Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Misaligned filters in the EFW?

  • Please log in to reply
63 replies to this topic

#1 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 05 March 2021 - 09:56 PM

So, tonight's my first night with the new 294MM and 8 position EFW with the Antlia 1.25" LRGBSHO filters. I got an hour of Ha on the Rosette and 5 minutes of O3... but trees got in the way. So, I decided to slew back east to try some LRGB on the Leo triplet.

 

When I took a test exposure with the L filter, I got this:

 

gallery_347158_15202_1208457.png

 

I didn't see this effect on either the Ha or O3 filter... cycling through the other filters and I can sort of see it on a couple of them.

 

I've never used filters or a filter wheel before, so I am unsure how to address this issue. I tried using the "Recalibrate" function in the EFW ASCOM panel, but that didn't seem to have an effect. I also tried using the "unidirectional" method... cycled through filters, then back to L... same thing. Tried recalibrating again after switching to unidirectional. Same thing.

 

Ideas?

 

Thanks!

 

PS - mods, if this belongs in the equipment forum, please don't hesitate to move it there.



#2 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 05 March 2021 - 10:29 PM

Is this just normal vignetting, but exacerbated because of NINA's autostretch? Something not quite centered properly? I'm back to being the complete newbie again smile.gif


Edited by jonnybravo0311, 05 March 2021 - 10:29 PM.


#3 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 01:50 PM

I'm following up on this in the hopes someone will have some ideas. Being as last night was my first night using the new camera and filter wheel, I expected to encounter issues. As shown in my original post with the screen grab of NINA, it looks like the filter isn't properly centered. This morning, I started processing my data to see if things would calibrate out. They did not. A lot of screenshots are forthcoming ;)

 

First, here's a raw Luminance filter light sub opened and auto-stretched in PI:

 

med_gallery_347158_15661_3186475.png

 

You'll notice that the top and bottom right look to be cut off. My guess is it's the threaded edge of the filter mounting. Next up is a Luminance filter flat auto-stretched:

 

med_gallery_347158_15661_5797461.png

 

The "cutoff" is a bit less pronounced, but I attribute it to the flat image being considerably shorter than the 90s light. The flat is 0.13" with a median 32844 ADU. Here's the contour graph for that flat:

 

med_gallery_347158_15661_1554238.png

 

That looks pretty normal to me, but I welcome feedback. Next up is the matched dark flat:

 

med_gallery_347158_15661_7154793.png

 

Nothing spectacular there. Now, let's move on to a 90" dark auto-stretched:

 

med_gallery_347158_15661_2485774.png

 

Interesting that the starburst/supernova (purposefully NOT calling it amp glow) pattern is so prominent here, whereas it's not even visible in the 90" light. Next, here's the result of running the WBPP script on the entire set of Luminance data (lights, darks, flats, dark flats):

 

med_gallery_347158_15661_2462764.png

 

Ummmm... what? OK, I can kind of understand the top and bottom right not calibrating out. After all, that doesn't really appear to be vignetting and falloff - which was the reason I opened this topic. However, what is the rest of this? Where did that strange pattern come from? Finally, here's what I get if I generate a new image from running the pixel math expression (Light - Dark) / (Flat - DarkFlat):

 

med_gallery_347158_15661_1634313.png

 

I can see the issue in the top right and bottom right, and it looks like the weird gradient pattern. I would expect that to be more pronounced after registration and integration. However, I still cannot wrap my brain around the result I actually got.

 

I have made each of the RAW files shown in the screenshots above available in my Google drive.

 

I appreciate any and all feedback. Thanks in advance!



#4 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 02:03 PM

Jonny,

 

Sorry to have missed this yesterday.

 

1. Take longer flats with this camera. Try to go above 0.2s. Even longer if you can. Dim you light source. Add layers of printer paper between the light source and the scope and take longer flats. See if it helps you.

The flats should calibrate your lights perfectly. I have seen no issues besides when taking very short flats. There are some threads about this in the experienced imaging forum if you are interested.

 

2. The darks look fine. It's stretched so obviously you're going to see the amp glow a lot more than in your lights with 90". It's normal.

 

3. It looks like you have some off axis alignment here. That is maybe you camera/filter wheel, is not sitting exactly on the optical axis and there is some uneven vignetting going on. Are you using some adapters somewhere? How is the camera/filter wheel attached to your scope?

