Jump to content


Photo

Single Vs. Double Stack

  • Please log in to reply
84 replies to this topic

#1 drksky

drksky

    Surveyor 1

  • *****
  • Posts: 1760
  • Joined: 31 Aug 2009
  • Loc: Bloomington, IL

Posted 24 December 2012 - 12:57 PM

I was curious to see the difference between a single and double stack configuration, so I did a little experiment with our SolarMax II 90. Here are the results:
Posted Image

I think the tuning of the DS image was slightly off, but you can see the marked difference between the two.

#2 *skyguy*

*skyguy*

    Surveyor 1

  • *****
  • Posts: 1960
  • Joined: 31 Dec 2008
  • Loc: Western New York

Posted 24 December 2012 - 02:15 PM

I actually find the surface detail more interesting in the single filter image.

#3 drksky

drksky

    Surveyor 1

  • *****
  • Posts: 1760
  • Joined: 31 Aug 2009
  • Loc: Bloomington, IL

Posted 24 December 2012 - 04:22 PM

In this one, yes, I think so, too. But as I said, I think the DS tuning was off. But other DS images I've taken have looked much better. I think I pushed the etalon too far in this one and killed the detail in the active areas.

On the other hand, though, I do think that active areas can be more interesting through a SS as the spot detail shows better.

#4 rigel123

rigel123

    Voyager 1

  • *****
  • Posts: 10749
  • Joined: 29 Jun 2009
  • Loc: SW Ohio

Posted 24 December 2012 - 04:45 PM

I would definitely want to "Try Before I Buy" due to the cost of a DS. It would have to be a huge improvement for me to spend that kind of $$$$. :grin: Evidently it must due to the number that have them!

#5 Pawel

Pawel

    Vostok 1

  • *****
  • Posts: 173
  • Joined: 27 Oct 2008
  • Loc: Tczew, Poland

Posted 24 December 2012 - 07:14 PM

I own two SM90 front etalons and I see biiiig differnce between single and double stack. Belive me, it's worth ;)

#6 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8581
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 24 December 2012 - 07:24 PM

I think a narrower bandpass is the way forward - trouble is its the $$$

#7 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 24 December 2012 - 08:14 PM

Double stacking air spaced etalons reduces the FWHM bandpass from 0.7 angstroms for the single etalon to less than 0.5 angstroms for the pair. This is a very significant benefit, as at 0.7 A FWHM, “parasitic” light form the photosphere is passed through, obscuring chromospheric detail.

I used Coronado SM 90's, and a single Baader D-ERF for each configuration. The images were processed identically, and demonstrate the single SM90 on the left vs. double stacked SM90's on the right. The added brightness of the disk on the right is due to photospheric light leaking through, significantly reducing the contrast of the chromospheric details. Also note that prominences are as easily visible at < 0.5 A. Indeed, due to the reduction of photospheric contamination, prominences generally appear equally if not more “prominent” at < 0.5 angstrom bandpass:

Attached Files



#8 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 24 December 2012 - 08:15 PM

The image on the left clearly shows the “double limb” artifact - the disk of the photosphere leaking through at 0.7 A - rendering a translucent appearance to the chromosphere, and again reducing the contrast of chromospheric detail. Chromospheric detail is much more easily seen and imaged with the DS <0.5 A bandpass on the right. Both images again are identically processed.

Attached Files



#9 dbowlin

dbowlin

    Ranger 4

  • *****
  • Posts: 342
  • Joined: 03 Sep 2009
  • Loc: Littleton, Colorado

Posted 25 December 2012 - 07:05 AM

Great prensentation, BYoesie. Thanks, now I want to DS my SM 60. Will I get similar results? With what little viewing oppritunity I have had this year I don't know if it will be worth it?
Dale

#10 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 25 December 2012 - 09:58 AM

Hi Dale,

You will get identical results with regard to contrast improvement and detail visability. I don't know of anyone who properly double stacks who doesn't see a tremendous improvement. There is an "old-wives" tail about it reducing the brightness of prominences, but as you can see this is false, for obvious reasons. The image will be dimmer overall - however there are measures you can take to improve this, mainly removing one (or more) of the ERF's:

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


#11 dbowlin

dbowlin

    Ranger 4

  • *****
  • Posts: 342
  • Joined: 03 Sep 2009
  • Loc: Littleton, Colorado

Posted 25 December 2012 - 06:02 PM

Thanks BYoesle, I enjoyed your older post. Great food for thought. Half the fun is deciding what to do and how best to do it.
Dale

#12 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15467
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 25 December 2012 - 10:05 PM

The image on the left clearly shows the “double limb” artifact - the disk of the photosphere leaking through at 0.7 A, rendering a translucent appearance to the chromosphere, and again reducing the contrast of chromospheric detail. Chromospheric detail is much more easily seen and imaged with the DS <0.5 A bandpass on the right. Both images again are identically processed.


Well, not quite. What the left image is showing on the limb is the Chromospheric "fringe"; the mass of fine spicules that, depending on how the filter is tuned, can show up as a thick band (sometimes called "the Spicule Forest"). Detuning the filter slightly can let the individual spicules show up in the filter, but the presence of the band on the limb does not necessarily mean that photospheric energy is leaking through. Zirin's book ASTROPHYSICS OF THE SUN on page 161 shows the effect of tuning a very narrow H-alpha filter on the spicule forest. A narrower bandpass filter set to the centerline of H-alpha will tend to screen out some of the slightly Doppler shifted detail (including some of the spicules), so the spicule forest band may begin to break down somewhat and be less prominent (or may show individual spicules that happen to be emitting in the passband of the filter). A broader filter will show less contrast for chromospheric disk detail but by its broadness, let in a little more off-band energy and may make the prominences a little easier to see than in a narrower filter. Thus, double stacking will show better disk detail contrast with a slight loss in the brightness and visible extent of some of the prominences. Clear skies to you.

#13 George9

George9

    Messenger

  • *****
  • Posts: 424
  • Joined: 11 Dec 2004

Posted 25 December 2012 - 10:39 PM

David, I don't doubt that a broader filter can pull in some more energy from spicules or prominences that are slightly off band, but I still assume that the reason for the presence of a dimmer spicule band above a brighter disk in Bob's left photo is photospheric leakage (i.e.,the brighter disk is leaked photosphere), no? In the Zirin p. 161 figure caption, Zirin seems to claim it is actually "the general chromosphere" instead, not the photosphere.

In that same p. 161 figure, I cannot find a citation to the bandwidth. Are you sure it is very narrow?

Also, even if it is nominally narrow, say 0.4 A, what if it is a Gaussian filter with relatively high tails. Couldn't it still be photospheric leakage?

Versus say a double-stacked filter with sharper shoulders and lower tails and therefore less photospheric leakage (i.e., Bob's right photo).

Which actually brings up a separate question about double stacking. Is the improved contrast from double stacking really due to the bandwidth going from 0.7 to 0.5 (the half-height width), or is it really due to changing the shape of the transmission curve from Gaussian to something with squarer shoulders and therefore lower tails? (That is, is most of the leaked energy right near the centerline, or is it spread further away in the tails?) I remember going over this with David Lunt in 1999 or 2000, and we did the math and it seemed to be the tails, not the nominal bandwidth. So a 0.7 A filter with sharp shoulders would also have noticeably better contrast than a regular 0.7 A filter.

George

#14 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15467
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 25 December 2012 - 11:12 PM

George9 wrote:

David, I don't doubt that a broader filter can pull in some more energy from spicules or prominences that are slightly off band, but I still assume that the reason for the presence of a dimmer spicule band above a brighter disk in Bob's left photo is photospheric leakage (i.e.,the brighter disk is leaked photosphere), no?


I am more concerned with his choice of words here. That band is indeed the chromosphere seen in profile (ie: from the side). It is composed mainly of spicules and short fibrils or filaments and is not an "artifact" of the photosphere but a very real feature. In a narrower filter when tuned correctly, you can still see that fringe, although it will be notably less easy to see in a distinct way from the limb in the narrower filter, especially if the filter is tuned to the very centerline wavelength. The photospheric "leakage" is really continuum leakage and visually is better expressed by the lower contrast of the disk detail rather than the presence or absence of the fringe on the edge of the limb. Thus, calling it an "artifact" isn't really correct. Clear skies to you.

#15 Spectral Joe

Spectral Joe

    Vostok 1

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

Posted 26 December 2012 - 12:53 AM

A thread from nearly a year ago that discusses the issue, with some references: http://www.cloudynig...5057295/page...
The "band" is indeed the chromosphere, the "artifact" is the smaller disk, which is continuum light leaking through. It's hard for people to accept that their $6000 solar telescope doesn't "pass only the light of hydrogen alpha" like the advertisements say, but that's the way it is. A single etalon system, unless a much better (and more expensive) blocking filter is used, will pass enough contimuum to be noticed. Lunt, Coronado, and others know this. These instruments are marketed to the amateur community, who for the most part can't afford the equipment that has the top performance in this regard. The cost differential is large. When these etalon based systems were conceived I don't think anyone thought about high dynamic range CCD cameras and powerful image processing software being applied to them. Once these techniques are applied the faults become apparent. Never fear, you can always double stack. For more money, of course.

#16 George9

George9

    Messenger

  • *****
  • Posts: 424
  • Joined: 11 Dec 2004

Posted 26 December 2012 - 08:00 AM

That is a great thread. And yes that's what I meant by artifact. Thanks so much. George

#17 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15467
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 26 December 2012 - 09:42 AM

A thread from nearly a year ago that discusses the issue, with some references: http://www.cloudynig...5057295/page...
The "band" is indeed the chromosphere, the "artifact" is the smaller disk, which is continuum light leaking through. It's hard for people to accept that their $6000 solar telescope doesn't "pass only the light of hydrogen alpha" like the advertisements say, but that's the way it is. A single etalon system, unless a much better (and more expensive) blocking filter is used, will pass enough contimuum to be noticed. Lunt, Coronado, and others know this. These instruments are marketed to the amateur community, who for the most part can't afford the equipment that has the top performance in this regard. The cost differential is large. When these etalon based systems were conceived I don't think anyone thought about high dynamic range CCD cameras and powerful image processing software being applied to them. Once these techniques are applied the faults become apparent. Never fear, you can always double stack. For more money, of course.


I so dislike the use of the terms "artifact" and "leaking through", as they imply that the filtering system isn't working properly when, in fact, it is. The passband shape of an etalon-based filtering system isn't infinitely sharp and the H-alpha line isn't square-wave sharp either. In fact, there is even just a little continuum energy from the sun at 6562.8 angstroms that is not being filtered out. What is happening here is purely a function of the passband width of the filter. The narrower the bandwidth, the greater the contrast of the disk detail tends to be. This is a combination of letting less continuum light in as well as screening out some of the slight Doppler shifted light detail. The fibril and filament disk structure tends to seem darker and a little sharper in narrower filters than in broader ones. It's just that simple. If you have a "leak", this implies that something well-off the primary passband's location is being let through, such as the unintended "leaks" found in some of the common nebula filters. This isn't the case in H-alpha solar filters. The primary passband is not leaking anything (it is just wider or narrower depending on the system's specifications or cost). As those with the old Daystar filters have sometimes found out, once the blocking or trimming filters start breaking down and actually "leaking" some of the other passbands of the "comb" created by the etalon, the H-alpha detail just goes away, so even a slight "leak" will kill the view and make it into just a very red version of the white light view. If you are seeing any chromospheric disk detail, the filter isn't really leaking anything. The same goes with calling some feature an "artifact", which in imaging and optical circles tends to mean something induced by the system that is not really there. This is not the case for the view of the spicule forest. It is a very real feature. Slight tuning variances can make it appear more distinct even in rather narrow filtering systems, so just because it is seen does not mean that this is some "artifact". If you want greater contrast, a double-stacked etalon can sometimes provide that, which is the plain and simple truth. Clear skies to you.

#18 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8581
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 26 December 2012 - 02:19 PM

An interesting read indeed chaps, thanks! :)

#19 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 26 December 2012 - 04:13 PM

Which actually brings up a separate question about double stacking. Is the improved contrast from double stacking really due to the bandwidth going from 0.7 to 0.5 (the half-height width), or is it really due to changing the shape of the transmission curve from Gaussian to something with squarer shoulders and therefore lower tails? (That is, is most of the leaked energy right near the centerline, or is it spread further away in the tails?) I remember going over this with David Lunt in 1999 or 2000, and we did the math and it seemed to be the tails, not the nominal bandwidth. So a 0.7 A filter with sharp shoulders would also have noticeably better contrast than a regular 0.7 A filter.



For reference, I thought I'd post an explanation on how double-stacking reduces bandwidth from David Lunt, original founder of Coronado Filters:

The result of two identical etalon filters in series is a convolution of the transmission bands of each. The single etalon has a passband shape which is Gaussian. If the bandwidth at 50% of maximum transmittance is w, then that at 10% of Tmax is 3.5w and that at 1% Tmax is 10w. The transmittance at any point in the spectrum of the stacked pair is T squared, where T is the transmittance of the single filter. The most important characteristic is that the bandwidth is reduced by the square root of 2. Given two etalons with bandwidths of 0.7A, the combined bandwidth becomes 0.5A, and the 1% bandwidth (or the "tails" of the passband) are reduced from 7A wide to ~1.8A. 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. Empahsis added.

Also see this discussion:

http://www.cloudynig...3435927/page...


I so dislike the use of the terms "artifact" and "leaking through", as they imply that the filtering system isn't working properly when, in fact, it is.

Sorry if I have offended anyone. I did not mean to imply that by use of the terms “leaking through” or “artifact” that a 0.7 FWHM filter is not working properly. Rather only that the filter system is indeed letting more off-band photospheric energy through, degrading contrast, and showing features that are not really a property or part of the chromosphere – such as more “spot detail,” as well as the “double limb” of the photosphere.

In the 50’s and 60’s there was some scientific debate as to whether or not the “double limb” was a true feature of the chromosphere (see the link in Spectral Joe's post). The “double limb” phenomena was determined to be an “artifact” of “parasitic” photospheric light contamination by blocking filters with insufficient narrowness to block side band harmonics. The double limb was determined in reality to be the border of the photosphere’s disk, and not a property or phenomena of the chromosphere. The double limb also appears identical with bandpasses over 0.5 A.

In a narrower filter when tuned correctly, you can still see that fringe, although it will be notably less easy to see in a distinct way from the limb in the narrower filter, especially if the filter is tuned to the very centerline wavelength... This is not the case for the view of the spicule forest. It is a very real feature. Slight tuning variances can make it appear more distinct even in rather narrow filtering systems, so just because it is seen does not mean that this is some "artifact".


I think this may be an issue of semantics - a narrow band filter eliminates the light from the wavelengths adjacent to the H alpha emission line, and therefore the boundary between the photoshpere and chomosphere disappears. Contrast is improved, and non-doppler shifted prominences are unaffected and as easily visible as at a wider bandpass.

There is no hint of the double limb, "fringe," or a separate “spicule forest” in my system when properly tuned and optimized at < 0.5 A . Of course if it is tuned off-band, more light form the photosphere will appear, showing the double limb artifact, or 'effect' if you prefer. By tuning a narrow band filter off band, one will of course introduce more continuum light, revealing the edge of the photosphere, and thereby producing the "spicule layer," or "fringe." However, in reality it is still the entire chromosphere (including spicules) on edge delineated by the now revealed photospheric boundary.

That the area of the chromosphere shown between the inner "double limb" shown at 0.7 A – the limb of the photosphere – and the remainder of the chromosphere lying beyond it, is sometimes mistakenly referred to by some as the "spicule layer/fringe" is understandable, but incorrect. It more accurately represents the entirety of the chromosphere lying above the photosphere (see diagram in the following post). Spicules themselves – features of the chromosphere - rise from the base of the chromosphere and penetrate the transition layer into the corona. These can be seen in the < 0.5 A double stacked animation link below just to the left of the surge prominence.

Again note there is no hint of the double limb from on-band light from the photosphere - or off-band light adjacent to the H alpha line - coming through, as it is far below the threshold (S/N ratio) needed for detection. Remember cooler Hydrogen in the photosphere is absorbing this wavelength far in excess of what the chromosphere is emitting, so with a sufficiently narrow bandpass the signal to noise ratio has rendered any off-band photospheric contibution invisible, and any on-band contibution from the photosphere is so weak as to also be invisible.

GIF https://dl.dropbox.c...mation2 gif.gif

AVI https://dl.dropbox.c... Animation1.avi

Attached Files



#20 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 26 December 2012 - 05:10 PM

Another recent CN post concerning double stacking with good representations of the effects of narrower bandpasses can be found here:

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

Note particularly the single, double, and triple stacking comparison images by Jesus...


Representation of the chromosphere and spicules courtesy of NASA:

Attached Files



#21 Bill Cowles

Bill Cowles

    MacGyver

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

Posted 26 December 2012 - 11:08 PM

Stack... :smirk:

Bill

Attached Files



#22 Andy Devey

Andy Devey

    Mercury-Atlas

  • -----
  • Posts: 2775
  • Joined: 06 May 2011

Posted 27 December 2012 - 11:35 AM

Hi Guys

This really is an excellent and informative thread many thanks to you all

Regards

Andy

#23 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 15467
  • Joined: 19 Apr 2004
  • Loc: southeastern Nebraska

Posted 27 December 2012 - 03:40 PM

BYoesle wrote:

That the area of the chromosphere shown between the inner "double limb" shown at 0.7 A – the limb of the photosphere – and the remainder of the chromosphere lying beyond it, is sometimes mistakenly referred to by some as the "spicule layer/fringe" is understandable, but incorrect.


No, it is not really all that incorrect. That is what the feature has been called in the professional literature (spicule forest, "sierra", "burning prairie", ect.). To cite Prof. Harold Zirin (Astrophysics of the Sun, p. 161-169):

Fig. 7.5 caption: "Spicules at various points in H-alpha. In the line center (0 Angstroms) we see the spicule forest extending up to 7000 km, marked at lower heights by the general chromosphere. Beyond +/-1/2 Angstroms, the narrow emission of the general chromosphere disappears, and only the broad-line spicules remain."

Text on page 169: "Although spicules appear to form a thick forest in H-alpha pictures at the limb, Cragg, Howeard and Zirin (1963) recognized that they occur only at the edges of the magnetic network and subsequent pictures confirmed this fact. In the center of the disk, the spicules are seen to protrude in all directions from the network elements, tracing the connecting flux loops. The lower velocity of the IN (intra-network) structure produces a fairly narrow Doppler profile. Thus this widespread structure, which may be thought of as the general chromosphere, is not seen further from the H-alpha line center than +/- 0.6 angstroms. At the limb (Fig. 7.5) spicule and IN chromosphere are all mixed up. In centerline H-alpha there is a fairly continuous band about 7000 km high: since this disappears at +/-1/2 angstroms, we identify it with the IN chromosphere. Outside this wavelength only spicules are seen on disk or limb. Although long exposures with a coronagraph will show spicules extending up to 10000 km the fact that most spicules are tilted means that they are no higher than the IN chromosphere. Below 1500 km there is little structure visible in centerline, but at H-alpha +/- 0.3 angstroms, a dark band, due to a minimum in the Doppler broadening, can be detected."

In imaging circles, a visual "artifact" is defined as "anomalies during visual representation of e.g. digital graphics and imagery". The fringe-like feature seen in H-alpha filters on the limb of the sun is not an artifact. Its appearance will vary depending on where the filter is tuned and its FWHM bandwidth, but it is a real feature. Clear skies to you.

#24 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4863
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 27 December 2012 - 07:39 PM

Hi David,

Thanks for the excellent citations you have referenced (I actually own and have read Zirin’s books as well.) No one has denied the presence or visibility of spicules in the chromosphere either at 0.7 or 0.5 A FWHM. Yet it is the entire chromosphere that is being shown by the presence of the photosphere at 0.7 A FWHM, not just spicules, spicule forest, ect.

I did specifically note in the next sentence following the quote above: “It more accurately represents the entirety of the chromosphere lying above the photosphere (see diagram in the following post).”

And obviously it is real.

I have not stated “the fringe like feature” on the limb of the sun is an artifact. I have stated the inner limb at 0.7 A FWHM is that of the photosphere, and in contradistinction, it is this photospheric light and limb that I and others have referred to as an artifact - not the chromosphere, or it’s components such as spicules, prominences, etc.

Initially it seemed you were referring to the layer above the photosphere (and perhaps still are) in my 0.7 A image as being more or less just the spicules: “the Chromospheric ‘fringe’; the mass of fine spicules that, depending on how the filter is tuned, can show up as a thick band (sometimes called "the Spicule Forest").” While this description “is not really all that incorrect,” it is incomplete. The layer simply is the entire chromosphere - including spicules, prominences, etc.

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.

I’m not trying to argue about semantics... I’m trying to clarify for CN’ers what they are seeing or imaging, what double stacking verses single stacking will show or not show, and to deal perhaps with expectations of the equipment they have and how it will perform.

Best wishes,

Attached Files



#25 George9

George9

    Messenger

  • *****
  • Posts: 424
  • Joined: 11 Dec 2004

Posted 27 December 2012 - 11:39 PM

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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics