Some more observations on filter performance

Use this section to discuss "standard" Baader/Coronado/ Lunt SolarView/ Daystar, etc… filters, cameras and scopes. No mods, just questions/ answers and reviews.
Post Reply
User avatar
Bob Yoesle
Im an EXPERT!
Im an EXPERT!
Posts: 473
Joined: Thu Mar 08, 2012 7:24 pm

Some more observations on filter performance

Post by Bob Yoesle » Mon Aug 24, 2015 4:05 am

As some of you know, I have been brash enough to state unequivocally "bandpass means nothing" when considering filter contrast performance, and "transmission profile ("finesse") means everything." http://solarchat.natca.net/viewtopic.ph ... 95#p166740 This position has come about over some time as I discovered the mysterious "double limb" observed and imaged by most single cavity filter systems was "parasitic" continuum light coming from disk of the photosphere. "Double stacking" essentially eliminates this continuum "leakage" - e.g. drastically improves filter system finesse - and usually far more than a single cavity filter of much narrower bandpass.

I first became aware of these issues a few years ago. Like many other non-professionals, I was quite uninformed as to filter theory, real performance, and chromosphereic features -- so when I first noticed these differences I thought there might be something wrong with my filter system, as it wasn't showing this often-imaged - and mis-identified - "spicule layer." All the more motivation to get to the truth of the matter.

Shown below are "normalized" single and double stacked transmission curves, and demonstrating the seeming irrelevancy of filter bandpass specification with respect to elimination of parasitic continuum, while at the same time showing the real difference lies between the filter system transmission "tails" where the continuum is actually leaking through:
Double stacking highlight.jpg
The identically processed images were also "normalized" to show equal brightness for prominence features, thereby clearly demonstrating the contribution of continuum to contrast degradation of disk features (and increased visibility of photosphere features - sunspots), and seemingly debunking the frequently-stated myth (even by the filter manufacturers) that a narrower bandpass (usually achieved via double stacking filters) decreases prominence details and visibility. This "decrease" in prominence visibility is apparently mostly due to the decreased overall brightness of the chromosphere, usually from the reduced transmission of a multiple cavity filter system or a narrower single cavity system.
H alpha single v double stacked sm Labled.jpg
H alpha single v double stacked sm Labled.jpg (287.2 KiB) Viewed 2727 times
Ken - in a recent post on digital heliospectroscopy verses filters - cited a reference which spelled out the situation clearly giving parameters to when this parasitic light form the photosphere begins to intrude into the chromospheric H alpha emission: http://solarchat.natca.net/viewtopic.ph ... 75#p167202
Bray and Loughhead, "The Solar Chromosphere" p18-19 talk at some length about Parasitic light.
"The absence of the photospheric limb on narrowband photographs free from parasitic light shows...that in the CORE of Ha the chromosphere is optically thick along a tangential line-of-sight to the limb"
"High resolution observations show that the photospheric limb first begins to re-appear at Ha +/_ 0.65A....As one moves further out from the line centre, the photospheric limb rapidly becomes more prominent and appears as a sharp boundary crossed at irregular intervals by isolated chromospheric features."
This is well demonstrated in their accompanying plates 2.8 and 2.10.
They also note the use of a 1A prefilter on their 1/8 A Halle filter suppressed the bandwidth and parasitic photosphere.
Last comment: The actual wings of Ha extend 8A on either side of the core. It is obvious that ANY off-band (red/blue wing) data will be only be obtained close to the core "shoulder" - at less than +/-1.25A and not out further in the extended wings which will give, with any filter/ SHG bandwidth, access to the photospheric continuum light.
Note the reference to DSing a 0.125 A FWHM filter (which apparently also shows a double limb) with a 1 A "prefitler," which increased system finesse enough to eliminate on-band parasitic photospheric intrusion. However, rather than "suppressed bandwidth," (highly unlikely with a 1 A pre-filter) it is more likely that suppressed transmission "tails" eliminated the parasitic intrusion of continuum form the photosphere into the chomospheric H alpha emission. Christian Valadrich has also established continuum suppression using a 4 A filter with 0.6 A DayStar system: http://www.astrosurf.com/viladrich/astr ... ntrast.htm