 

I'll download your files now and have a look.

 

*** is the flat that you uploaded is a master flat? After dark flat calibration?


  • dswtan, limeyx, Mike in Rancho and 1 other like this

#5 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 02:15 PM

Jonny,

 

Sorry to have missed this yesterday.

 

1. Take longer flats with this camera. Try to go above 0.2s. Even longer if you can. Dim you light source. Add layers of printer paper between the light source and the scope and take longer flats. See if it helps you.

The flats should calibrate your lights perfectly. I have seen no issues besides when taking very short flats. There are some threads about this in the experienced imaging forum if you are interested.

 

2. The darks look fine. It's stretched so obviously you're going to see the amp glow a lot more than in your lights with 90". It's normal.

 

3. It looks like you have some off axis alignment here. That is maybe you camera/filter wheel, is not sitting exactly on the optical axis and there is some uneven vignetting going on. Are you using some adapters somewhere? How is the camera/filter wheel attached to your scope?

 

I'll download your files now and have a look.

 

*** is the flat that you uploaded is a master flat? After dark flat calibration?

Thanks, Eyal. I appreciate you taking a look.

 

I forgot about this camera needing longer flats. I'll do that next session. I'm using an iPad as the light source, and I can easily dim it further than it is now. Also, I can add more diffusion (paper, another tee-shirt, etc).

 

I should have mentioned my imaging train. It is:

 

294MM -> EFW -> T2-to-T2 adapter -> 11mm ring -> 16.5mm ring -> WO 6AIII (set to ~6.1mm adjustment to get me the 62.1mm total back focus distance) -> GT81

 

All files I uploaded are raw, straight from the camera. Single subs.



#6 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 02:18 PM

So, I'm not sure exactly what are you doing over there with WBPP but I managed to calibrate your lights with just a bit of residual.

It does not calibrate perfectly and I think maybe I know why. It might be that in the time that passed between capturing your light and your flats, there was some movement in your optical train hence the bit of mismatch. I'm only guessing here. But I do think you should take longer flats first to see how does that change.



#7 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 02:21 PM

Thanks, Eyal. I appreciate you taking a look.

 

I forgot about this camera needing longer flats. I'll do that next session. I'm using an iPad as the light source, and I can easily dim it further than it is now. Also, I can add more diffusion (paper, another tee-shirt, etc).

 

I should have mentioned my imaging train. It is:

 

294MM -> EFW -> T2-to-T2 adapter -> 11mm ring -> 16.5mm ring -> WO 6AIII (set to ~6.1mm adjustment to get me the 62.1mm total back focus distance) -> GT81

 

All files I uploaded are raw, straight from the camera. Single subs.

I think if you can create a master dark flat, calibrate and integrate your flats and make a master and also a master dark and upload all, it will be maybe easier to pin point why aren't you able to calibrate properly. I can help you with the manual process I do so we can be consistant. If you want that then PM me.

 

The cause of vignetting might be coming from your filter wheel or maybe something else. To know for sure, take out the filter wheel and shoot a sub without filters. Just hook up the camera and put a 16mm extension to compensate for the missing FW to reach the backfocus.

If the vignetting is the same then it is not the FW.


  • jonnybravo0311 likes this

#8 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 02:27 PM

Great idea. That was going to be my next diagnosis step... getting rid of the filter wheel and just connecting the camera directly to the scope.

 

Also, I do usually manually calibrate, register and integrate. I used WBPP here because... well, I was freaking tired after working all night on a production issue and not getting to sleep until the skies were starting to brighten with dawn. It's entirely possible I botched something using the scripts... I'll go through and create things manually to see if I get any better results.



#9 Ken Sturrock

Ken Sturrock

    Cardinal Ximenez

  • *****
  • Administrators
  • Posts: 10,066
  • Joined: 26 Sep 2009
  • Loc: Denver, CO

Posted 06 March 2021 - 02:31 PM

Like Eyal's result, my quick and dirty with just the regular calibration routine produces an image much like yours with the PixMath formula. It was still a stubborn artifact. A default ADB pass still left a mark.



#10 dswtan

dswtan

    Viking 1

  • *****
  • Posts: 572
  • Joined: 29 Oct 2006
  • Loc: Morgan Hill, CA

Posted 06 March 2021 - 02:35 PM

BTW, on the corner vignetting specifically, if I'm understanding your imaging train and calculating properly, you seem to be at <f/5 and at the limits. Just mechanical tolerances may be against you here.

 

"The distance between the filter and sensor is around 10mm when you connect it to [a 4/3 camera sensor]. So 1.25″ filters won’t have vignetting up to F5 focal ratio scope and 31mm filter won’t have vignetting up to F2 lens."

 

Ref (and the thread):
https://www.cloudyni...mera/?p=7913315


  • jonnybravo0311 likes this

#11 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 627
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 06 March 2021 - 02:39 PM

Jonny,

 

There are some threads about this in the experienced imaging forum if you are interested.

See?  Told ya.

 

Anyway, if you use paper for more diffusion check that it is as flat and clean as you can get it.  Mine ended up having grain.  I know you're not supposed to see anything so far out of focus, but I swear sometimes I can see that grain lol.  So I now take a few more flats, shift the paper over say every 5, and then even rotate it: up, across, and even opposing 45 degree diagonally.


  • jonnybravo0311 likes this

#12 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 02:48 PM

I use this camera with 1.25" filters and this FW in a F/4.3 system and my lights and flats are evenly illuminated and the vignetting is not coming from the FW but from the focuser. Moreover, I also have an OAG which does not interfere in the light path. There is something off axis here.
 

BTW, on the corner vignetting specifically, if I'm understanding your imaging train and calculating properly, you seem to be at <f/5 and at the limits. Just mechanical tolerances may be against you here.
 
"The distance between the filter and sensor is around 10mm when you connect it to [a 4/3 camera sensor]. So 1.25″ filters won’t have vignetting up to F5 focal ratio scope and 31mm filter won’t have vignetting up to F2 lens."
 
Ref (and the thread):
https://www.cloudyni...mera/?p=7913315



#13 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 05:50 PM

Just following up on this. I've now tested a number of different scenarios both with and without the reducer and with/without the EFW.

  1. 294MM -> EFW -> t2-t2 -> 11mm -> 16.5mm -> 6AIII -> GT81
  2. 294MM -> 11mm -> 20mm -> 16.5mm -> 6AIII -> GT81
  3. 294MM -> EFW -> t2-t2 -> 11mm -> 16.5mm -> GT81
  4. 294MM -> 11mm -> 20mm -> 16.5mm -> GT81

In all cases, when I've got the EFW on, I get that cut off on the right side. When I remove the EFW and just shoot with the camera, no issue. So, the problem is definitely with the EFW as I originally suspected. Does anyone have a clue how to tell the EFW to move a couple more clicks once it thinks it's at the correct position?



#14 RogerM

RogerM

    Ranger 4

  • *****
  • Posts: 351
  • Joined: 13 Aug 2020
  • Loc: California

Posted 06 March 2021 - 06:04 PM

I've gone so far as to perform a manual test of my EFW, unmounted, to check on it's positioning performance.  In multiple tests I've observed it to maintain proper centering of each of the eight filter positions.  ZWO has a recalibration feature in their ASCOM driver that you can try running.  There's also a built in recalibration routine in the ASIAir app as well.



#15 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 06:37 PM

I've done the calibration multiple times. I've done it with the "unidirectional" setting and without. I've done it attached to the imaging train and detached. I've done it on two different computers. Every time I take an image with the EFW attached to the train, I get the effect shown in earlier posts. I'm wondering if I somehow botched threading the filters themselves into their positions. I mean, I didn't think that was even a possibility, but at this point, I'm ready to believe in gremlins :).



#16 RogerM

RogerM

    Ranger 4

  • *****
  • Posts: 351
  • Joined: 13 Aug 2020
  • Loc: California

Posted 06 March 2021 - 06:49 PM

I've done the calibration multiple times. I've done it with the "unidirectional" setting and without. I've done it attached to the imaging train and detached. I've done it on two different computers. Every time I take an image with the EFW attached to the train, I get the effect shown in earlier posts. I'm wondering if I somehow botched threading the filters themselves into their positions. I mean, I didn't think that was even a possibility, but at this point, I'm ready to believe in gremlins smile.gif.

Welcome to the next level of insanity...lol.gif

 

You brought up an interesting thought, about filter thicknesses (mounted?) and the internal clearance in the EFW.  There is only one correct side in which to mount the filters to and I'm pretty sure you know which side that is.  On another thread was discussed about someone finding some residual tape (the EFW was purchased second-hand) so perhaps opening the EFW up and checking the wheel's sensor targets is in order...maybe it's a tiny dust gremlin?



#17 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 08:03 PM

Jonny, open the FW up and rotate it via the computer and see if there is something going on with the rotation mechanism. Worst comes to worst you get a new FW. But you're not there yet. At least it's not something in the other parts of optical train, FW is the easiest problem to have.

#18 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 08:08 PM

Insanity it is. I've since pulled apart my entire imaging train. I took the filter wheel completely apart (well, as apart as I dared). I inspected everything, but nothing seemed obviously wrong. Of course, I'm not an expert at filter wheel construction, so I'm not sure how much I should trust my evaluation :).

 

I've now put everything back together and started taking test flats again. I _might_ have gotten things a tiny bit better because now the image circle doesn't seem so slewed. However, it really looks like the image circle produced by these filters is too small for the 294MM. Here's a screenshot of NINA after I took a flat using the B filter:

 

med_gallery_347158_15202_1399299.png

 

Notice the bottom left, bottom right and top right.

 

The L filter is still pretty badly skewed... bottom right is cut off pretty significantly.


  • dswtan and Mike in Rancho like this

#19 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 08:11 PM

Jonny, open the FW up and rotate it via the computer and see if there is something going on with the rotation mechanism. Worst comes to worst you get a new FW. But you're not there yet. At least it's not something in the other parts of optical train, FW is the easiest problem to have.

I had the filter wheel completely apart on my desk. Plugged it into a laptop and switched filter positions and then ran the calibration routine. As I wrote in my previous post - everything LOOKS like it works. There's really nothing standing out to me here. I watched the wheel rotate smoothly and when it stopped, it certainly _looked_ centered.



#20 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 08:18 PM

Are you sure you put the FW facing the right side to the camera? I'm working at F/4.3 and my flats are fine with this FW camera combination.

#21 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 08:21 PM

Are you sure you put the FW facing the right side to the camera? I'm working at F/4.3 and my flats are fine with this FW camera combination.

Ummm... if the mount is in home position and I'm standing behind it facing north, when I look at the EFW, the bottom "bump" is facing me. So...

 

EFW%E6%AD%A3%E9%9D%A21-460x460.jpg

 

That is the camera side. If that's the issue, I'm going to laugh about it pretty long and hard.


Edited by jonnybravo0311, 06 March 2021 - 08:22 PM.


#22 imtl

imtl

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 06 March 2021 - 08:25 PM


I would not worry about how NINA is showing your flats. I would try and calibrate and see what I get. The fact that you are getting different vignetting with different filter position to me looks like uneven rotation of the FW. Hence malfunction in the mechanics.

Another thing to try, which might be stupid. Take out all the filters and image with no filter but FW in the image train. See how the vignetting looks like
  • jonnybravo0311 likes this

#23 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 627
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 06 March 2021 - 08:31 PM

It really does look like an image circle, now that you have one centered.

 

Any way to slap your DSLR on the end of things, to see if it proves out to be the image circle using a bigger sensor?  Or is all the spacing too unique to an astrocam?


  • jonnybravo0311 likes this

#24 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 08:46 PM

I would not worry about how NINA is showing your flats. I would try and calibrate and see what I get. The fact that you are getting different vignetting with different filter position to me looks like uneven rotation of the FW. Hence malfunction in the mechanics.

Another thing to try, which might be stupid. Take out all the filters and image with no filter but FW in the image train. See how the vignetting looks like

I'll give that a try.

 

 

It really does look like an image circle, now that you have one centered.

 

Any way to slap your DSLR on the end of things, to see if it proves out to be the image circle using a bigger sensor?  Or is all the spacing too unique to an astrocam?

Unfortunately my DSLR also has a 4/3 sensor. So, even if I _could_ somehow find a way to attach... there would be a significant difference in the distance between filter and sensor... with the 294MM, it's ~10mm. With my G9 it would be ~25mm.



#25 jonnybravo0311

jonnybravo0311

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,533
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 06 March 2021 - 09:42 PM

I created an album in my gallery with single flat subs for each and every filter.

 

flat L
Album: Fun with Flats
26 images
0 comments

 

In case anyone is interested.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics