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

Not all filter bandpasses are created equal.

This topic has been archived. This means that you cannot reply to this topic.
44 replies to this topic

#1 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 19 October 2013 - 10:40 AM

In a previous thread, the effects of double stacking H alpha filters were extensively discussed:

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

I have wondered whether the effect of double stacking two filters, which narrows the filters bandpass, is equivalent to having a single filter with the same bandbass (FWHM), such as the DayStar and Solar Spectrum filters.

As can be seen in the diagram below, George9 demonstrated that the passband curve for identical FWHM filters can vary appreciably (all curves have a FWHM bandpass of 2 Å). From the discussion above, we know that double stacking filters not only decreases the FWHM bandpass, but also considerably steepens the filter transmission shape.

I have recently viewed several images taken with narrow band – 0.5 Å FWHM and narrower – sub-Angstrom Ha filter systems from DayStar and Solar Spectrum, all of which show the “double limb” of the sun. This seems to demonstrate that the transmission shape of these filters is not as steep as the equivalent FWHM double stacked filters:

Erio Inglante Rossi (using 0.5 Å Solar Spectrum filter):

http://s2.www.astron...ssets/2-sun.jpg

Philippe Tosi (0.5 Å DayStar filter):

http://www.cloudynig...rd=Imaging&a...

Fred Bruenjes (using 0.45 Å DayStar filter):

http://www.daystarfi...p?/8/category/5

Richard Antal (using 0.3 Å DayStar filter):

http://www.daystarfi.../897/category/3

From what I have seen, it appears one must go to < 0.2 Å with a single filter to get a similar result (even at that FWHM, a double limb can still be identified):

http://www.cloudynig...199014/Main/...

Thus it can indeed be seen that most of the benefits of double stacking come from the steepening of the transmission curve and the suppression of the “tails” of the filter transmission shape – which eliminates side-band light from the photosphere from leaking through the filter system.

In this representation, th blue curve is a single stacked 2 Å filter, and the red curve is a double stacked 2 Å filter:

Attached Thumbnails

  • 6146510-Compare curves.jpg


#2 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 19 October 2013 - 10:40 AM

With double stacking two filters, the transmission of light through the “tails” of the filter curve is curtailed -- the photosphere is effectively extinguished -- and therefore there is no double limb at < 0.5 Å with this filter arrangement:

Attached Thumbnails

  • 6146511-0.7 v 0.45 A compare jpg sm.jpg


#3 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 19 October 2013 - 10:56 AM

OOPs I put this in the wrong forum, could the mods please move it - thanx. :foreheadslap:

#4 George9

George9

    Surveyor 1

  • *****
  • Posts: 1613
  • Joined: 11 Dec 2004

Posted 19 October 2013 - 04:04 PM

Bob, very clever to look at those images to answer the question.

My question was going to be whether the air-spaced etalons have a different profile than the solid ones. The front-mounted air-spaced Lunt etalons have a single filter element and have a Lorentzian curve (as I learned on CN in that great thread). For the solid etalon, I am not sure if it is a single solid element or a stack of wider solid elements in series. If it was a stack of elements, then the profile should be Gaussian with sharper shoulders like (or better than) a DS.

Your images imply that even a solid filter is not a stack but a single Lorentzian.

Years ago, David Lunt implied that in fact it was the tails and not the center width that made the difference, and what you found confirms it.

George

#5 Tom and Beth

Tom and Beth

    Skylab

  • -----
  • Posts: 4180
  • Joined: 08 Jan 2007

Posted 19 October 2013 - 04:28 PM

Perhaps it's my monitor, but not only do I see a lack of a "double limb " in the lower comparison, but the surface is more detailed. I would not have expected that with two systems each passing 0.5 Å.

Interesting.

#6 George9

George9

    Surveyor 1

  • *****
  • Posts: 1613
  • Joined: 11 Dec 2004

Posted 19 October 2013 - 08:34 PM

Here is what I get in terms of how much of the signal you see in the eyepiece is from outside a given bandwidth given two filters. Both are 0.5A FWHM, but the first is SS .5A and the other is DS .5A.

width - SS - DS
.5A - 50% - 35%
1A - 29% - 11%
2A - 16% - 2%
4A - 8% - 0.3%
8A - 4% - 0.04%

So a SS 0.5A filter has 16% of the visible energy coming from outside of a band 2A wide centered on H-alpha. DS has 2%.

(Let me know if I messed up.)

George

#7 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 20 October 2013 - 01:33 AM

Sounds about right George -- but this is not my area of expertise...

Perhaps it's my monitor, but not only do I see a lack of a "double limb " in the lower comparison, but the surface is more detailed. I would not have expected that with two systems each passing 0.5 Å.

Interesting.


You are correct T&B, but the top image is that of a single 0.7 Å FWHM filter -- obviously showing the double limb, as well as decreased contrast due to the light from the photosphere. I included it to show how the other images with narrower FWHM bandpass filters show the same feature, but that a pair of double stacked 0.7 Å FWHM filters -- with a resultant < 0.5 Å FWHM (bottom) -- does not.

_________________________________

On the other hand -- :john:

It occurred to me that another reason that the images taken through the DayStar and Solar Spectrum filters may show a wider bandpass is that their optical systems may not be fully optimized: Etalons placed on the front of the telescope are in an ideal position for optimum performance -- no instrument angles, and the minimum field angles. Internal etalons are usually configured with specific collimation optics matched to the objective to keep instrument angles and field angle magnifications to a minimum.

However, most rear filters are used with telecentric lens systems designed for a specific focal length objective -- 800 mm for the Baader Telecentrics -- or an unknown objective FL for "telecentric" barlows (TV Powermates) which do not feed the etalon true telecentric light bundles required by a narrow band etalon unless they have the designed-for objective focal length. This may cause a broadening of the filters bandpass.

David Lunt addressed some of these issues in a 2001 posting, which may give a partial answer:

If the filter is positioned near the focus of the instrument, different parts of the Solar disc are imaged by different parts of the filter causing the above described filter variations to be imparted on the image. It is commonly considered that such variations across the field are solved by the use of a telecentric lens or system which will render all the rays parallel through the filter. This is incorrect. A telecentric component only converts the primary rays in the off-axis bundles parallel. The bundles traversing the filter still have the focal ratio of the system and are still specific to a portion of the image ...

It is for the above reasoning that we position our filters as far from the focal plane as possible or position them within an afocal [collimator] lens system. The downside of moving away from the focal plane is a consequent necessary increase in the aperture of the filter. However, with the construction that we use, the increase in aperture is not as much of a problem as trying to solve the problems associated with proximity to the focal plane. The afocal configuration is far superior to the telecentric one when it comes to maintaining the performance of the filter. This mode does result in parallel rays traversing the filter, only limited by the field of view corresponding to the angular acceptance of the filter; - a better condition than having no part of the field of view imaged at the nominal spectral resolution of the filter due to the cone of the focal ratio. (Emphasis added.)


Therefore, the appearance of the double limb (e.g. the widening of the bandpass) may not be the result of the filter transmission shape -- it could be more due to the optical system used: The narrower the bandpass of the filter, the narrower the focal ratio cone would have to be in order to achieve the specified bandpass, even with telecentric lenses optimized for the focal length of the objective...

To the best of my knowledge, here’s how these issues are dealt with for narrowband H alpha filter use:

A narrow band etalon will only perform well if the angles - both instrument and field angles - fall within the “acceptance angle” of the filter for it’s specific bandpass. The narrower the filters bandpass, the smaller the acceptance angle will be. If these angles exceed this critical value, the filters performance will suffer - parts of the image, or the entire image, may fall off-band as the bandpass is broadened by the larger ray angles passing through the filter.

In the diagram below, the top diagram shows an etalon placed in front of the objective. There are no instrument angles, and the field angle (at the Sun’s limb) is that subtended by the Sun itself - 0.25 degree. This is the best that can be done with respect to angles passing through an etalon employed for solar use.

The middle diagram shows a collimator system, where the collimator is optimized to keep the field angles within the acceptance angle of the etalon for it’s specified bandpass. This generally requires the collimator to have about half the objectives focal length, and etalon about half the objectives diameter. If these general conditions are not met, the filter can fall off-band the farther you get off the optical axis, and a “sweet spot” is developed.

The bottom diagram shows a telecentric configuration. Here again the telecentric lens system should 1): be specifically designed to achieve telecentricity for the objective’s focal length (if not there will be no telecentric postition for the etalon - see Baraff: http://home.comcast....erving/Tele.pdf ), and 2): the resultant light cone angles (instrument angles) should fall under the acceptance angle for the bandpass specification of the etalon. If this condition is not met, the etalon's bandpass will broaden for all parts of the image -- and the filter will give less than the specified bandpass -– a 0.2 to 0.5 Å FWHM filter may actually perform no better than a 0.7 Å filter, as possibly seen in the images linked to above.

If neither condition is met, one might have both sweet spot issues and bandpass broadening.

Attached Thumbnails

  • 6147799-Filters and angles.jpg


#8 Peter Z

Peter Z

    Mariner 2

  • -----
  • Posts: 284
  • Joined: 03 May 2010

Posted 20 October 2013 - 12:59 PM

Hi Bob.
Your conclusion about the contribution from the wings of the bandpass profile is very sensible and (as owner of a Daystar) a little disheartening! I've gone to some lengths to design and model a telecentric focal extender specifically for my objective (100mm f/12) to give me f/34 at the filter. According to the ray tracing program I use (WinLens3D), the chief ray angle through the filter is exactly zero and the ray cone consistent with f/34 (Daystar recommends f/30). I always observe the double limb but attributed this to the (nominal) bandpass of the Daystar as 0.6 Angstroms.

It looks like lack of telecentricity might not be the cause of the problem but, as you suggest, leakage from the continuum in the wings of the profile. .... Darn!
The next thing to do is try f/48 at the filter and see what happens with the tighter cone.

I'm also experimenting with a double-pass arrangement through the Daystar (to tighten up the wings and reduce the bandpass a little) but my first attempts have been unsuccessful.

Cheers.
Peter.

#9 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 21 October 2013 - 12:32 AM

Hi Peter, I'm impressed that you formulated your own custom telecentric lens system. You should post it on the ATM/Optics/DIY forum and describe what your process of optimization was. I think this would be very useful to others to gain the most from their DayStar or Solar Spectrum filters.

F48 should rule out the issue of optics vs. bandpass. Unfortunately I'd have to agree with the postulate that it is the wings letting the continuum through, especially if the transmission shape is that of a single filter vs. a multiple filter stack per George's diagram.

#10 Montana

Montana

    Soyuz

  • -----
  • Posts: 3659
  • Joined: 02 Jan 2010

Posted 22 October 2013 - 06:52 AM

Depends on your blocking filter as well :(

Alexandra

#11 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 22 October 2013 - 09:53 PM

As Alexandra states, a problem with a blocking filter can also manifest itself with the same appearance of a widened bandpass and a showing of a “double limb.”

As can seen below, an H alpha etalon is actually a “comb filter” that not only passes the intended wavelength, but harmonic multiples of this wavelength as well. The blocking filter acts to block these harmonic peaks. If the blocking filter (also an interference filter) has too wide a bandpass, or if the blocking filter’s center wavelength is shifted to one side (such as can be done by tilting/blue shifting), some of the energy from the adjacent harmonic peak(s) can be let through. This would be off-band continuum light, and will have the identical effect and appearance of showing the photosphere and the double limb.

Alexandra, so sorry to hear you’ve had the blocking filter issue...

Attached Thumbnails

  • 6152657-Etalon peaks & BF.jpg


#12 ValeryD

ValeryD

    Vendor (Aries)

  • ****-
  • Vendors
  • Posts: 1931
  • Joined: 26 Nov 2005

Posted 23 October 2013 - 02:40 AM

As were said before, the band pass shift in a blocking filter - due to high angle of rays from the edge of the aperture, has harmful influence on Ha details contrast due to leaking of continuum light from other (nearest blue one) harmonic peak of etalon transmission.

I have experimented with different F/numbers in my 150mm solar telescope.

It has two refocusing objectives (exchangeable) F/5 and F/7. The F/5 system provides the brightest images, but the lowest contrast. F/7 is already significantly better, but still not perfect. The real good contrast starts from system F/10-12. For large scale photography I routinely use F/12 and for prominences F/5 to F/8 - depends of desired scale and prominences brightness.

Even after F/12, say at F/15, contrast continues to increase. And, the double limb is the most pronounced at F/5.




Valery.

#13 David Knisely

David Knisely

    Hubble

  • *****
  • Posts: 16971
  • Joined: 19 Apr 2004

Posted 23 October 2013 - 06:41 PM

BYoesle wrote:

As Alexandra states, a problem with a blocking filter can also manifest itself with the same appearance of a widened bandpass and a showing of a “double limb.”


The term, "double limb", isn't really accurate. It is the "spicule fringe" that is appearing, so such a description isn't a real defect in the filter or system (other than having a passband that is broader than it might be). As for the blocking filter having problems, that causes a dramatic drop in contrast across the entire field, eventually resulting in little to no H-alpha detail being seen. The DayStar systems (solid spacer crystal between the etalon plates) used both a blocking and trimming filter (i.e., a low-pass and a high-pass filter) to get rid of the unwanted "comb" passbands. When they began to fail in my 0.7 angstrom DayStar T-Scanner, the H-alpha detail quickly vanished. After about 15 years, the older ones needed to be replaced, although I understand that the newer coatings tend to be somewhat longer-lived. Clear skies to you.

#14 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 24 October 2013 - 12:44 AM

Valery: I seem to recall David Lunt stating that when configured normally the blocking filter would perform best at f10 or greater:

...the focal ratio ideally should be greater than f10 to maintain performance of the blocking filter....but in fact you will see good results at considerably shorter f ratios than this.


but also stated:

The focal ratio is not critical. If one was using the blocking filter as a primary filtering element, it can be used without compromise to F/8. However, in the Solarmax system, it is only used as a blocking element for the main filter. In these circumstances it can be used to about F/4 without any problem.


What you may be observing is the effects of using the blocking filter in the refocused light cone fed by the collimator, changing the ability of the blocking filter to work at the traditional f ratios found on an front mounted filter system.

___________________________


The term, "double limb", isn't really accurate. It is the "spicule fringe" that is appearing, so such a description isn't a real defect in the filter or system...


David: I think there is a misunderstanding as to what constitutes the “double limb,” the “spicule fringe,” and also what is going on with Alexandra’s blocking filter.

The double limb as I refer to it has little to do with the “spicule fringe” or “spicule layer.” Rather it is continuum light from the photosphere intruding on - and adding to - the light from the chromosphere, both reducing the contrast of details of the chromospheric disk, as well as demarking the edge of the photosphere beneath the entire on-band chromosphere. This is easily seen in the above comparison images with a 0.7Å FWHM (top) and < 0.5 Å FWHM double stacked filter system (bottom). Indeed, one can verify this is light from the photosphere by the increased detail seen in the small sunspot group just off center on the 0.7Å filter system image, and the decreased prevalence of the same group in the narrower bandpass filter system.

The “spicule fringe” as you refer to it seen in that 0.7Å FWHM image is actually the entire on-band chromosphere, including prominences (which are quite obvious), surges, and macrospicules, etc. Note the features are nearly identical to the < 0.5Å filter system - without the double limb of the brighter photosphere intruding. Spicules themselves are fast-moving Doppler-shifted features. While the “spicule layer” or “fringe” is part of the chromosphere, it is better seen when a filter system is tuned off-band to the red (~ 656.36 nm) or blue (~ 656.2 nm) wing of the H alpha line, as seen in the de-tuned < 0.5Å FWHM filter system image below: the prominences of the chromosphere have disappeared, and the spicule layer has appeared in the off-band blue-wing image (bottom). The “double limb” of the photopshere is again seen to emerge due to the intrusion of light from the photosphere coming from the wing of the filter transmission passband, which is blue shifted by tilting the filter system off-band.

Of course, the broader the bandpass of the filter system, the more these off-band spicules will be included in the overall view of the chromosphere.

... 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


Alexandra has two blocking filters to use with her double stacked 0.5Å filter system, one of which apparently allows side band light from the photosphere through. The DS blocking filter that apparently is letting the continuum light through is not deteriorated, as the detail looks quite good: http://www.flickr.co...ra4/9597964395/ . Rather it likely is letting harmonic peak continuum light through the filter system, as described above, which exactly mimics the image from a wider bandpass filter of ~ 0.7Å FWHM. Since a blocking filter's purpose is to block the harmonic side band peaks of the primary etalon(s), I believe this would be considered a defect, and not deterioration.

Top: < 0.5Å filter on-band. Bottom: < 0.5Å filter off-band showing the actual spicule layer ("fringe") and the photospheric limb:

Attached Thumbnails

  • 6154735-on off proc jpg sm jpg.jpg


#15 Peter Z

Peter Z

    Mariner 2

  • -----
  • Posts: 284
  • Joined: 03 May 2010

Posted 24 October 2013 - 12:14 PM

Hi Bob.

Thanks for initiating a very interesting thread and for that very informative last post! It alleviated my own confusion about what constitutes the spicule fringe and and the double limb.

By the way, I think you'll be interested in the attached diagram if you aren't already aware of it. It's a transmission curve for various configurations of interference bandpass filters from the Melles Griot catalogue. The drawing shows curves for 1,2,3 and 4 cavity filters - essentially equivalent to single, double, triple and quadruple stacked single-cavity filters. The results totally support your argument. (Hope I'm not infringing any copyright laws by posting it.)

Cheers.
Peter.

Attached Thumbnails

  • 6155545-Melles Griot Filters.jpg


#16 mhilscher

mhilscher

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 13 Jul 2011

Posted 24 October 2013 - 02:26 PM

wow, what a deep post. thank you contributors

#17 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 25 October 2013 - 08:45 AM

:waytogo:

Hi Peter,

I had not seen that particular diagram -- really nice to have another reference.

George produced a similar (non-logarithmic) set of curves for the post cited above. It shows normalized transmission curves for a single 0.8 Å FWHM filter (dark blue), single 0.7 Å FWHM (red), 0.7 Å double stacked (e.g. ^ 2 @ 0.45 Å FWHM - green), followed by triple (^3 @ 0.35 Å FWHM) and quadruple stacked (^4 @ 0.30 Å FWHM) transmission curves, respectively.

One can see that the most significant effect is the reduction of the filter curve transmission "tails" - which greatly suppresses side band energy getting through the filter system. The biggest "bang for the buck" appears to be double stacking, both with regard to bandpass reduction and tail suppression.

Note also that because the curves are normalized, they do not show the reduction in overall transmission (peak intensity) that also occurs with stacking multiple etalons...

Attached Thumbnails

  • 6157066-Compare curves2.jpg


#18 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9006
  • Joined: 19 Aug 2008

Posted 26 October 2013 - 05:31 AM

A great thread as always Bob, thanks! ;)

#19 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 26 October 2013 - 12:00 PM

:thanx: Mark!

#20 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 21 December 2013 - 06:30 PM

There appears to be an answer as to whether the < 0.5 Å solid etalon filters used by DayStar (and perhaps Solar Specturm) use multi-cavity filters to achieve a narrower bandpass. From the following information posted on the DayStar web site (and a phone call) the answer seems to be “no.”

DayStar does not double-stack filters to accomplish narrower bandpass. The coatings on the etalon crystal dictate the bandpass of each filter...

A 0.4 [Å filter] is all about chromosphere. This specific bandpass shows the soft fuzzy boundary layer of the sun's chromosphere.


Emphasis added.


See:

http://www.daystarfi...iew/categori...

The picture for the 0.4 Å filter definitely shows the “double limb” of the photosphere.

Therefore the appearance of the double limb with <0.5 Å solid etalon filter appears to be related to the nature to the filter construction itself (a single cavity), as well as (at least in some cases) the lack of optical system optimization to obtain the necessary field and instrument angles.

On the other hand, we know that a double stacked (two cavity) air-spaced etlaon system with a <0.5 Å FWHM cumulative bandpass blocks continuum leakage that creates the appearance of the “soft fuzzy boundary layer” resulting from light from the photosphere. Double stacking suppresses the tails of a single cavity filter - e.g. increases the “steepness” of the filter systems cut-off, thereby reducing the transmission of out-of-band light. A single cavity filter such as used with a < 0.5 Å solid filter will not have the steeper transmission profile, and therefore the double limb form the photosphere shows itself.

The downside to the use of multi-cavity filters (i.e. multiple stacking) is the reduced overall transmission and reduced image brightness. If a single cavity filter has a transmission of 60%, a double stacked filter system of identical filters will have a transmission of 0.6 x 0.6 = 0.36, or 36%.

Given that most of the benefit of multiple stacking apparently occurs due to the suppression of the filter “wings” or “tails,” and not the incremental reduction of the FWHM, triple and quadruple stacking would appear to offer less benefit due to bandpass reduction, while incurring significant reductions in image brightness. For example a triple stack of 60% etalons: 0.6 x 0.6 x 0.6 = 22%, and a quad stack would be 13%.

Many (including myself) have assumed that the bandpass specification is an absolute criteria by which to judge a narrow-band solar filter system, and a reliable indicator of what the system performance - and the appearance of the sun - will be. This apparently is not the case. A more important criteria seems to be the filter system’s construction, and the resulting shape of the transmission profile.

Additionally, for the solid filters, the accessory optics for providing a telecentric light bundle trough the filter, and the narrowness of the light cone iteself from the primary optcs, play an important role in the filters performance.

The bottom line is that a FWHM bandpass measurement/specification is not a reliable indicator of how a solar filter system will ultimately perform “in real life.”

#21 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9006
  • Joined: 19 Aug 2008

Posted 22 December 2013 - 04:27 AM

The bottom line is that a FWHM bandpass measurement/specification is not a reliable indicator of how a solar filter system will ultimately perform “in real life.”


I completely agree with you Bob. Great piece of detective work and a great thread. The question now is how can we apply this to designing our own filter systems...

#22 pbsastro

pbsastro

    Viking 1

  • *****
  • Posts: 671
  • Joined: 21 Mar 2007

Posted 23 December 2013 - 11:47 AM

The bottom line is that a FWHM bandpass measurement/specification is not a reliable indicator of how a solar filter system will ultimately perform “in real life.”


I completely agree with you Bob. Great piece of detective work and a great thread. The question now is how can we apply this to designing our own filter systems...


The problem is double stacking does not seem to work very well with two rear mounted filters, due to ghosting from the filters being so close to each other. And the use of a front mounted filter is limited to 90-100mm which is too much limited, not only on resolution but also on light needed for high magnification.

Two years ago, I considered going for a rear double stack with a LS152 and a regular air-spaced rear filter (from a smaller Lunt or Coronado scopes) which have lenses of about 190mm focal, compared to the 350mm of the LS152 and LS100, so about 160mm separation between the filters.
In the end I decided to go for a 0.3A SolarSpectrum which I am still waiting delivery after 2 years.

I have a Coronado front-mount 90mm in a TV102 and a 40mm rear-mount that I use in my 150 and 175mm refractors. Sometimes I use the 40mm as double stack with the 90mm, and as you say the photosphere (double) limb disappears. The image is nice at 0.5A but dim, which limits magnification. I much prefer the high mags (240x-400x), even at 0.7A and with photosphere intruding. But that may be because I seem to prefer proms and limb over disk. From what has been talked here, the 0.5A SS may have less detail/contrast on disk than the 0.5A DS.

Pedro

#23 George9

George9

    Surveyor 1

  • *****
  • Posts: 1613
  • Joined: 11 Dec 2004

Posted 23 December 2013 - 05:05 PM

Thanks for finding that, Bob. It makes sense.

I completely love my LS80 DSII, but I do wish I could go to higher mag when seeing warrants. I'll have to get a second scope someday.

George

#24 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 9006
  • Joined: 19 Aug 2008

Posted 24 December 2013 - 02:57 AM

I've tried repeatedly over the past year to double stack a pair of 20mm rear mount etalons: To get them both 'on band' to get them to give a double stacked view their outer surfaces are virtually parallel and there is a ghost image of the sun overlapping / superimposed on the true image. If you tilt the etalon to throw the ghost then you get banding effects visible because the etalon is tilted too much. I also tried putting a circular polariser between the two etalon sets but this dimmed an already dim image and when I tried imaging with this setup also showed rather annoying interference effects. I think with the right polariser - designed for Ha transmission there may be some mileage, but i've yet to find the right polariser, and i've tried a few...

#25 BYoesle

BYoesle

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6577
  • Joined: 12 Jun 2004

Posted 24 December 2013 - 07:12 AM

Hi Mark, Pedro, and George.

The question now is how can we apply this to designing our own filter systems...


Based on my experience and what I have learned and seen from others, I have come to the tentative conclusion that for optimum H alpha viewing and imaging, a two telescope solution is necessary for doing both full disk and hi-res viewing/imaging, and that these should ideally be equipped with multi-cavity filters:

For full disk viewing and imaging, the best solution is a double stacked objective mounted etalon system of as large an aperture as can be afforded. This provides highly uniform disk contrast at < 0.5 Å FWHM bandpass with the steeper transmission shape of a two cavity filter system needed to eliminate the “double limb,” which is created by parasitic light from the photosphere getting through the filter system via the transmission tails of a single cavity filter.

For higher resolution viewing and imaging, greater aperture is needed than can be reasonably obtained with front etalon(s). Here the optimum would seem to be an etalon(s) placed in a collimator lens system, unless one can design and fabricate a telecentric lens for their particular objective’s focal length. However, the telecentric system will still be constrained by the FWHM spec of the etalon for a very narrow (e.g. a f30 to f60) light cone to reach its theoretical bandpass, and even with that, may still pass continuum light due to the use of a single cavity filter.

Using internal etalons with a collimator lens can be tricky, as Mark has discovered. Here field angles are magnified by the ratio of the collimator to objective focal length, and unless both the diameter of the etalon(s) and collimator focal length is properly sized you’ll get vignetting, sweet spots, or excessive banding – especially if one or both the etalons require overmuch tilting to bring them on-band and/or eliminate reflections. The general rule dictated by the geometry and field angle math is that the collimator needs a focal length at least half the objective’s, and the etalon needs to be at least half the objectives diameter to avoid excessive vignetting.

If double stacking an internal etalon system for use with a collimator, one would ideally use an etalon that requires little tilt to be on band for one of the pair, and a second etalon would be of the type that requires a bit of tilt to be on-band to remove the reflections, but not so much tilt as to introduce excessive banding. These seem readily available as dedicated double stacking front etalons.

My solution for full disk viewing and low and medium resolution work is already in hand - my double stacked ED100 SM90/90/BF30 system.

For the higher resolution work, my tentative solution will be to attempt to place this existing system into the optical path of a 150 mm diameter Newtonian based reflector system which uses a collimator lens (with an auxiliary field lens to reduce vignetting), with the existing 100ED telescope objective acting as the refocusing lens. This system meets all the aforementioned criteria needed to theoretically provide reasonable field angles, minimal vignetting, and < 0.5 Å two cavity filter performance (e.g. no double limb). A refractor based system did not work, as the collimator lens (a standard refractor objective) introduces significant spherical aberration, which may be elimated by adjusting the conic constant of the primary mirror (e.g. a "degree of freedom" not available with a refractor primary objective). Additionally, such a reflector could be used for other OTA’s used for Continuum and CaK, and therefore I think it more similar to a solar patrol telescope of the past, with many details yet to be worked out:

Attached Thumbnails

  • 6270050-3D layout labled 2 SM.jpg



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