Back to forums

What's wrong with my filters?

Equipment Gradient Removal Calibration
22 replies•618 views
Are my filters worse than you'd expect?
Multiple choice poll 9 votes
33% (3 votes)
22% (2 votes)
0% (0 votes)
44% (4 votes)
You must be logged in to vote in this poll.
bobtobb avatar
bobtobb avatar
Hi!

I'm getting some immense gradients in my final stacked images, which "should not happen" since I have proper Flats. The filters are in an EFW and they look clean (basically placed directly in the EFW straight from purchase). They don't have fingerprints. It's 36mm filters on an APS-C camera.

I have ZWO RGB, and Chroma SHO - the problem is particularly bad on the RGB, and quite severe in the integrated G.

Attached integrated RGB + single frames out of all Flats, all stretched with STF. Anyone experienced this before? Wouldn't you expect all filters to have a similar gradient (from the OTA) but different spots (from dust)?

Steps I've tried so far:
* Take out and double-check that all filters look good. Used a blower to remove some dust.
* Triggered EFW calibration in ASIAIR, didn't help.

Next step is to clean them all with alcohol but wanted to reach out before I do that..

Helpful Engaging
Jan Erik Vallestad avatar
Jan Erik Vallestad avatar
The green looks to be cut off in the corner for some reason. But have you checked for light leaks in the whole image train? How do you take the flats?
bobtobb avatar
bobtobb avatar
The green looks to be cut off in the corner for some reason. But have you checked for light leaks in the whole image train? How do you take the flats?

Yeah! The G filter looks horrible (but still looks OK when I visually examine it)! It has a huge effect on quality of my G channel after integration, and I consider G to be mostly broken...  I take all the flats in the beginning of the session, after focusing - so when the rig is roughly polar aligned.. I guess it could be a light leak from a nearby light, but then all flats should have the same leak..
Tony Gondola avatar
Tony Gondola avatar
Some colors will show gradients more than others, this is normal. What does concern me is the falloff in the corner of frame one. Are you using an off axis guider?
bobtobb avatar
bobtobb avatar
Tony Gondola:
Some colors will show gradients more than others, this is normal. What does concern me is the falloff in the corner of frame one. Are you using an off axis guider?

Yes indeed I am - it's the ZWO OAG-L. From the factory it was set so low it cast a shadow on the frame, so I have put it quite high up (to the point that the guidecam's lower 20% are i shadow (which I figured was a better compromize).

Will they show more gradients than other even though all flats are made with a flat-panel? I didn't move the flat panel between filters.
jmdl101 avatar
jmdl101 avatar
To me it looks like your green filter isn't lining up correctly and the edge of the filter is cutting off that corner. I'd say that's more likely over a prism shadow because the rest of your filters look fine. Everything else looks like light leaks.

What scope are you using?
Well Written Concise
bobtobb avatar
bobtobb avatar
Ah, so maybe if the wheel isn't aligning that slot I can try swapping green and red, and see if I now get the error on red..
This is on the ZWO FF80.
jmdl101 avatar
jmdl101 avatar
That should be an easy way to check it. The zwo EFW comes with darkening rings to prevent light leaks, do you have yours installed? There could be  a light leak somewhere else though, an easy way to find it is to put on the lens cap and take a flashlight and shine it around while watching a live view of the camera.
Helpful
bobtobb avatar
bobtobb avatar
Yes I have the darkening rings installed on both ends of the EFW - I'll try the flashlight to validate, thanks. I may very well have some light leak in there - but I'm thinking that's a potential additional problem to the other problem since I get the same gradient every night when I take flats. (I move my rig and set it up every session, so I always start with Flats).
I'll tear it apart tomorrow and clean the filters using alcohol, just to be safe.
Shinpah avatar
Shinpah avatar
There should be a driver setting if this is an alignment issue that can force the filter wheel to only rotate in one direction. That solved positional issues for me with my zwo 7pos 36mm.
Concise
bobtobb avatar
bobtobb avatar
There should be a driver setting if this is an alignment issue that can force the filter wheel to only rotate in one direction. That solved positional issues for me with my zwo 7pos 36mm.

Ah yeah I'll try that, thanks!
Willem Jan Drijfhout avatar
Willem Jan Drijfhout avatar
The Green looks indeed like some kind of mis-alignment. Are you using a ZWO filterwheel? If so, have you tried to re-calibrate it? I've had alignment issues with an EFW in the past and re-calibration solved it for me. I'm using KStars, and in the INDI-panel there is a ReCalibration button you can click. If you use ASCOM, the ASCOM driver also has a recalibration option if I'm not mistaken.
The recalibration process takes a bit longer (like a minute or perhaps a bit longer) and the wheel makes some scary noises, but that's supposed to be ok. An easy and quick option to test if you havent already done so.
Helpful Supportive
bobtobb avatar
bobtobb avatar
Willem Jan Drijfhout:
The Green looks indeed like some kind of mis-alignment. Are you using a ZWO filterwheel? If so, have you tried to re-calibrate it? I've had alignment issues with an EFW in the past and re-calibration solved it for me. I'm using KStars, and in the INDI-panel there is a ReCalibration button you can click. If you use ASCOM, the ASCOM driver also has a recalibration option if I'm not mistaken.
The recalibration process takes a bit longer (like a minute or perhaps a bit longer) and the wheel makes some scary noises, but that's supposed to be ok. An easy and quick option to test if you havent already done so.

