Jump to content


Photo

Single Vs. Double Stack

  • Please log in to reply
84 replies to this topic

#26 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15450
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 28 December 2012 - 02:21 AM

BYoesle wrote:

The photosphere too is real, and I suppose in this way it is not literally an “artifact” by the definition you are using. A better word might have been “defect” as it applies to the filter’s performance, if your desire is for it to only show the chromosphere. But it seems you would argue this too is false, as the filter is performing within specification. That would be true, and I believe that is the point


Yes, it is not an artifact or a defect, so the use of that word would also definitely be incorrect. If you see this in a filter with a designed bandwidth of 0.7 angstroms, the filter is not defective. It is performing about as well as it can given the design limitations. It is the appearance of the sun with the given bandwidth that is the only question., and again, the narrower the filter's bandwidth specs are, the higher the contrast of disk detail tends to be. That is the only point to be made here. Clear skies to you.

#27 ValeryD

ValeryD

    Vendor (Aries)

  • *****
  • Vendors
  • Posts: 1127
  • Joined: 26 Nov 2005
  • Loc: Kherson, Ukraine.

Posted 28 December 2012 - 03:29 AM

Post deleted by ValeryD

#28 Bill Cowles

Bill Cowles

    MacGyver

  • *****
  • Posts: 2810
  • Joined: 16 Apr 2006
  • Loc: Utah

Posted 28 December 2012 - 02:25 PM

:waytogo:

Bill

#29 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 28 December 2012 - 10:37 PM

I believe there was expressed a wrong point of view how the narrower band wide can be achieved.

If both identical etalons will be adjusted that the vertexes of their bandwides will fully coincide then we have the same bandwide. And narrower bandwide can be achieved ONLY if we will shift one bandwide vs second one. In this case we will have significantly lower maximum of transmittance, narrower bandwide and much supressed wings.

In such way we can achieve the same narrow bandwide even with, say, two 0.9 etalons. However maximum transmittance will be lower and wings suppression will be lesser.


Hi Valery,

This is incorrect. I have very successfully used two identical bandwidth filters to narrow the bandwidth as stated by David Lunt in the post above. The ideal is to have one filter be on band with no tilt, and to have a second filter that has the identical bandwidth/center when slightly tilted to remove ghost reflections between the two etalons.


Now I get it (I think). I had misunderstood what Zirin meant by the "general chromosphere." Figure 7.5 is neat because the shots are close in time and the spicules are the same from frequency to frequency. You can see individual spicules uncovering themselves as you go off band.

So the bottom line is that the lower disk you see is indeed the photosphere, and it marks the base of the chromosphere, and the spicules start more or less at the base of the chromosphere, although we only see their tops because of the intranetwork (non-spicule) structure, which Zirin calls the general chromosphere.

Bob, have you tried to duplicate Zirin's photo, using a SS filter so that the photosphere is visible and trying to see closer to the base of the spicules by going slightly off-band? That might be a fun experiment.

George


Hi George,

The “general chromosphere” as Zirin refers to it, is shown in the on-band images of Fugure 7.5. It is also as demonstrated by the images I made above for both the 0.7 and < 0.5 A bandpasses which were taken on-band.

Spicules are made more visible with off-band tuning, and this of course introduces some more continuum energy into the image, showing the limb of the photosphere. This is what David was referring to.

So while I haven’t tried to image spicules @ 0.7 A with off-band tuning, they would be just as easily imaged with a 0.5 A filter with off-band tuning. When a < 0.5 A filter is used on-band, only the top of a few larger or macrospicules become visible, as in the animation above. The vast “spicule forest” on the limb remains hidden by the general chromosphere when the filter is on-band.

In another publication, Johannesson and Zirin noted that when using a very narrow <0.5 A filter:

The image through this pure filter [the Rakuljic (Rakuljic & Leyva 1993 ) holographic filter] does not show the spurious inner limb, which was shown by White and Simon (1966) to be due to continuum passing through the Lyot filter sidebands...

... Images in the wing of H-alpha show Doppler-shifted moving features in the chromosphere, namely spicules. Despite the evidence of the Dunn-Zirker images and various eclipse measurements, there is a general belief among solar physicists that the general chromosphere extends only to 2000 Km, as in the VAL model (Vernazza, Avrett & Loeser 1981). It is thought that the higher H-alpha emission is due to spicules. Our result establishes the qualitatively obvious fact that, except for macrospicules, the general chromosphere extends above the average spicules.