Thanks to this reference from Ken, I thought I would try and look to see what might actually be going on. First, there is the problem of using "normalized" transmission curves, which do not accurately represent the true situation when double stacking etalons, and the entire DS curve is stretched taller than it should be. Here, using normalized transmission values for SS verses DS, and the figures stated above for the +/- 0.065 nm limits for the intrusion of continuum, we have a value near 22% for single stacked transmission level, and about 8% for double stacked:
Double stacking data.jpg
However, using a reasonable assumption of about a 60% peak transmission value for a typical DS etalon system compared to a SS, we can see a number of interesting areas:
SS v DS non-normalize data 3 SM.jpg
SS v DS non-normalize data 3 SM.jpg (94.09 KiB) Viewed 2727 times
Note that the non-normalized DS transmission curve crosses the +/- 0.065 nm threshold at ~ 0.048 (4.8%) instead of ~ 8%. This has led to a significantly larger and more accurate representation of area covered by the difference in the two transmission profiles (especially adjacent to the +/- 0.065 nm limit), and perhaps showing a larger amount of flux difference permitted to intrude into the chromospheric emission by the SS profile.

The region labeled 1 indicates an area where at under the +/- 0.065 nm limit, the SS curve apparently does not contribute significant additional continuum to the chromosphere. Area 2 indicates the region where additional continuum can leak into the chromosphere emission. However, from this we need to consider that at the point where the DS curve crosses the +/- 0.065 nm limit - area 3 - there is apparently insufficient flux (area under the curve) to make a noticeable contribution of continuum to the chromosphere. This leaves areas 4 and 5 as the remaining regions to consider, and establishes that a transmission level above 0.048 (4.8%), continuum may be highly suspect to be leaked. Moreover, the extended region in the SS area 4 stretches way out in either direction into the continuum regions, and also has significant area. Ken notes that at +/- 0.125 nm beyond the "core shoulder," continuum will dominate, and as can be seen there is significant area in region 4 beyond this limit, whereas in the DS configuration, area 3 is well within this limit and would seem to have no significant contribution. These extended areas are where the flux from the photosphere becomes quite large, especially as there is overall about a 100,000:1 relationship between the brightness of the photosphere and chromosphere.

It would be interesting to analyze what parts of areas 4 and 5 actually contribute most significantly to "parasitic" continuum leakage (e.g. transmission height vs. wavelength) -- and see a more rigorous analysis from a mathematical perspective -- but that is not my forte - mostly a visual guy here... ;)
Last edited by Bob Yoesle on Tue Aug 25, 2015 3:39 pm, edited 5 times in total.
Diagonally parked in a parallel universe.

Curiosity is the father of knowledge; uncertainty is the mother of wisdom.

Dark-Sky Defenders

Solar B
Ohhhhhh My!
Ohhhhhh My!
Posts: 139
Joined: Mon Jul 27, 2015 9:29 am
Location: Dunfermline , Scotland

Re: Some more observations on filter performance

Post by Solar B » Mon Aug 24, 2015 8:22 am

Many Thanks Bob .... now whilst I do not completely understand all the terminology
that you discuss here (I am at least trying to) this is a "very" important lesson
for me personally .... meaning that in future I will not be perusing rear mounted
SS etalons but will revert back to DS front mounted options.
It does make me ponder as to the visual performance of eg Daystar filters at
0.7 - 0.6A ? when I'm currently using a supposedly <0.5A having tested two in number...however there is no doubt as to their imaging capabilities.

Brian
" Gentlemen only ever use Refractors "

Luke Stacy
Ohhhhhh My!
Ohhhhhh My!
Posts: 156
Joined: Tue Aug 19, 2014 1:25 pm

Re: Some more observations on filter performance

Post by Luke Stacy » Mon Aug 24, 2015 6:43 pm

Thanks, Bob, that is very interesting reading, I would love to understand much more about these filters than the very little I do know. Some of this is over my head but I will re-read it and try and get to understand it all. It may take me some time :lol:

I use a Daystar Quark and have used a bright and a dark Chromosphere. My very clear impression was that proms were even better in the bright one. That said, I had a dark Quark, much darker than the bightest Chromosphere, and proms were still very visible in it, though I would say they did not look quite as impressive as the bright one. Is it down to mind/brain/kidding myself?

I have two Quarks that I can image with. One is darker and requires longer exposure time than the other one so part of me would have thought that is the one to image with. However, pending final tests to double check (difficult with our poor summer here...), I find the brighter one actually performs better and shows more detail, so my conclusion was that the darker one presumably is narrower bandpass but bandpass alone perhaps does not indicate how well a filter performs.

Brian, what filter are you comparing the Quark to? I find it works very well for visual, I used SolarMax 60 single and double stack before and have looked though Lunt 60s. My wife has looked though SF70. We are both very impressed with the Quark visually. I love using it in my ED100 for visual where of course there is a big difference in aperture versus the SM60, and for closer up detail, aperture is king.

We both sold our SM60's.

Of course the dedicated h-a scopes have pros and cons just like the Quark so which is best will depend on needs, budget and preferences. I see a lot of good comments on Quark for imaging but to me I enjoy using Quark for visual just as much and I highly recommend it, especially if you like the idea of higher mag and finer detail on proms and active regions and a dedicated h-a scope or front filter in those apertures it out of budget.

Solar B
Ohhhhhh My!
Ohhhhhh My!
Posts: 139
Joined: Mon Jul 27, 2015 9:29 am
Location: Dunfermline , Scotland

Re: Some more observations on filter performance

Post by Solar B » Mon Aug 24, 2015 9:02 pm

Hi Luke

I'm not able to compare with any DS filters at the mo as I don't have any ... but
was relying on my memory of using such as you mention (60mmDS) .... and as
I recall the contrast and detail the quark offers is somewhat less ....
I currently use the Quark with a equinox 66 and 40mm EP which does offer full disc views but I probably really prefer those of my PST !!
Sure for resolution the quark is going to beat almost any dedicated solar scope / filter with its choice of apertures but I think its quality over quantity views that I'm after.

I'm still getting to grips with it but do find it hard work visually compared
to every other HA filter/scope I've used to date .... plus I'm starting to get a bit paranoid as all I hear is how good they are .... but think I'd probably still prefer an SM60.

Brian
" Gentlemen only ever use Refractors "

Derek Klepp
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 12660
Joined: Wed Oct 19, 2011 10:02 am

Re: Some more observations on filter performance

Post by Derek Klepp » Mon Aug 24, 2015 9:24 pm

Thanks Bob at least I know now what I am actually seeing .In terms of professional research these parameters would be very important.

User avatar
Bob Yoesle
Im an EXPERT!
Im an EXPERT!
Posts: 473
Joined: Thu Mar 08, 2012 7:24 pm

Re: Some more observations on filter performance

Post by Bob Yoesle » Tue Aug 25, 2015 4:33 pm

Thanks guys.
...meaning that in future I will not be perusing rear mounted
SS etalons but will revert back to DS front mounted options.
I think rear-mounted single etalons have a great place in hi-res applications with larger apertures. Just look at some of the work posted on this site.

The potential shown by Christian Valadrich, and mentioned in the Bray and Loughhead article, for stacking a ~ 1 or 1.5 A filter to further suppress the leakage of parasitic continuum needs to be explored much further. The fact that the commercial makers of these filter systems have not done this already may speak more to their lack of familiarity surrounding the issue, not the ability (or lack thereof) to potentially accomplish a good result.

Beyond this, using a front (external) or collimmated (internal) etalon with a rear etalon would also seem to be a viable DS alternative, if the rear filter is properly configured with appropriate secondary (telecentric) optics...
Diagonally parked in a parallel universe.

