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

Nighttime H-Alpha filter for Solar viewing? Part 2

  • Please log in to reply
27 replies to this topic

#1 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6569
  • Joined: 12 Jun 2004
  • Loc: Washington USA

Posted 21 August 2013 - 12:25 PM

This is a continuation of a previous thread; see Part 1 here.

I recently made a presentation on solar filters, and came up with this visual representation of why you can not use a white light solar filter with a nighttime H alpha filter to see features of the chromosphere such as prominences, filaments, flares, etc.

In pane 1 below, a bright white painted exterior wall represents the intensely bright photosphere, and the window is the broad dark H alpha absorption line in the continuous spectrum of the photosphere. This can be thought of as looking into a dark room with no other windows, in which a candle glows. This candle represents the H alpha emission line from the chromosphere, ~ 100,000 times dimmer than the photosphere.

In pane 2 we use a standard white light filter to dim the photosphere to a safe level for observation. This has required a 100,000 times reduction in brightness, but we have also dimmed the brightness of the emission line (the candle) from the chromosphere 100,000 times as well. Adding a nighttime H alpha filter (pane 3) reduces the brightness of the photosphere a bit by eliminating all but the red wavelenghts around the H alpha absorption line, but cannot do anything to bring back and reveal the H alpha emission (the candle), which has been greatly reduced in brightness by the white light filter.

The only way to see the H alpha emission (the candle) from the chromosphere is to completely eliminate light coming from the photosphere (the wall), while leaving the window un-dimmed. We can then peer into the window (the absorption line), where the chromosphere - the candle - can now be seen in pane 4.

Attached Thumbnails

  • 6038355-Combination SM 2.jpg

  • Doug D., Live_Steam_Mad, shaesavage and 3 others like this

#2 mhilscher

mhilscher

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 13 Jul 2011
  • Loc: Cologne, Germany

Posted 21 August 2013 - 02:28 PM

very nice. thanks for sharing

#3 astrovale

astrovale

    Surveyor 1

  • *****
  • Posts: 1986
  • Joined: 19 Feb 2006
  • Loc: Rimini, Italy

Posted 21 August 2013 - 04:57 PM

Very clear and easy explanation Bob! Thanks for sharing it.

Luca

#4 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9005
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 23 August 2013 - 05:25 PM

That is a really good explanation Bob! :cool:

#5 highfnum

highfnum

    Aurora

  • *****
  • Posts: 4628
  • Joined: 06 Sep 2006
  • Loc: NE USA

Posted 24 August 2013 - 04:47 AM

Cool
I guess thid expla ins hiw prom scope works
It has oculting disk

#6 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6569
  • Joined: 12 Jun 2004
  • Loc: Washington USA

Posted 27 August 2013 - 12:41 PM

Thanks guys, appreciate your taking a look!

Yep Jon, that's it exactly, and why the moon makes such a good occulting disk!

#7 jeffpkamp

jeffpkamp

    Ranger 4

  • -----
  • Posts: 322
  • Joined: 11 Jan 2013

Posted 12 November 2013 - 10:33 PM

I'm a bit late to this party, but was refered to this post from elsewhere.

I realize that your post is on using Ha Filter for visual work on the chromosphere, but has anyone tried it for solar AP? I'm just curious. I'm looking at doing some solar stuff next year with my SCT. If you added an occluder could you do it then?

#8 jeffpkamp

jeffpkamp

    Ranger 4

  • -----
  • Posts: 322
  • Joined: 11 Jan 2013

Posted 12 November 2013 - 10:38 PM

I also find it a little ironic that your visual example is almost the exact opposite of what is happening in nature. Ie in your example its a small flame in a tiny room surrounded by too much light, where as with the sun, it's a tiny window shining way too much light to see the little bit of light that is surrounding it.

A good example none the less :)

#9 brianb11213

brianb11213

    Cosmos

  • *****
  • Posts: 9047
  • Joined: 25 Feb 2009
  • Loc: 55.215N 6.554W

Posted 13 November 2013 - 04:30 AM

I realize that your post is on using Ha Filter for visual work on the chromosphere, but has anyone tried it for solar AP? I'm just curious.

Same thing.

I'm looking at doing some solar stuff next year with my SCT. If you added an occluder could you do it then?

Build a coronascope using a SCT as a basis?

1. I think there would be way too much scatter in the corrector plate & off the two mirror coatings.

2. You'll definitely fry the secondary unless you're using an energy reduction filter overthe obvective. A good ERF big enough to fit a SCT will likely cost more than the SCT itself, and there is nowhere a subaperture ERF could be fitted without severely restricting the aperture.

The only practical ways to get Ha chromospheric detail from a SCT are to use a front-mounted etalon / ERF / blocking filter set, picking a suitable sized etalon to fit in an off-axis adapter (Lunt / Coronado / Solarscope), or to use an off-axis subaperture ERF over the objective and a back-end etalon / blocking filter set (Daystar).

The mechanical issues with SCTs used in the daytime (princiaplly tube currents due to differential heating by solar radiation) are such that it is almost certain that a dedicated solar scope will perform better for similar expenditure & will certainly be more convenient to use.
  • farsight and eyedoc1984 like this

#10 jeffpkamp

jeffpkamp

    Ranger 4

  • -----
  • Posts: 322
  • Joined: 11 Jan 2013

Posted 13 November 2013 - 09:22 AM

I am getting a baadar solar film filter if that is what you mean by an ERF. Can't be destroying my telescope or setting it, my eyes, or my camera on fire now :). I was going to build a full aperture filter and a sub aperture filter and see which worked better. I don't even have a Ha filter, so I'm not terribly worried about seeing the chorosphere at the moment, I was just curious if people had tried it and why it didn't work.

#11 brianb11213

brianb11213

    Cosmos

  • *****
  • Posts: 9047
  • Joined: 25 Feb 2009
  • Loc: 55.215N 6.554W

Posted 13 November 2013 - 10:28 AM

I am getting a baadar solar film filter if that is what you mean by an ERF. Can't be destroying my telescope or setting it, my eyes, or my camera on fire now :).

True, but the solar film will be far too dense to allow any chance of seeing anything but the photosphere. An energy reduction filter (ERF) removes dangerous ultra violet & infra red radiation & most of the visible spectrum too but has a high transmission at the wavelength (or wavelengths) of the spectral line(s) it's supposed to work with.

I was going to build a full aperture filter and a sub aperture filter and see which worked better.

Good idea, the seeing is very often too bad for an aperture larger than about 4 inches to be of any help at all, though just occasionally (especially when the sun is still lowish in the morning) you may get a reasonably steady view even with the full aperture of an SCT. A light coloured towel or something else providing insulation, wrapped around the metal tube of an SCT, usually helps reduce tube currents.

#12 vtornado

vtornado

    Vanguard

  • *****
  • Posts: 2093
  • Joined: 22 Jan 2016
  • Loc: Northern Illinois

Posted 23 March 2016 - 10:13 PM

I'm sure someone is going to say this won't work, but ...

 

What if before the telescope objective, you place a   UV/IR cut filter, then a wratten #29 filter, then the objective, then the H/A nebula filter.  Orion sells one that has a .7 angstrom pass.  How much energy is going to get through the H/A?

 

Of course any experimenting I do will be to a web-cam.  I can fry a $30 web cam.  I only have two somewhat working

eyes.



#13 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6569
  • Joined: 12 Jun 2004
  • Loc: Washington USA

Posted 25 March 2016 - 12:25 AM

You will indeed fry your web-cam, and just get a blindingly and dangerous bright red continuum image.  Orion sells no such filter -- their "extra-narrowband" H alpha filter is 7 nm (nanometers) FWHM, which is 70 Angstroms:

 

http://www.telescope...CFU1gfgodyM8KkQ



#14 jspielberg

jspielberg

    Sputnik

  • -----
  • Posts: 36
  • Joined: 25 Oct 2014

Posted 24 April 2016 - 03:29 PM

Bob -  great explanation!

 

<offtopic>

vtornado -

In case you don't trust Bob.

Orion says it right at the bottom of the page he linked.

 

"Note: This filter should NOT be used for solar observation."

Seems pretty cut and dried.

</offtopic>



#15 vtornado

vtornado

    Vanguard

  • *****
  • Posts: 2093
  • Joined: 22 Jan 2016
  • Loc: Northern Illinois

Posted 03 May 2016 - 10:18 AM

You will indeed fry your web-cam, and just get a blindingly and dangerous bright red continuum image.  Orion sells no such filter -- their "extra-narrowband" H alpha filter is 7 nm (nanometers) FWHM, which is 70 Angstroms:

 

http://www.telescope...CFU1gfgodyM8KkQ

Thanks oops, my bad memory, I thought an angsrom was 10  ^ -8, it's 10 ^ -10.  Thanks for catching my faulty memory.

 

Yes I trust bob.  I don't trust Orion.



#16 lowdowndan

lowdowndan

    Lift Off

  • -----
  • Posts: 15
  • Joined: 22 Sep 2016
  • Loc: Los Angeles, CA

Posted 15 March 2018 - 05:58 PM

I don't get it.  First, the sun provides so much light that it has to be dimmed, even when using a hydrogen alpha filter. Even they have prefilters to make sure that the scope and internal optics don't absorb too much heat.  My white light filter does the same, although to a level that allows me to observe with the naked eye.

Therefore, all of the light, including the hydrogen alpha line has been reduced in intensity.

Now I put on the 7nm hydrogen alpha filter, which filters out all the remaining light outside the 7nm window.  I believe that my filter transmits approx. 85% at the H alpha peak, but don't quote me on this.  Therefore, the remaining light outside the 7 nm, or 70 A window is blocked and all that is left is a greatly attenuated spectrum centered, presumably, around the H alpha wavelength.  The background light intensity is approximately 140 times the amount of light transmitted through a 0.5 A filter (70/0.5), while the amount of remaining hydrogen alpha light is the same. Therefore, the peak-to-background ratio is greatly reduced, to the level that an observer can't see the hydrogen alpha structures at all.  

Now I use my digital SLR. There is plenty of light to take photos using the filter combination, obviously much more that with night viewing.  I have a general background that is 140 times as intense as with the 0.5A systems, but I also have 14 bits (factor of 16000) of dynamic range to play with.  Why can't I adjust the contrast to eek out some of the hydrogen alpha contrast in the image?  If the rest of the 70A light is relatively homogeneous (true?), I should be able to crank the contrast to visualize at least some of the Hydrogen Alpha structure.  Is it that the rest of the 70A light is not relatively homogeneous and the contrast increase will just accentuate the black body variations? What about prominences? 

Finding a one part in 140 signal in a system that has a dynamic range of one part in 16000 should not be that difficult.

What am I missing? In any case, I will try it and see what I get...



#17 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6569
  • Joined: 12 Jun 2004
  • Loc: Washington USA

Posted 15 March 2018 - 08:14 PM

 

 

Finding a one part in 140 signal in a system that has a dynamic range of one part in 16000 should not be that difficult.

What am I missing? In any case, I will try it and see what I get..

 

You are missing that the brightness ratio is not 1:140 based on the filter bandpasses. It is 1:100,000 based on the brightness difference between the photosphere and the chromosphere, and you will have done nothing to improve this difference, other than to make both the photosphere and the chromosphere ~ 100,000 times dimmer with the white light filter. All you will be able to image is the photosphere - reduced a bit in brightness due to the bandpass of the 7 nm filter.

 

By all means, give it a try and see. waytogo.gif



#18 vincentv

vincentv

    Vostok 1

  • *****
  • Posts: 141
  • Joined: 16 Apr 2017
  • Loc: Costa Rica

Posted 16 March 2018 - 09:10 AM

As a newbie my math is probably off in more than one way, but here's what I think will happen:

The h-halpha line is a bit under 2 angstroms wide total, including wings. It has a brightness of roughly 16% of the surrounding continuum.

That means that for every 100 photons per angstrom in the continuum, you'll get 32 of halpha. You have 70 angstroms - 2 = 68 angstroms of continuum*.

You will receive 6800 photons of continuum for very 32 of halpha. That's before you take into account camera noise. You'll have to be extremely careful to take out noise sources to even approach some of the signal. A narrower filter will dramatically improve your chances.

Your best bet will be prominences since you will only fight atmospheric and equipment dispersion. Even then chances are low. But go ahead and have fun experimenting, i'm actually curious of how much data you can extract.

 

 

 

*: I'm being generous here. Given the shape of the filter's transmission curve and how FWHM is measured you'll actually get more continuum leaked. I'm just not sure how to calculate it.



#19 lowdowndan

lowdowndan

    Lift Off

  • -----
  • Posts: 15
  • Joined: 22 Sep 2016
  • Loc: Los Angeles, CA

Posted 16 March 2018 - 12:00 PM

As a newbie my math is probably off in more than one way, but here's what I think will happen:

The h-halpha line is a bit under 2 angstroms wide total, including wings. It has a brightness of roughly 16% of the surrounding continuum.

That means that for every 100 photons per angstrom in the continuum, you'll get 32 of halpha. You have 70 angstroms - 2 = 68 angstroms of continuum*.

You will receive 6800 photons of continuum for very 32 of halpha. That's before you take into account camera noise. You'll have to be extremely careful to take out noise sources to even approach some of the signal. A narrower filter will dramatically improve your chances.

Your best bet will be prominences since you will only fight atmospheric and equipment dispersion. Even then chances are low. But go ahead and have fun experimenting, i'm actually curious of how much data you can extract.

 

 

 

*: I'm being generous here. Given the shape of the filter's transmission curve and how FWHM is measured you'll actually get more continuum leaked. I'm just not sure how to calculate it.

 

I agree with your math. The math, like mine, is back of the envelope. Your assessment was a photon ratio of 212, while my relative contrast degradation was 140, these numbers are not so far off, again given that the camera is able to manage 14 bits of dynamic range.  Photon counting is irrelevant, this is not deep sky photography, we are not trying to take long exposures, so photon counting and dark current should be irrelevant. I suspect the exposure will be quite fast. I looked through the scope by eye with the two filters and the view was relatively bright, nothing like a nebula.  In any, case, I plan on taking the scope out tonight to align it (LX200 classic, I don't know how to align during the day) and take my first pictures tomorrow morning.  Then off to photoshop processing to see if I can pull image out.  I will expose "to the right", making sure no pixel is saturated, and see if I can pull the 7 stops of signal out of the background. Photos (good or bad) should be up tomorrow...



#20 lowdowndan

lowdowndan

    Lift Off

  • -----
  • Posts: 15
  • Joined: 22 Sep 2016
  • Loc: Los Angeles, CA

Posted 16 March 2018 - 12:21 PM

Sorry to ramble, but let's say that the photosphere's structure is too complex to do a simple contrast enhancement. The process I propose assumes that the background is relatively homogeneous so that the overall brightness level contrast can be globally expanded, which of course may not be true at the 1:200 ish level.  But what if I also take a full-spectrum image (no H alpha filter) with my SLR, isolate the red channel, then use that (grayscale) as a background image for subtraction (of course the H alpha image will also employ only the red channel)? I just need to determine an optimal normalization for the full-spectrum image, which I assume I could do iteratively, perhaps maximizing the local contrast of the resulting net image.  

 

I will only have to carefully align the two images so that the subtraction makes sense within the photosphere structure. I could even do a deformable registration using the photosphere's apparent structure in both images to guide the registration, which should be no problem since, as was described, the H alpha signal is a factor of 200ish below the photosphere signal, so wouldn't have an impact on the registration. This would reduce the problems associated with atmospheric distortion of the sun causing local misalignment that would make the subtraction incorrect.  B-spline using mutual information?  I really don't see why this wouldn't work.  That said, the proof will be in the pudding.

 

More thinking: I have a wider H alpha filter, the Baader UHC-S filter, that has a 35 nm width around the H alpha line.  The other transmission bands of this filter are outside the red, so the red camera channel will exclude those.  While the camera has an IR filter that cuts 75% of H alpha, it cuts about the same from both filters, so again we only reduce overall photons, of which the sun provides plenty.  Therefore, I will use the 35nm UHC-S filter to provide the "background" and the 7 nm to provide a combination of the "background" and H alpha signal and do the normalized, registered subtraction.  

 

Hmmmm



#21 vincentv

vincentv

    Vostok 1

  • *****
  • Posts: 141
  • Joined: 16 Apr 2017
  • Loc: Costa Rica

Posted 16 March 2018 - 12:54 PM

You'll probably need to saturate the disk to even have a chance with prominences. 

You can do the match backwards:

Assume a halpha telescope has a transmission of 0.1 to 0.2. With a canon DSLR i get prominences with this settings: iso 200, F 8.3, 1/25s (40ms).

 

The baader ND5 film has a transmission of 0.00001. You'll need an exposure 10000 to 20000 times longer to capture the same amount of halpha light. Holy cow, 400 seconds with the above settings!

 

Let's try baader ND3.8 *photography* filter. Transmission is 0.00016. That means you need between 625 and 1250 longer exposure times to match the halpha scope. 

 

I had not done this numbers before and the results are waaay beyond what I expected. I invite everyone to double check.

 

Edit: For surface detail I get some results with ISO200, 1/80s (12ms), F9.


Edited by vincentv, 16 March 2018 - 01:00 PM.


#22 lowdowndan

lowdowndan

    Lift Off

  • -----
  • Posts: 15
  • Joined: 22 Sep 2016
  • Loc: Los Angeles, CA

Posted 16 March 2018 - 01:12 PM

You'll probably need to saturate the disk to even have a chance with prominences. 

You can do the match backwards:

Assume a halpha telescope has a transmission of 0.1 to 0.2. With a canon DSLR i get prominences with this settings: iso 200, F 8.3, 1/25s (40ms).

 

The baader ND5 film has a transmission of 0.00001. You'll need an exposure 10000 to 20000 times longer to capture the same amount of halpha light. Holy cow, 400 seconds with the above settings!

 

Let's try baader ND3.8 *photography* filter. Transmission is 0.00016. That means you need between 625 and 1250 longer exposure times to match the halpha scope. 

 

I had not done this numbers before and the results are waaay beyond what I expected. I invite everyone to double check.

Since this is all academic until we try, let me ask this. What is the width of your transmission filter, 0.5 Angstroms? 



#23 vincentv

vincentv

    Vostok 1

  • *****
  • Posts: 141
  • Joined: 16 Apr 2017
  • Loc: Costa Rica

Posted 16 March 2018 - 01:18 PM

Single stack so around 0.7 or 0.6 angstroms. 


  • hahied likes this

#24 lowdowndan

lowdowndan

    Lift Off

  • -----
  • Posts: 15
  • Joined: 22 Sep 2016
  • Loc: Los Angeles, CA

Posted 16 March 2018 - 02:15 PM

The photosphere has a dip in output at the hydrogen alpha line.  According to BASS2000, the solar intensity has an 80% dip at the hydrogen alpha wavelength due, I assume, to hydrogen atoms in the photosphere absorbing that wavelength.  Therefore, the contrast between photosphere and chromosphere is improved by a factor of 5 at that wavelength, which hurts attempts to use broader wavelength filters. 



#25 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6569
  • Joined: 12 Jun 2004
  • Loc: Washington USA

Posted 17 March 2018 - 11:08 PM

Indeed.

 

Here's the solar photosphere absorbtion profile for 1 nm (10 Å) centered on the H alpha line at 656.28 nm, shown with the transmissions for a 0.7 Å and 0.5 Å (double stacked) H alpha solar filter:

 

0.7 v 0.5 FWHM.jpg

 

Here's the transmission profile of the Baader 7 nm H alpha filter, with the 1 nm horizontal scale shown above indicated in grey:

 

ha_7nmfiltercurvesm2.gif

 

Note that continuum begins to manifest itself at +/- 0.65 Å on either side of the H alpha line at 6562.8 Å. See here:

 

https://www.cloudyni...-2#entry6792695


Edited by BYoesle, 18 March 2018 - 09:53 AM.



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