The off-band (spicule) limb is always lower than the centerline limb by an average value of 500 Km (0.7 arcsec). Because of the self reversal of the chromospheric H-alpha line, the off-band chromosphere is twice as bright as the centerline. As a result, the gradient of the off-band limb profile is considerably steeper just above the photosphere. Thus the popular view that spicules rise above the chromosphere is incorrect, except insofar as the macrospicules are concerned.

Johannesson, A. & Zirin, H. (1996) The Pole-Equator Variation of Solar Chromospheric Height. Astrophysical Journal, 1996, Volume 471, pp. 510-520


So to get back to the more salient issue of this post’s topic, and the point of narrowing the bandpass via double stacking:

Narrow passbands of 0.5 A or less... are required to take advantage of the highest opacity in the cores of Ha, Hb, and Ca H and K and thus achieve adequate contrast to study the disk structure.

Foukal, P.V. (1990) Solar Astrophysics (p. 293). New York: Wiley & Sons.


It is therefore established that a 0.7 A FWHM filter will admit continuum light from the photosphere when it is centered on-band. This reveals the limb of the photosphere showing a boundary between the photosphere and the general chromosphere. This is an inherent property of a filter with the wider bandpass. It should not be confused with a filter tuned off-band to show greater spicule details.

It is furthermore established that a < 0.5 A FWHM filter will not admit continuum energy from the photosphere when tuned on-band, and shows only the general chromosphere, without showing the limb of the photosphere. When tuned off-band, it too will admit continuum energy, revealing the disk of the photosphere.

It is furthermore established the chromosphereic spicules are made more visible when the filter is tuned into the wings of the H alpha line. This will also introduce continuum energy, revealing the limb of the photosphere, irrespective of bandpass.

Filtergrams below from the Astrophysical Journal, 1996, Volume 471, The Pole-Equator Variation of Solar Chromospheric Height, Fig. 1:

(a) An off-band chromospheric image. Several arrows mark the locations of dark and bright spicules crossing the limb. The spicule is brighter than the sky, but normally is in absorption against the disk, even near the extreme limb. (B) A centerline image at the same position angle but a few hours later showing the inner limb due to sidebands of the Zeiss (Lyot) filter. © A limb image on a different date with a pure filter (Rakuljic filter), with the inner limb barely detectable from the morphology.

Attached Files



#30 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 28 December 2012 - 11:32 PM

Very very good, Bob.

I cannot find an image anywhere, but if you take two Gaussians [this should have been Lorentzians] and multiply them (i.e., two filters in series), you get the best output when their centerlines coincide. When you start to shift them, the output just becomes dimmer, not narrower.

George

#31 ValeryD

ValeryD

    Vendor (Aries)

  • *****
  • Vendors
  • Posts: 1127
  • Joined: 26 Nov 2005
  • Loc: Kherson, Ukraine.

Posted 29 December 2012 - 03:04 AM

Post deleted by ValeryD

#32 bob71741

bob71741

    Apollo

  • *****
  • Posts: 1022
  • Joined: 16 Feb 2008

Posted 29 December 2012 - 08:22 AM

Post deleted by bob71741

#33 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 29 December 2012 - 11:34 AM

For Lorentzian, as you shift the filters with respect to each other, it not only becomes dimmer, it also becomes wider. I tried to show it in the attached figures. The first shows (on an arbitrary frequency scale centered at 0) a single stack (SS) and double stack (DS center), the latter being the square of the first.

The second figure shows two SS's each offset in opposite directions but with the same bandwidth. The product is "DS shift," and "DS shift norm" is the same curve normalized to better see the width.

The third figure compares DS center to DS shift. The normalized version demonstrates that it gets wider.

As always, I could be wrong, but these were the kind of figures I was looking for. Sorry about the wrong function earlier.

George

#34 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 29 December 2012 - 11:35 AM

Figure did not attached. Let me try again. George

Attached Files



#35 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 29 December 2012 - 09:01 PM

Very nice diagrams George!

Question: Bob, have you tried to duplicate Zirin's photo, using a SS filter so that the photosphere is visible and trying to see closer to the base of the spicules by going slightly off-band? That might be a fun experiment.

George



Answer:

Hi George,

... So while I haven’t tried to image spicules @ 0.7 A with off-band tuning, they would be just as easily imaged with a 0.5 A filter with off-band tuning. When a < 0.5 A filter is used on-band, only the top of a few larger or macrospicules become visible, as in the animation above. The vast “spicule forest” on the limb remains hidden by the general chromosphere when the filter is on-band.

Bob


The top illustration shown below exhibits the transmission shapes to scale for a single H alpha etalon with 0.7 A FWHM (green) verses 0.5 A FWHM (red) pair of double stacked etlons (superimposed on a representation of the sun’s absorption spectrum). Your diagrams too show very well the changes that occur with double stacking per David Lunt’s description:

... Thus the effect is to narrow the actual bandwidth and increase the visibility of chromospheric detail, while the steeper shape of the passband reduces the out of band transmission, thus significantly improving contrast.


The much broader “tails” and (and to a lesser degree wider bandpass) demonstrates why the photosphere becomes visible at 0.7 A, while it is invisible with a < 0.5 A DS system with it’s steeper sides and suppressed tails.

But also of interest (and related), when viewed with the emission spectrum (bottom) of the of the chromosphere at the same scale (distance between the “horns” is +/- 0.7 A = 1.4 A), we may see why the “general chomosphere” is all that can be seen with the narrower bandpass filter when it is "on-band" (as in filtergram c above), while at a wider bandpass the photosphereic disk appears, along with evidence of the network of Doppler-shifted spicules (filtergram b above). When a < 0.5 A filter is tuned off-band to the red or blue wing, the Doppler-shifted spicules should reveal themselves clearly, albeit with less continuum (photospheric) interference than the wider bandpass.

I will try to image this at < 0.5 A blue shifted (all I can do s tilt my filters to the blue wing) when the weather improves this spring, but some others (perhaps Bill in sunny Utah for instance ;)) might be able to show this to us much sooner.

The H-alpha spectrum is most interesting (Figure 2)... Bright emission horns about twice the brightness of the central core appear at either wing. The structures seen in the horns are due to the slit crossing several spicules... The photospheric absorption line continues into the chromosphere without change, but bright emission horns of brightness equal to the photosphere appear at +/- 0.7A extending up to about 4000 Km. With good seeing, these break up into emission from individual spicules, extending to greater line shifts than the static chromosphere. At greater heights the horns disappear...

Johannesson, A. & Zirin, H. (1996) The Pole-Equator Variation of Solar Chromospheric Height. Astrophysical Journal, 1996, Volume 471, pp. 510-520


Lower image: Astrophysical Journal, 1996, Volume 471, The Pole-Equator Variation of Solar Chromospheric Height, Fig. 2:

Attached Files



#36 doctomster

doctomster

    Lift Off

  • -----
  • Posts: 13
  • Joined: 04 Feb 2008
  • Loc: North Carolina

Posted 29 December 2012 - 09:42 PM

I use a Lunt 152 LS to which I have added a Coronado 90 mm etalon up front to provide < 0.5 Angstrom bandwidth. The views are great as long as I don't dial up the magnification too high. Using an eyepiece less than 8 mm then the central obstruction of the Coronado begins to make its presence known.

#37 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 177
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 29 December 2012 - 09:53 PM

By far, the most significant contributor to the visible photosphere in a single etalon system is the poorly suppressed etalon modes blue and red of the desired mode. The combined transmission of these can easily exceed 5%, and they are located well outside any part of the H alpha absorption, where the intensity is 5 times that in the line core. In a phone discussion with one of the GONG design staff earlier this year the subject came up, relating to why imaging disk and proms was difficult with a single stack system. The NSO staffer (who's been doing this since before most of us were born) said that in single etalon systems the sideband leakage would present exactly the situation observed, and that professionals would be using these instead of the expensive alternatives if the performance was the same.
So why does a double stack eliminate the problem? Manufacturing tolerances. A difference of just 20 microns in the spacing in an air spaced etalon, or less of a variation in the thickness of a solid etalon, will shift the offending sidebands to the point where they fall beyond the skirts of the passband of a similar etalon (of the original spacing). This, combined with the usual blocking filter, results in a near total elimination of the unwanted light passed in these sidebands, if two slightly different etalons are stacked together. Evidence shows that the thickness is not that well controlled. Simply reducing the bandpass from 0.7 to 0.7 Angstroms isn't enough, but math shows that the sideband leakage is there, and that eliminating it makes enough difference to explain what is observed.

#38 Bill Cowles

Bill Cowles

    MacGyver

  • *****
  • Posts: 2810
  • Joined: 16 Apr 2006
  • Loc: Utah

Posted 29 December 2012 - 09:55 PM

Sunny Utah? :p

Bill

Attached Files



#39 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 177
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 30 December 2012 - 12:10 AM

Here's a movie, scanning from 2 Angstroms blue of H alpha to 2 Angstroms red, 0.35 Angstrom bandwidth. The transition from photosphere to chromosphere is smooth, you can see the image "swell" as it passes through the line center. The seeing was bad and variable, as you can see from the ripples on the limb, but you can get the idea.

Posted Image

#40 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 30 December 2012 - 12:49 AM

I have learned a great amount in this thread. Thanks again, Bob.

I always wondered what manufacturers meant when they said they were "matching" two filters to make sure that they could be double stacked. Based on Joe's post, were they just making sure that the sidebands did not happen to line up?

David Lunt worried that my ASP-60 and a friend's ASP-60 would not make a good double stack, but it worked well. Perhaps the fact that they were manufactured a year apart actually helped.

George

#41 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 177
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 30 December 2012 - 01:28 AM

George, your plots are correct, if you're still wondering. I've been lazy in using Gaussians to show the effect of stacking filters. I had a Gaussian spreadsheet already built and didn't want to take the time to do it right. I spent some time today to make a Lorentzian version and it agrees with yours.

#42 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 30 December 2012 - 01:50 AM

Wow Joe - that’s a spectacular video! :bow: I note how in the red and blue wings chromospheric disk details like filaments (as well as prominences) disappear, but the spicule forest of bright patches and mottles really pops out - just as they do (i.e. spicules) at the limb:

...we can now see that Secchi’s needles (the “spicules”) are arrayed in long picket fences... Spicules lie at the borders of supergranule cells...

Zirker, J.B. (2002) Journey from the Center of the Sun (p. 142). New Jersey, Princeton University Press.



Hi George: The inside story on ‘matching’ of front mounted etalons can be found here:

http://www.cloudynig...Board=solar&...

Then see this post for my solution:

http://www.cloudynig...rd=solar&amp...


In a phone discussion with one of the GONG design staff earlier this year the subject came up, relating to why imaging disk and proms was difficult with a single stack system. The NSO staffer (who's been doing this since before most of us were born) said that in single etalon systems the sideband leakage would present exactly the situation observed...

[Joe]


This is another reason for double stacking to reduce bandwidth (as noted above). Besides improving contrast, removing the light of the photospheric disk equalizes the brightness of the disk and the prominences, making capturing both in the same exposure much easier, and eliminates the need to combine separate exposures for each - note the filaprom @ ~ 10:30:

June 15, 2012 21:37 UTC

Attached Files



#43 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 30 December 2012 - 02:40 AM

When you process a properly exposed 0.7 A disk filtergram to bring out disk detail with increased contrast, you usually end up wiping out the limb chromosphere and prominenecs. That’s why you may be required to make two separate exposures, or use more involved processing to obtain both disk and prominence details:

Attached Files



#44 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15450
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 30 December 2012 - 03:36 AM

Spectral Joe wrote:

Observing the Sun with complex optical systems since 1966, and still haven't burned, melted or damaged anything.
Not blind yet, either!


I love that sig! A friend of mine lives up in the wilds of the Paul Bunyan State Forest in Minnesota on Mantrap Lake and I went to visit him one year. We had his 10 inch f/5 Newtonian out at the lake shore (three inch off-axis ERF and "Barlowed" to f/31) with his DayStar T-Scanner and were watching a post-flare arcade develop on the limb when some people came by on a boat and shouted out, "You'll go blind!!". We just smiled and went back to the scope. Clear skies to you.

#45 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 30 December 2012 - 11:14 AM

George, your plots are correct, if you're still wondering. I've been lazy in using Gaussians to show the effect of stacking filters. I had a Gaussian spreadsheet already built and didn't want to take the time to do it right. I spent some time today to make a Lorentzian version and it agrees with yours.


The thing is, when I incorrectly did the Gaussian the day before, I discovered that as you shift, it becomes dimmer but not wider. That is, when you normalize, all the curves coincide. I guess it's a moot point unless there are also Gaussian filters out there, but it's an interesting property. I knew it was supposed to get wider, but I didn't make the connection until seeing Valery's post. Bottom line, I guess, is that Gaussian turns out not to be a good shortcut.

Thanks for the double stacking links, Bob. So the matching is more mechanical. I remember seeing your great innovation when you first posted it. I just put up with the two dark ERFs and the TMax. Although when one of my filters decontacted after 12 years and I had to hacksaw the ASP-60 apart to fix it, I made it so I could leave one ERF off, and that brightened the double stack.

George

#46 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 30 December 2012 - 01:14 PM

Hacksaw?! :foreheadslap: You’re a brave man George!

I made it so I could leave one ERF off, and that brightened the double stack.



:waytogo:

Just in case others are as rusty on their math and statistics as I am, I thought I’d post what the shapes for Gaussian (“normal” or “Bell”) distributions verses Lorentzian (“Cauchy”) distributions look like (upper diagram).

It seems that unless you “do the math,” it can be hard to tell the difference given the variables of amplitude and deviation (lower diagram) - where the Gaussaian vs. the Lorentzian distribution looks almost identical to the DS v. SS profiles above...

Attached Files



#47 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 30 December 2012 - 05:25 PM

Your bottom curve really shows it well. I threw in a double and quadruple stack, all at the same FWHM. It looks suspiciously like a large number of Lorentzians will approach Gaussian.

(I had to hacksaw. Lunt said neither they, nor Meade, nor Isle of Man would be able to do it cost-effectively, so I had nothing to lose. Luckily the spacers remained stuck to one element.)

George

Attached Files



#48 bill1234

bill1234

    Viking 1

  • -----
  • Posts: 768
  • Joined: 21 Apr 2011
  • Loc: NJ, USA

Posted 30 December 2012 - 07:41 PM

excellent post...thanks for starting and continuing...

#49 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4861
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 30 December 2012 - 08:18 PM

Thanks for the plots George.

I was wondering if you might run the following bandpasses and post them as you have above at the same amplitude:

Lorentzian - 0.8, 0.7, 0.5, and 0.35 Angstroms FWHM respectively (this last value is what I assume results from a triple stacked series of 0.7 FWHM etalons). Is quad stacking actually a 0.24 A FWHM bandpass?

It would be nice to have a comparison for what Jim Ferreira is imaging at 0.8 A (which obviously shows a near-photospheric disk and lots of Doppler-shifted as well as on band prominences), as well as what I and others are seeing at 0.7 A and 0.5 A, and what Jesus Munoz has shown with his SS, DS, and TS images seen here:

http://www.cloudynig...rd=solar&amp...

From what Jesus’ images show in the thread above, and interpolating a curve between your green Lorentzian ^2 curve (double stacking) and the red Lorentzian^4 curve (quad stacking), I’m not sure I see much benefit to triple stacking compared to double stacking - but then again the curves shown above are at the same FWHM. It would be nice to see some quantitative data or graphical representations, and to perhaps show what the narrower DayStar or Solar Spectrum filters actually accomplish - assuming these solid etalons provide similar shaped transmission curves.

Joe might press his awesome spectrohelioscope into service for showing some controlled comparison images to go along with these graphs - or perhaps he already has. It’s hard to find solid info like this anywhere...


Addendum

Finally had some sun today, but the seeing was horrible (1-2/10). Temp -1 C, solar altitude ~ 22 degrees (local noon), relative humidity 26%, winds SE 5-10 km/h. Close ups of the limb spicules off-band will have to wait for a much higher sun, and hopefully warmer temps. But I did get images of the chromosphere double stacked @ < 0.5 angstroms: top on-band @ 20:43 UTC, bottom ~ 0.7 A to the blue wing @ 20:44 UTC. Tuning was accomplished using the etalon closest to the objective (i.e both etalons were tilted together)...

The change off-band is obvious, showing the spicule (mottle) forest and much diminished general chromosphere features (Joe’s video is much better). Both images with identical exposures (5 ms), gain (575), gamma (1500), and processing. Paradoxically, the off-band image is dimmer - any ideas about this??? From the discussion above more continuum from the wing should be mkaing it through (especially considering tilting also widens the bandpass slightly), and therefore it should be brighter - no?

The thing is, when I incorrectly did the Gaussian the day before, I discovered that as you shift, it becomes dimmer but not wider.

[George]


Hmmm.....

Attached Files



#50 George9

George9

    Messenger

  • *****
  • Posts: 421
  • Joined: 11 Dec 2004

Posted 30 December 2012 - 11:46 PM

Very good image despite the conditions. I gave it a visual try today, but seeing was bad for me, too.

Here is what I ran:

1. 0.8 FWHM of single stack
2. 0.7 FWHM of single staack
3. double stack of 0.7, which is 0.45052 FWHM
4. triple stack of 0.7, which is 0.35688 FWHM
5. quadruple stack of 0.7, which is 0.30449 FWHM

In the figure, "Lorentzian" is 0.7 unless otherwise specified. Remember that a real filter won't have 100% transmission, so each increase in stacking will actually be dimmer. I welcome someone to check my derived FWHM's. I just iterated to the answers. And very happy to replot if I misunderstood.

(On the Gaussian, yes it seemed odd. You could send one up to violet and one down to infrared and theoretically it would be just as narrow, but just very dim. Would need to double check that.)

George

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics