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

Hydrogen alpha flats - Need help

  • Please log in to reply
17 replies to this topic

#1 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 18 June 2019 - 05:49 PM

I decided a few weeks ago to try making flats, but I ran into an issue.  I can't get my camera out of focus enough to generate a flat through defocus, and everything I tried using to blur the incoming light, be it plastic films and bags or what have you, turned out to completely block the hydrogen alpha light. 

Short of throwing a piece of plate glass at some #600, does anyone have suggestions?



#2 rigel123

rigel123

    Hubble

  • *****
  • Posts: 16501
  • Joined: 29 Jun 2009
  • Loc: SW Ohio

Posted 18 June 2019 - 08:00 PM

If your camera is attached by a nosepiece did you try pulling it out a bit after racking the focuser back?  Or are you trying to get a flat with the full disk in the FOV which I find just doesn’t work well for flats.  I've only found flats to work well with closeups of the surface.



#3 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 18 June 2019 - 08:56 PM

I was hoping to defocus until the disk was in the full field, but that requires a hell of a lot of travel outside of focus, thus my attempting to blur the image prior to the objective.  I guess I can get a barlow and do a closeup... gives me another reason to get a barlow I suppose.



#4 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9003
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 19 June 2019 - 12:14 AM

Why are you wanting to do flats?



#5 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 19 June 2019 - 12:52 AM

For science. 

No, I want to compare the flat-corrected images with the software-corrected images to deal with my camera's issues.



#6 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9003
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 19 June 2019 - 08:25 AM

What's wrong with your camera?



#7 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 19 June 2019 - 11:57 AM

Short answer: it uses a Sony EXMOR sensor.

Less short answer: certain sensors (IMX174, IMX178) have poor performance in monochromatic light at long wavelengths and it induces a noticeable grid pattern.

I don't really want to poop all over other forum users, that is not my intent here, but I want to show you an example.  I don't have any readily available images since i've started using software to help mitigate the impact, but once you know what to look for, it is easy to spot.  Take a look at the images here: 

You'll notice horizontal lines and blocky structures on some of the closeups.  These are not jpeg artifacts, these are indeed visible during the live preview through SharpCap/FireCapture if you zoom in sufficiently, as you can verify from the image in the second link.  Some people seem to have this issue and others do not, but I've watched carefully and even people claiming to not have the problem, do indeed have the problem periodically from what I can see.  The issue tends to creep up during very short exposures, so single-stacked setups are typically more susceptible than double-stacked.


Edited by jwestervelt, 19 June 2019 - 12:31 PM.


#8 Great Attractor

Great Attractor

    Mariner 2

  • -----
  • Posts: 262
  • Joined: 23 Oct 2012
  • Loc: Poland

Posted 19 June 2019 - 01:36 PM


everything I tried using to blur the incoming light, be it plastic films and bags or what have you, turned out to completely block the hydrogen alpha light

 

I use a semitransparent plastic shopping bag. It does attenuate the light significantly, but there's still enough after I crank the exposure and gain to maximum (additional noise due to that doesn't matter for making a flat anyway); without the bag I need only 0.7 ms exposure in the disc center (no gain).

 

Though I need it only to counter vignetting; not sure if increasing the gain like this wouldn't change your sensor's "fixed-pattern noise" and invalidate the flat.



#9 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 19 June 2019 - 02:13 PM

changing the gain doesn't remove the fixed-pattern noise, it just makes it harder to discern through the random pattern noise from the gain.  The fixed pattern is still there, and will come back after stacking eliminates the random noise.

I tried several semi-transparent bags that were available to me, and all of them appeared to completely attenuate the hydrogen alpha line.  Ideally I'd keep the gain very low for this, and exposure times over 6ms appear to begin eliminating the pattern.   At around 12ms the pattern seems to be about half as prominent, and above 16ms it appears to be absent entirely.  I can probably de-tune the etalon to get more light from the photosphere in, but I'm thinking rough-polished glass flats are going to be the way to go.  I'll try a barlow first though since that idea would likely be sufficient.



#10 Great Attractor

Great Attractor

    Mariner 2

  • -----
  • Posts: 262
  • Joined: 23 Oct 2012
  • Loc: Poland

Posted 19 June 2019 - 02:23 PM

If it's a matter of simply increasing the exposure time, I suggest adding a 1.25" neutral density filter (e.g. Baader; they come in a few different transmission variants). 16 ms doesn't sound too bad (at most, I've used about 11-15 ms for either white light at f/15 with Solar Continuum filter, or in H-alpha for weak prominences).



#11 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9003
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 19 June 2019 - 03:00 PM

The fixed noise pattern issue keeps doing the rounds.

 

I'm not sure i've seen it anywhere with the 174 chip.  The 178 just seems to be a duffer camera for solar, personally I would sell it and get another sensor...  My gut feeling is that the rolling shutter plays a part.

 

An easy solution with this camera would be to rotate it 90 deg relative to the norm, so chip long axis is vertical, and then to allow some drift whilst taking exposure.  Stacking enough frames should cancel out the fixed pattern noise.



#12 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 19 June 2019 - 03:37 PM

174s have the issue as well, as do the 183s.  https://bbs.astronom...opic.php?t=8012  It seems to affect most, if not all, of the backlit exmor sensors.  

Allowing for drift is the intuitive answer, but I have played with this and found that stacking software tends to lock onto the fixed noise pattern during alignment since it is a stronger signal.  The end result is that the other details (filaments, spicules, etc) are softened a bit because they are not registered very well.  Software like autostakkert will align against the grid pattern, causing the other features in the AP to shift around a bit.  I have confirmed this by using the same .SER for input, and running through Autostakkert with the same settings, but removing the grid pattern through software before stacking.  The resulting stacked image prior to grid removal appears to have a soft focus by comparison due to the poor registration.

I realize that a different camera is probably the best bet, but which?  And if the problem can be solved in software, why not do so?  I'm only trying to get a clean flat so I can properly calibrate my software and validate the results.
 



#13 ASTROTRUCK

ASTROTRUCK

    Viking 1

  • -----
  • Posts: 582
  • Joined: 02 Jan 2010

Posted 20 June 2019 - 10:21 AM

question using drift method - if i am imaging disk 250  images fraction of a second each  - do i shut off the tracking before beginning exposing all the 250 shots so all are drifting or do only say 100 images drifting then 150 tracking then processing all together as 250  ?

 

i would expect all the images should be taken drifting with the tracking off then processing so that somehow the issues are removed - i have a very strong sweet spot almost centred so the rest looks in shadow. - should drifting remove this  & if so i assume  all my solar images should be drifting other than for prominences being a bright image anyways 

 

thanks



#14 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 20 June 2019 - 11:12 AM

drifting could help with the sweet spot, but I haven't tested it to verify.  It might make the issue worse so I'd have to let someone else with experience have a crack at that.  The way I see it, if you have a sweet spot that covers 1/3 of the sun's disk, then for any given point on the sun's image, 2/3 of your frames are going to contain a non-sweetspot version and these may cause the better data in the sweetspot to be removed.



#15 DesertRat

DesertRat

    Fly Me to the Moon

  • *****
  • Posts: 6266
  • Joined: 18 Jun 2006
  • Loc: Valley of the Sun

Posted 20 June 2019 - 01:57 PM

Fixed pattern noise (FPN) in cmos cameras is largely ignored by planetary imagers since it is well corrected for by atmospheric dithering (and imperfect tracking) that occurs at long focal lengths.  For prime focus imaging of the entire solar disc it would be more problematic.

 

I am a little surprised that FPN for these cameras has any well defined grid pattern.  There was a conjecture early on that for some of these mono cmos cameras a bayer artifact was a remnant due to all the chips being made had color filter arrays which were removed in the end for the less popular mono devices.  Seems strange but if any remaining bayer patterns were left there would be a more well defined pattern especially in a halpha image.  Normally FPN patterns are more random in nature.  I don't have one of these cameras (174/178/183) so I cannot verify.

 

I am also surprised that a stacking program could lock onto FPN, though I admit the possibility exists.  A good flat should involve no alignment, no tracking, in order to get a good record of FPN.   Some solar imagers use a lower gamma (or more contrasty setting) for surface images in order to more easily track halpha features.  And a higher gamma for proms.  This is not the recommended linear imaging for scientific purposes but we are only looking at getting a nice image.  If the alignment boxes are too small that may be contributing to what you and others are seeing.

 

What one could do is take a camera lens fitted with a red filter and take a flat using a elp panel or similar flat source.  Red filter because FPN does have a wavelength dependency and a W25 filter might get you close enough.  Again a simple addition of all frames with no alignment is important.  We are assuming here that vignetting, dust spots etc are not the primary concern.

 

A defocused solar disc will not work since it will be larger than the focused image and include surrounding black space. 

 

Never saw a well defined sweet spot as small as reported above.

 

Interesting topic, thanks for posting your investigation.

 

Glenn


  • marktownley and LFORLEESEE like this

#16 jwestervelt

jwestervelt

    Viking 1

  • -----
  • topic starter
  • Posts: 554
  • Joined: 01 Sep 2012
  • Loc: Phoenix, Arizona

Posted 20 June 2019 - 02:44 PM

I will try a red filter and see if I can replicate this.  It would be a simple matter of pointing the camera at the open sky (without a scope attached) if that is sufficient enough to induce the grid pattern. My understanding it that it primarily happens on longwavelength and sufficiently monochromatic light, but I have heard of people running into this with lunar imaging through a red filter, so it is worth a shot. Thanks!



#17 ASTROTRUCK

ASTROTRUCK

    Viking 1

  • -----
  • Posts: 582
  • Joined: 02 Jan 2010

Posted 21 June 2019 - 08:42 AM

regarding drifting anyone tried it ??- should i stop tracking which might make the sun image speed too fast in which case track at stellar or lunar speed ?

 

i could try all three but most days are cloudy here in toronto so would like to get advice so i could image knowing which is best rather than experimenting & kill a whole day - then clouds for a month or more -

 

maybe better to take up computerized tiddly-winks good on cloudy days  



#18 rigel123

rigel123

    Hubble

  • *****
  • Posts: 16501
  • Joined: 29 Jun 2009
  • Loc: SW Ohio

Posted 21 June 2019 - 09:11 AM

regarding drifting anyone tried it ??- should i stop tracking which might make the sun image speed too fast in which case track at stellar or lunar speed ?

 

i could try all three but most days are cloudy here in toronto so would like to get advice so i could image knowing which is best rather than experimenting & kill a whole day - then clouds for a month or more -

 

maybe better to take up computerized tiddly-winks good on cloudy days  

Not sure on the drift but I took up nature photography and cloudy days are OK for that!


  • jwestervelt likes this


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