Yup  it's the ZWO 36mm 7-slot wheel - I've tried recalibrate and it didn't help (it's available as an option on the ASIAIR too) but I haven't tried setting it to only rotate in one direction as suggested above. It'll be interesting if that helps. Thanks for the suggestion!
Willem Jan Drijfhout avatar
Willem Jan Drijfhout avatar
Willem Jan Drijfhout:
The Green looks indeed like some kind of mis-alignment. Are you using a ZWO filterwheel? If so, have you tried to re-calibrate it? I've had alignment issues with an EFW in the past and re-calibration solved it for me. I'm using KStars, and in the INDI-panel there is a ReCalibration button you can click. If you use ASCOM, the ASCOM driver also has a recalibration option if I'm not mistaken.
The recalibration process takes a bit longer (like a minute or perhaps a bit longer) and the wheel makes some scary noises, but that's supposed to be ok. An easy and quick option to test if you havent already done so.

Yup  it's the ZWO 36mm 7-slot wheel - I've tried recalibrate and it didn't help (it's available as an option on the ASIAIR too) but I haven't tried setting it to only rotate in one direction as suggested above. It'll be interesting if that helps. Thanks for the suggestion!

Yes, uni-directional rotations is worth enabling.
TiffsAndAstro avatar
TiffsAndAstro avatar
The green looks to be cut off in the corner for some reason. But have you checked for light leaks in the whole image train? How do you take the flats?

Yeah! The G filter looks horrible (but still looks OK when I visually examine it)! It has a huge effect on quality of my G channel after integration, and I consider G to be mostly broken...  I take all the flats in the beginning of the session, after focusing - so when the rig is roughly polar aligned.. I guess it could be a light leak from a nearby light, but then all flats should have the same leak..

You're not a asking flats for each filter used?
I've read even parfocal filters benefit for doing that
bobtobb avatar
bobtobb avatar
TiffsAndAstro:
You're not a asking flats for each filter used?
I've read even parfocal filters benefit for doing that

You mean if I take Flats for each filter? Yes I do! In all cases the lightbox remains static. I fire off all Flats at the beginning of each session.
TiffsAndAstro avatar
TiffsAndAstro avatar
TiffsAndAstro:
You're not a asking flats for each filter used?
I've read even parfocal filters benefit for doing that

You mean if I take Flats for each filter? Yes I do! In all cases the lightbox remains static. I fire off all Flats at the beginning of each session.

does your focus change for each filter change for flats?
Oscar avatar
Oscar avatar
Willem Jan Drijfhout:
The Green looks indeed like some kind of mis-alignment.


I agree
bobtobb avatar
bobtobb avatar
TiffsAndAstro:
TiffsAndAstro:
You're not a asking flats for each filter used?
I've read even parfocal filters benefit for doing that

You mean if I take Flats for each filter? Yes I do! In all cases the lightbox remains static. I fire off all Flats at the beginning of each session.

does your focus change for each filter change for flats?

I don't know, I have it set to autofocus after filterchange just to be safe. I have 3 different thicknesses of filters in the wheel (RGB + L + SHO) so I ticked the setting to always refocus just to be safe.
David Fercho avatar
David Fercho avatar
It makes sense to me that the rapid drop-off in the corner of the green filter would suggest the filter is not centered in the light path.  But I'm interested in an explanation of why calibration with the flats isn't correcting the issues at all.  Since the green flat is capturing the same drop-off as the lights, why doesn't calibration with the flat brighten that area of the master light?  Also, the red flat has a pronounced brightening on the right side, which may be what the others here have associated with a light leak.  But, again, the calibration result seems to be the opposite of what is expected.  The red master light is also very bright on the right side.  Shouldn't the bright zone of the red flat reduce the brightness in the same zone of the master light?

I am asking all of this for my own understanding, and with the hope that any answers will also be useful to bobtobb.

CS,
Dave
Well Written Insightful Respectful Engaging
bobtobb avatar
bobtobb avatar
David Fercho:
It makes sense to me that the rapid drop-off in the corner of the green filter would suggest the filter is not centered in the light path.  But I'm interested in an explanation of why calibration with the flats isn't correcting the issues at all.  Since the green flat is capturing the same drop-off as the lights, why doesn't calibration with the flat brighten that area of the master light?  Also, the red flat has a pronounced brightening on the right side, which may be what the others here have associated with a light leak.  But, again, the calibration result seems to be the opposite of what is expected.  The red master light is also very bright on the right side.  Shouldn't the bright zone of the red flat reduce the brightness in the same zone of the master light?

I am asking all of this for my own understanding, and with the hope that any answers will also be useful to bobtobb.

CS,
Dave

Yeah I've been scratching my head about that too. I've re-run WBPP several times to rule out any settings errors and then it hit me! The most logical explanation is simply that the Flats aren't used to compensate the Lights - and that is exactly what is happening! Everything except L+Flats was shot in Bin2, so WBPP doesn't use the flats since they're shot with a different bin. That makes sense, and sorry for the wild goose chase! But the filter issue is still a mystery to me (and it very much affect all sessions). I'll check for light leaks in a bit.
Helpful Insightful Respectful
David Fercho avatar
David Fercho avatar
I always check the calibration tab in WBPP before starting the run.  There should be green check marks in the dark and flat column for each group of lights shown there.  If your calibration frames are incompatible with the lights, or if WBPP doesn't properly match filters, keywords, etc., the green mark will not be there, indicating your lights will not be calibrated. Checking that saves wasted processing time when something's not right.
Well Written Helpful Concise
bobtobb avatar
bobtobb avatar
Thanks all for your suggestions! I wanted to give you an update;

I disassembled the filterwheel to look for any obstructions, but everything looked great, and I could not find anything wrong with slot 3 where the G filter was. Eventually I concluded that the issue is with the filter itself, though it looks perfect with visual inspection (attached!).

I've been thinking about replacing these ZWO RGB filters with Chroma RGB for a while, so that's what I ended up doing; and now the problem is gone.

Well Written Concise