutech OAG-5 as yours. I love it. I have Helical focuser for the guide port and I also love it. I highly recommend Helical focuser.

Peter

[niteman1946]

Thanks for your response, Peter.

Am I correct in understanding that Nebulosity might do a better calibration job than PixInsight?

Looks like uploading zipped files for the Bias, Flats, Darks, and Lights in SendSpace (alternatively in PixInsight's Endor) is going to take forever.  Zipped files of half the stacks will likely take around 8hrs.  My upload is really slow, sorry.  That means your downloads will take a substantial time.  Is it worth the effort?

Mark

[topboxman]

Hi Mark,

I just don't think I am using PixInsight calibration correctly. Nebulosity is easier and quicker. I am suggesting to try a different software to see if you get different results. I use Bad Pixel Mapping and I can't get BPM'ed images to work with PixInsight so I gave up. BPM works better than dark subtraction with Sony CCD cameras.

Uploading images is always slower than downloading images. I am not asking for every single sub of bias, dark and flats, just a stack of each and only one light subs totaling four files. For me it probably take me about 30 minutes to upload but probably less than 5 minutes to download. Don't worry about it.

Is your last image calibrated with flats?

Peter

[niteman1946]

Thanks Peter,

Sorry to be so dense,
do you want my Masters in Flats, Bias and Darks?  Or do you want a few each of the individual Flat, Bias and Dark subs?

Yes, the M78 image was fully calibrated, aligned and integrated in PI.

Thanks,

Mark

[topboxman]

Yes, Masters in Flats, Bias and Darks and one light sub totaling four files.

When I said a stack, I meant Masters.

Did you check every single sub of bias, darks and flats before creating a Master of each? Sometimes there can be at least one bad sub due to a glitch during download from camera to computer.

Peter

[topboxman]

Hi Mark,

Just to be on the same page, your camera has a mechanical shutter which means you may need to expose about a minimum of three to four seconds when taking flats to avoid shutter shadow effect. When taking flats, make sure the peak is at about the middle of the histogram. I used to look for maximum ADU of the image but now I rely on histogram and get good results.

Peter

[niteman1946]

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

[topboxman]

Hi Mark,

The first two has issues downloading. I am able to download the last two in your list. Your flat looks really clean when opened with PixInsight.

I forgot that if you created a Master with PixInsight, I will not be able to calibrate with Nebulosity because Nebulosity cannot open FIT files created by PixInsight. So I will not be able to calibrate your light sub with Nebulosity.

Hopefully others will be able to help you.

Peter

[Dave Watson]

Hi Mark,

Being having a play with your images in Pixinsight.......ummmm, very odd

The Bias master seems very strange, it is streeked, I use the same camera and don't get anything like this problem.

I both batch pre-processed and integrated your images and I get the same result as you. Its almost as if the Flat master is not being taken into account.

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

I will keep playing.

Dave

[topboxman]

Hi Mark,

I was able to download the first two images this time.

Dave is correct about the Master Bias. It does not look like your Master Dark. Did you capture Bias subs at the same temperature as Lights? What exposure times did you use for Bias? It looks like as if each sub may have RBI (Residual Bulk Image). Maybe try put a 10 second delay in between subs so that ghost image won't be transferred to the next sub.

Are you capturing using native or Atik ASCOM driver? If you are using ASCOM driver and did not upgrade latest Atik ASCOM driver, then older Atik ASCOM driver has issues with taking Bias at less than certain exposure times because of mechanical shutter. The new Atik ASCOM driver now allow fastest shutter speed regardless of mechanical shutter as long as you cover the camera or capture subs in the dark.

I always play it safe by putting a 10 second delay for Flats, Darks and Bias to avoid possible RBI.

Also, are your Master Flat and Dark calibrated with Bias? If so, it might be more accurate to create another set of Master Flat and Darks again but not calibrated with Bias so we can evaluate a little better.

Peter

[niteman1946]

Hi Dave,
The Master Bias was taken using Artemis at 0.001 seconds.  I previously used Images Plus to capture bias subs, and Mike Unsold for some reason limits the shortest bias exposure to 0.20 seconds.  What I understand is that the Bias sub should be at the shortest exposure possible.
In both cases the bias subs showed a somewhat darker bottom than top.  When integrated, this became more pronounced.  The streaking is there with either the 0.001s or 0.20s groups.  However it is coarser with the 0.001s.

Would you mind posting an image of your stretched and non-stretched Master Bias?

Your comment about Flats not being taken into account rings true to me.  This has been my opinion “at times” also.  I suspect the fainter the object, the more necessary the stretch, and then the more noticeable the vignetting.  I asked this question of tech support previously.  One thing they suggested was moisture and ice contamination (presumably not on the flats, but on the lights).  I’ve since cooked the desiccants twice to minimize that possibility.

Yes, I do mostly follow Vincent’s summary (mostly).  Actually I have a hard copy of a more detailed summary that Astropixel did on the PI forum.  It no longer comes up unfortunately.  I take one exception to Vincent’s approach.  I do not calibrate the Flat and Darks initially.  I simply integrate both groups and then select calibrate at the Master Dark and Flat section during Image (Light) calibration.  I have tried it both ways with no noticeable difference.

Mark

[niteman1946]

Hi Peter,
See my comments to Dave on the Bias.  My records do not show that I cooled the bias subs.  The Bias master I uploaded was made from 0.001s subs.
I had never seen the term RBI and now have come across it twice in one week.  I will make sure there is at least a 10s separation between subs.
I can (and did) take the bias subs at very short exposure (i.e. 0.001s) using Artemis.  Ascom is another story.  I can’t get Ascom Platform 6sp1 to play nice with my scope.  It won’t connect. So I’m staying with Platform 5a until the 6sp1 problem is resolved.

I do not initially calibrate the darks and flats with the Master Bias.  All calibration is done within the Image (Light) calibration operation.  I have tried both ways with no noticeable difference.

Appreciate your and Dave’s help.

Mark

[Dave Watson]

I have not heard of RBI either, but as Mark commented I will also take Peter's lead and give a 10s delay for Biases, Darks and Flats in the future, althought I have not had problems todate.

Here are the links for my files

http://www.qdigital-astro.com/dropbox/Bias_Master-20.fit.zip
http://www.qdigital-astro.com/dropbox/Dark_Master_300-20.fit.zip
http://www.qdigital-astro.com/dropbox/Flat_Master_L-20.fit.zip
http://www.qdigital-astro.com/dropbox/L_300-20_i.fit.zip

All files are have not been stretched, I don't have stretched versions of Bias, Dark or Flat masters.
The file L_300-20_i.fit is a batch pre-processed and integrated of M31, that has not been stretched or had any other work done on it.

I use Vincent's tutorial to the letter as it seems to work for me.

The exposure setting for Bias frames should be the minimum that the camera will achieve, for the 383L this is 0.2s. The 383L specification states that this minimum exposure is the recommended exposure to minimise the effects of vignetting, but find the minimum exposure is more like 3-4s to prevent shutter smearing when I take Flat frames.

Dave

[niteman1946]

Hi Dave,
Interesting stuff.  Are you running anything similar to my optical train setup?  That is 12”LX200, FF/FR, OAG, EFW2, Atik?  How are you setup?
I do suspect that my setup is causing this vignetting.  Although, although …., I was informed that F5.0 or greater should not suffer this.  Oh well.

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 ?)
b) What was your temp setting? (mine was OFF)
c) How many bias subs were used? (20 ?) (mine was 30)

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?
a) What was the exposure? (300s ?) (mine was 600s)
b) What was the temp setting? (mine was -10C).
c) How many subs were used? (20 ?) (mine was 30)

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)
b) What was the individual sub ADU value? (25K-35K ?) (mine was about 30k)
c) What was the temp setting? (mine was -10C).
d) How many subs? (20 ?) (mine was 30)

4. M31:  Damn, that looks good!  That wasn’t taken with a large scope, no?

Mark