arrghh newton rings! why flats! whyyyy!

[Kitsunegari]

So my flats effectively are not functioning properly in any program i use.  What could i be doing wrong?

The flats are 10,000 frames , the image is 3000.

 

 

test2.tif 0001.bmp

Edited June 12, 2020 by Kitsunegari

[Star101]

flats are measured in ADU's else by taking a Flat and watching the spike on the Histogram is only 1/3 from the left. i.e ].....iIi..........[   If that makes sense!

 

[Kitsunegari]

1 hour ago, Star101 said:

flats are measured in ADU's else by taking a Flat and watching the spike on the Histogram is only 1/3 from the left. i.e ].....iIi..........[   If that makes sense!

 

well i will have to try making sense of it, or else i need to fork over some money to the camera tilter.;   these software issues are pretty annoying.  Now i am getting errors in autostakkert thats cutting off half my image!

 

I have some great data and autostakkert is being stubborn with it.

 

 

 

[Star101]

I dont think its flats causing the issue. 

Have you tried without flats?

 

[Kitsunegari]

52 minutes ago, Star101 said:

I dont think its flats causing the issue. 

Have you tried without flats?

 

yea,   the newton rings are throwing back alot of errors during stacking with autostakkert.    In my full disc shots the limb is totally void like it cannot align.  In the above sample you can see the lower right of the image lost all its detail as well.

 

My 2x barlow has less newton rings;  but they are still slightly visible.   i used registax on this barlowed shot because i didnt want to waste another couple hours on AS3! again.   The flat does no good in registax either.

 

I wish i could use registax for everything here, but there is no automation process like AS.       There is a batch setting, but i dont think its any good for stacking stuff.   

 

My main focus is time-lapsing, and now im losing hours to errors on what is otherwise very detailed data.   I guess ill just take the suckers way out and get the camera tilter.

 

 

 

 

[Ibbo!]

Which camera are you using - some are more prone to NR's than others.

Try cutting back to 20 to 50 frames for flats I have very mixed results using flats.

Cut down the AVI or SER capture to 30 to 45 secs and stack less frames., I use between 51 and 350 for a stack.

For time lapse I have done 10/15 secs capture every 1-2 min using a stack of 50 frames in the past but it did involve some tweaking to line up to produce the final result.

I have a tilter in place for NR but remember every tweak you make tends to alter something else.

[Kitsunegari]

for these particular captures i was using the IMX287 basler camera,   

 

The .ser files were set to 10,000 frame captures.  7GB each file.  captures were set to 10 second intervals.

attached is the data.txt for the capture.      

33 minutes ago, Ibbo! said:

Which camera are you using - some are more prone to NR's than others.

Try cutting back to 20 to 50 frames for flats I have very mixed results using flats.

Cut down the AVI or SER capture to 30 to 45 secs and stack less frames., I use between 51 and 350 for a stack.

For time lapse I have done 10/15 secs capture every 1-2 min using a stack of 50 frames in the past but it did involve some tweaking to line up to produce the final result.

I have a tilter in place for NR but remember every tweak you make tends to alter something else.

Thanks for the advice Ibbo:  i will attempt using capturing a cascade of flats next session, at 10 frame increments up to 1000.

 

Mars_161133.txt

[Macavity]

How are you doing the Flats? For PART disks, I use two layers of "crossed" cling film...
slew to the centre of the solar disk and adjust the gain to give a curve centred at 50%.
Newtons rings usually show up most (for me) when I use a Barlow Lens for part disks.
Fortunately my full disks don't need flats... 'cos it's a LOT more fiddly to do them! 😛

Flats in AS!3 work well for me in removing the "blobs" due to dust on the chip etc.
But I have to admit I never had too much success removing Newtons Rings via flats.
Eventually, I had to "experiment with" (buy sadly! lol) various Cameras, until finding
ones that didn't produce NRs in the first place --  CCD chips rather than CMOS etc.
An old-style style DMK41, before "graduating" to my ICX445 based "Chameleon 3".

I also found that time lapse disks (I sense that is what you are attempting) are very
prone to showing even the smallest residual Newton's Rings - Too faint to see in 
conventional (static) DISK images. NR's are no problem for prominences, but...  🤔

Edited June 13, 2020 by Macavity

[Kitsunegari]

So here is a retake using Regstax,  the details that were destroyed with autostakkert have been preserved;  but the newton rings are still prominent.      I think the autostakkert detail around the active region looks more presentable however, so there is clearly some discrepancy here on how data is handled between the two apps.

 

Guess i have some sleuthing to do on this because i have a great setup dying to be remedied.

[Nigella Bryant]

I've never taken flats. I use the rowan astronomy tilt plate, easy to use and set up. Alway's keep it on the scope. 

[Macavity]

Hmmm... As the exposure time is increased, the Sun "drifts around" on the
chip (the mount tracking in not perfect!) and the Newtons rings "follow" it?
So the image you produce (as a result of stacking/animation) contains fixed
solar features (sunspots etc.) and now the Newtons rings drift around?!? 😛

A bit like the effect I am getting here...

This an animation of the disk over a couple of hours (Note drifting NRs!):

This is a single frame (The newtons rings are not really noticable): 🙂



The human eye/brain is clearly attracted more by Movement?!?! 😛

I suspect you'll either need a Camera Chip that doesn't give NRs...
OR some method of reducing them at source... like a "Tilter" etc. 🤔

[Zakalwe]

Flats are very hard to get right on partial discs. Do you have a lot of drift in the avi files?

 

[Zakalwe]

Also, cut down the number of frames. 3000 is way too much... Don't forget that the surface of the Sun is incredibly dynamic. Spicules and proms are moving at many thousands of miles per hour and a too-long avi file can blur them.

[Zakalwe]

On 13/06/2020 at 02:06, Kitsunegari said:

I have some great data and autostakkert is being stubborn with it.

 

The data can't be that great as AS!3 would stack it. You need to solve the NRs to get good data...they can't be solved afterwards (well, they sorta can be solved with artificial flats, but it is far better to solve them before stacking.)

 

 

On 13/06/2020 at 03:09, Kitsunegari said:

I wish i could use registax for everything here, but there is no automation process like AS.       There is a batch setting, but i dont think its any good for stacking stuff. 

Believe me, you don't.  Registax is ancient technology (last updated 8 years ago). It is slow, clunky, awkward to use and uses outdated algorithms (for colour  stacking). Autostakkert is leagues ahead of it.

 

On 13/06/2020 at 03:09, Kitsunegari said:

My main focus is time-lapsing, and now im losing hours to errors on what is otherwise very detailed data.   I guess ill just take the suckers way out and get the camera tilter.

You're trying to run before you can walk. Fix the NRs and be confident that you can acquire, stack and process single images first, before you waste hours trying to acquire, stack and process tens of images to make animations.

Edited June 16, 2020 by Zakalwe

[Freddie]

1 hour ago, Zakalwe said:

Also, cut down the number of frames. 3000 is way too much... Don't forget that the surface of the Sun is incredibly dynamic. Spicules and proms are moving at many thousands of miles per hour and a too-long avi file can blur them.

I don’t think the length of AVI is an issue as a 10,000 frame AVI would be captured in 34 secs with the parameters used. You have to consider image scale and at this image scale, movement over that time isn’t going to be an issue. I wouldn’t however stack anywhere near 3000 of those 10,000 frames though.

[Zakalwe]

On 13/06/2020 at 04:12, Macavity said:

How are you doing the Flats? For PART disks, I use two layers of "crossed" cling film...
slew to the centre of the solar disk and adjust the gain to give a curve centred at 50%.
Newtons rings usually show up most (for me) when I use a Barlow Lens for part disks.

Flats work great when the Sun fills the FoV. Well worth doing.

Televue PowerMates are a pain for generating NRs too.

 

On 13/06/2020 at 04:12, Macavity said:

Fortunately my full disks don't need flats... 'cos it's a LOT more fiddly to do them! 😛

There's the problem with trying to create flats when the Sun doesn't fill the chip. Even with cling film you do not get even illumination across the FoV. The flats then under correct in parts of the image. Personally I never use flats unless I'm using the Quark or barlowing the Lunt 60mm for this very reason.

[Merlin66]

I use TV PM’s for most of my solar imaging and yes I experience NR, but equally I find the same problem of NR with various Barlow’s. IMHO I think it’s the higher f ratio that aggravates the generation of NR.

The basic cause is the design of the camera being use. The interference between the built-in cover plate and the underlying silicon based chip.

 

[Zakalwe]

9 minutes ago, Merlin66 said:

I use TV PM’s for most of my solar imaging and yes I experience NR, but equally I find the same problem of NR with various Barlow’s. IMHO I think it’s the higher f ratio that aggravates the generation of NR.

The basic cause is the design of the camera being use. The interference between the built-in cover plate and the underlying silicon based chip.

 

Yep....NRs can appear on one camera and not on the other.

The worst NRs that I have ever experienced was when I mounted an ASI1600 on my Lunt L60.

[Rusted]

Newton's rings are better seen with near parallel light normal to the surface being tested.
Monochromatic light is required to emphasize the rings. 
That is precisely how large optical flats are/were once tested in a simple partially reflective rig.
H-a and a stack of plane-parallel surfaces are a perfect scenario for creating NR.

[Macavity]

34 minutes ago, Zakalwe said:

There's the problem with trying to create flats when the Sun doesn't fill the chip. Even with cling film you do not get even illumination across the FoV. The flats then under correct in parts of the image.

Good stuff! 😎 My DMK camera gave FEW NR's, but then had a "slightly dodgy" chip. 😕
I did use flats to remove some uneveness in the pixels.

Using Flats for full disk resulted in images that wer (you guessed) un-attractively FLAT! 🥳
I found a website that generated "real life" solar disks from algorithms that I could add...
But ultimately, it seems better to try to reduce them "at source" (as you/others suggest).

The thread did get me thinking: If we get some nice BIG sunspots (I'm still alive!) I might
try to "drift out" the residual faint NRs I had on the above animation... But who knows? 🤔
(I had some success altering tracking rates... drifting images across "dodgy" chips etc.)

I suspect I should buy a spare "Chameleon Cam" --- With my "lockdown savings"... 🙄🤣

Edited June 17, 2020 by Macavity

[Zakalwe]

28 minutes ago, Macavity said:

Good stuff! 😎 My DMK camera gave FEW NR's, but then had a "slightly dodgy" chip. 😕
I did use flats to remove some uneveness in the pixels.

Using Flats for full disk resulted in images that wer (you guessed) un-attractively FLAT! 🥳
I found a website that generated "real life" solar disks from algorithms that I could add...
But ultimately, it seems better to try to reduce them "at source" (as you/others suggest).

The thread did get me thinking: If we get some nice BIG sunspots (I'm still alive!) I might
try to "drift out" the residual faint NRs I had on the above animation... But who knows? 🤔
(I had some success altering tracking rates... drifting images across "dodgy" chips etc.)

I suspect I should buy a spare "Chameleon Cam" --- With my "lockdown savings"... 🙄🤣

CCD cameras like the old DMKs were much less prone to generating NRs. Mind you, they also had tiny sensors with not a lot of real estate to be affected.

Drifting can work as long as the rings aren't very strong. It's a waste of sensor real-estate though. And in some scenarios, drift is the last thing that you want to happe. For example, I can just about squeeze a full disk onto the 178 chip when using my Lunt 60mm. Any drift whatsover and I lose part of the image which means back to mosaics. I bought the camera to specifically remove the need for mosaics!

The best outcome is to get rid of them at source. Tilt adjusters are a powerful tool in kitbag for achieving just this. I have my adjuster at the minimum needed tilt as tilt can introduce an uneven focal plane which can show up on big sensors. Any faint residual rings can be removed with an artificial flat-field

https://photographingspace.com/tag/artificial-flats/

[Pete Presland]

I certainly hardly ever see them on my DMK21au618 camera, i have a recently purchased a tilt adapter for when i start using a ZWO camera for Solar imgaing.

[inFINNity Deck]

On 17/06/2020 at 11:08, Zakalwe said:

The best outcome is to get rid of them at source. Tilt adjusters are a powerful tool in kitbag for achieving just this. I have my adjuster at the minimum needed tilt as tilt can introduce an uneven focal plane which can show up on big sensors.

Hi Stephen,

when using a tilt adjuster, what process do you use to adjust them? I understand that in static images the NRs are hard to see, so what is the easiest and best way (presumably these result in two answers)? I noticed that my last animation also show faint NRs, so could do with improvement, although an artificial flat would do.

Nicolàs

PS: Inspired by our last conversation, I created a second solar imaging tutorial on a Dutch forum, of course referring to you (should translate nicely in Chrome).

[Zakalwe]

2 hours ago, inFINNity Deck said:

Hi Stephen,

when using a tilt adjuster, what process do you use to adjust them? I understand that in static images the NRs are hard to see, so what is the easiest and best way (presumably these result in two answers)? I noticed that my last animation also show faint NRs, so could do with improvement, although an artificial flat would do.

Nicolàs

 

It's just trial and error, to be honest. One thing that I do is to put the axis of tilt along the long axis of the sensor. This keeps the difference in focal plane across the chip to the minimum.

 

 

2 hours ago, inFINNity Deck said:

PS: Inspired by our last conversation, I created a second solar imaging tutorial on a Dutch forum, of course referring to you (should translate nicely in Chrome).

Cool!  I'll have a look later.

[vineyard]

On 13/06/2020 at 08:29, Nigella Bryant said:

I've never taken flats. I use the rowan astronomy tilt plate, easy to use and set up. Alway's keep it on the scope. 

Hi Nigella, may I ask whether you keep this tilter between the diagonal & the focus draw tube (so that you can still use an EP) or after the diagonal?

I've just gotten a Solar Quest.  What a brilliant idea - makes set-up & tracking so much easier.  But it seems to do its job too well b/c since the Sun doesn't move in the image, I'm getting NR even without a 2.5x PM now!

I'm hoping that a T2 tilt adaptor could be permanently setup between the diagonal & the focuser (esp as I use an ASI120mm mini and the 1.25 nose on that would make connecting it directly to a tilt adaptor slightly problematic.

Thank you,

Vin