Curiosity is the father of knowledge; uncertainty is the mother of wisdom.

Dark-Sky Defenders

Solar B
Ohhhhhh My!
Ohhhhhh My!
Posts: 139
Joined: Mon Jul 27, 2015 9:29 am
Location: Dunfermline , Scotland

Re: Some more observations on filter performance

Post by Solar B » Wed Aug 26, 2015 3:09 pm

Totally agree Bob ... rear mounted etalons are superb ... however what
I had wrongly thought / believed before trying some and reading your article and
indeed others ... looking to increase they're contrast ... is that a 0.3A(rear mounted) would noticeably outperform (on contrast & surface detail) a 0.5 DS front mounted pair !!
Quarks etc are good but give me a DS 40 - 60 any day :)

Brian

Also have been trying to get my head round what fwhm actually means ?
"In other words, it is the width of a spectrum curve measured between those points on the y-axis which are half the maximum amplitude"
Too much for my wee brain ?
" Gentlemen only ever use Refractors "

User avatar
Merlin66
Librarian
Librarian
Posts: 3169
Joined: Tue Dec 13, 2011 1:23 pm
Location: St Leonards, Australia
Contact:

Re: Some more observations on filter performance

Post by Merlin66 » Wed Aug 26, 2015 11:02 pm

FWHM is the width of the Gaussian curve at the position of 50% the peak intensity. It is also based on the Airy Disk, the accepted measure of resolution - when two Airy Disks (simple example - an equal magnitude double star) sit close to each other such that the first dark ring of one coincides with the peak of the other the spacing between them will be equal to the FWHM result.
Harrison Fig 4_3.jpg
Harrison Fig 4_3.jpg (32.14 KiB) Viewed 2480 times
Harrison Fig 4_4.jpg
Harrison Fig 4_4.jpg (27.02 KiB) Viewed 2480 times
(Bob, this also defines, for a Gaussian distribution of intensity - which I think all filters give us - that the curve reaches zero intensity at +/- FWHM from the central wavelength)
Last edited by Merlin66 on Fri Aug 28, 2015 10:53 pm, edited 1 time in total.
Reason: images added
"Astronomical Spectroscopy - The Final Frontier" - to boldly go where few amateurs have gone before
https://groups.yahoo.com/neo/groups/ast ... scopy/info
"Astronomical Spectroscopy for Amateurs" and
"Imaging Sunlight - using a digital spectroheliograph" - Springer

User avatar
Bob Yoesle
Im an EXPERT!
Im an EXPERT!
Posts: 473
Joined: Thu Mar 08, 2012 7:24 pm

Re: Some more observations on filter performance

Post by Bob Yoesle » Fri Aug 28, 2015 9:19 pm

Hi Ken,

Strictly speaking, I believe most single-stacked filters are a Lorentzian distribution as opposed to a Gaussian. Indeed, multiple stacking is what brings the filter transmission profile closer to a Gaussian: http://www.cloudynights.com/topic/40132 ... try5150574 This property is also what appears to suppress parasitic continuum -- For equal FWHMs the base of the profiles is significantly wider for the Lorentzian verses the Gausssian, and the Gaussian base indeed appears to be about 1 + times the FWHM from the centerline -- base (X axis) scale in FWHM units:
Compare curves 4 resized.jpg
Compare curves 4 resized.jpg (112.28 KiB) Viewed 2469 times
Adapted from George's original diagram. Note all the profiles illustrated have the same FWHM (bandpass). This explains the improvement from stacking, and why single filters with a narrower bypass don't have the contrast performance of a DS with the same FWHM.

As George notes:
It looks suspiciously like a large number of Lorentzians will approach Gaussian.
For our purpose a fewer number of Lorentzians than indicated above may get closer to a Gaussian due to the concurrent reduction in FWHM with double stacking...

Absorption profile & filter bandpass SM.jpg
Absorption profile & filter bandpass SM.jpg (53.19 KiB) Viewed 2456 times
Bob Algoni recently sent me some very interesting plots showing single stack 0.7 A and double stack 0.5 A output plots, apparently derived by a convolution of the transmission curves with the absorption line curve of the photosphere as shown above.

0.7 A SS convoluted output:
Single stack output.jpg
Single stack output.jpg (185.97 KiB) Viewed 2473 times
0.5 A DS convoluted output:
Double stack output.jpg
Double stack output.jpg (169.97 KiB) Viewed 2473 times
Comparison output:
SS v DS output compare.jpg
SS v DS output compare.jpg (150.11 KiB) Viewed 2473 times
However, the DS output again appears generated from a "normalized" DS transmission plot. I therefore made the comparison plot below based on an assumed 60% of SS peak for the DS peak, and added the +/- 0.065 nm limit for the photosphere leakage cited by Bray and Loughhead:
SS v DS output compare non-normalized 0.65 A limit 3.jpg
SS v DS output compare non-normalized 0.65 A limit 3.jpg (169.56 KiB) Viewed 2456 times
These convoluted filter ouptut profiles appear to go even farther in demonstrating where the improvement in chromospheric contrast - and suppression of photospheric continuum - is coming from with double stacking. The transformation from a Lorentzian to more of a Gaussian transmission profile appears to be a key feature, which seems to reinforce the position that filter bandpass alone is an insignificant factor by which to judge filter system contrast performance.

Moreover, if continuum is sufficiently reduced by double stacking - which seems to be the case - it may also suggest why there is not much to gain from triple stacking, while the peak transmission (and image brightness) would be significantly reduced.
Last edited by Bob Yoesle on Sat Aug 29, 2015 2:33 pm, edited 2 times in total.
Diagonally parked in a parallel universe.

Curiosity is the father of knowledge; uncertainty is the mother of wisdom.

Dark-Sky Defenders

User avatar
Valery
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 3053
Joined: Sun Nov 18, 2012 3:13 pm

Re: Some more observations on filter performance

Post by Valery » Sat Aug 29, 2015 11:16 am

Hi Bob,

Thanks a lot for you very well illustrated and very interesting info here.

Also, thanks to all discussion participators.

Valery
"Solar H alpha activity is the most dynamic and compelling thing you can see in a telescope, so spend accordingly." (c) Bob Yoesle.

Largest full size 185 - 356mm Dielectric Energy Rejection Filters (D-ERF) by ARIES Instruments.

Solar B
Ohhhhhh My!
Ohhhhhh My!
Posts: 139
Joined: Mon Jul 27, 2015 9:29 am
Location: Dunfermline , Scotland

Re: Some more observations on filter performance

Post by Solar B » Sat Aug 29, 2015 11:32 am

Oh dear ... picture that moment in the film "Scanners" where the guys head
explodes ... that's how I feel trying to get mine all round this !
and thanks Ken for your help.

Brian
" Gentlemen only ever use Refractors "

User avatar
marktownley
Librarian
Librarian
Posts: 24536
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Contact:

Re: Some more observations on filter performance

Post by marktownley » Mon Aug 31, 2015 8:35 pm

Good article!
Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!

User avatar
Merlin66
Librarian
Librarian
Posts: 3169
Joined: Tue Dec 13, 2011 1:23 pm
Location: St Leonards, Australia
Contact:

Re: Some more observations on filter performance

Post by Merlin66 » Tue Sep 01, 2015 12:01 am

Good report.
I personally think the DS filter would be used at a longer exposure to real get into the core of the Halpha line.
It would be interesting to see the wing comparison when each filter has the same peak intensity.
"Astronomical Spectroscopy - The Final Frontier" - to boldly go where few amateurs have gone before
https://groups.yahoo.com/neo/groups/ast ... scopy/info
"Astronomical Spectroscopy for Amateurs" and
"Imaging Sunlight - using a digital spectroheliograph" - Springer

christian viladrich
Im an EXPERT!
Im an EXPERT!
Posts: 439
Joined: Sun Jun 14, 2015 4:46 pm
Location: France
Contact:

Re: Some more observations on filter performance

Post by christian viladrich » Fri Sep 04, 2015 12:20 pm

Hello Bob,
Thanks for this very interesting discussion.
Maybe one small addition : what is called "finesse" has little to do with the shape of the transmission curve. Whatever the finesse or FWHM, the transmission curve is still a Lorentzian curve for an air-spaced (or mica-spaced) Fabry-Perot etalon.
Steeper transmission curves are acheived by Lyot filters or, as you said, multi-stacked F-P etalons.

I have not been have to test the combination of a 1.5 A and 0.6 A filter. Still need to find a good 1.5 A filter ...

BTW, I've read recently that old "Professional grade" Daystar filters had two stacked mica-spaced F-P and not only one. Has anyone more information about this ?
Christian Viladrich
Co-author of "Astronomie Planétaire"
http://www.astroplanetes.com/
Co-author of "Astronomie Solaire"
http://www.astronomiesolaire.com/

User avatar
Bob Yoesle
Im an EXPERT!
Im an EXPERT!
Posts: 473
Joined: Thu Mar 08, 2012 7:24 pm

Re: Some more observations on filter performance

Post by Bob Yoesle » Sat Sep 05, 2015 2:49 pm

Thanks everyone. However, I still think there is some additional analysis that can and needs to be done... I'll have to get myself up to speed with GNUPLOT and convolution theorems/algorithms to see what else might be derived.

Christian: the Beloptik 1.5 A filter seems to have disappeared from the website. However, Bob Luffel of Alpine Astronomical here in the States -- the Baader USA Rep. -- tells me Baader will soon have a 1 or 1.5 A H alpha filter for use in a "chronograph." Being from Baader we know it will be expensive, but of very high quality. I told him of your work and the benefits of DS continuum suppression, and I'm pretty sure he's on board to try a DS using the new Baader filter with his soon to be delivered Solar Spectrum filter when it arrives.

Somewhere I recall reading that for achieving a < 0.3 A FWHM, either DayStar or Solar Spectrum DS mica etalons...
Diagonally parked in a parallel universe.

Curiosity is the father of knowledge; uncertainty is the mother of wisdom.

Dark-Sky Defenders

User avatar
Merlin66
Librarian
Librarian
Posts: 3169
Joined: Tue Dec 13, 2011 1:23 pm
Location: St Leonards, Australia
Contact:

Re: Some more observations on filter performance

Post by Merlin66 » Sat Oct 10, 2015 6:11 am

"Astronomical Spectroscopy - The Final Frontier" - to boldly go where few amateurs have gone before
https://groups.yahoo.com/neo/groups/ast ... scopy/info
"Astronomical Spectroscopy for Amateurs" and
"Imaging Sunlight - using a digital spectroheliograph" - Springer

Gert
Ohhhhhh My!
Ohhhhhh My!
Posts: 50
Joined: Wed Nov 13, 2013 11:42 am

Re: Some more observations on filter performance

Post by Gert » Mon Oct 12, 2015 6:49 am

Hi Bob,

I agree with your analysis that the stacked filter reduces the peak transmission and was wondering if anybody might be able to contribute data to the same. I checked the Lunt web page for their front etalons which is basically a plain etalon in parallel light, but they would not give a peak transmission.

Wikipedia says that high finesse (very desirable) etalons would have 95% mirror reflectivity.
https://en.wikipedia.org/wiki/Fabry%E2% ... rferometer

Then the TecOptics page here:
http://www.tecoptics.com/etalons/theory.htm

gives the peak transmission function. If we take the sin^2 part to be 0 then the transmission becomes 1.0. But that's ideal. What would one assume for the real etalon? One could numerically plug the H-A profile and one resp. two etalons with give peak transmission into a spread sheet and see.

Clear Skies,
Gert
Gert Gottschalk, Ph.D.

Tri-Valley Stargazers, Eastbay Astronomical Society, WFS Berlin

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests