Page 1 of 1

Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Thu Jul 07, 2016 6:57 pm
by christian viladrich
Dear all,
I had some fun with the following test comparing 0.3 and 0.6A Ha filters.
Here is a first comparison with the same processing applied to both images. There is no local processing. Only linear visualisation with the objective of having about the same luminosity on the disk.
The 0.3 A filter is on the left and the 0.6 A on the right :
Image

In the 0.6A image, the photosphere is visible through the chromosphere. This is why the contrast is lower.
The intensity of the chromospheric fringe is about 33% the intensity of the disk with the 0.6A, versus 46% with the 0.3A filter.

Here, I processed each of the image in order to get the most of each:
Image
The granulation of the photosphere is clearly visible through the chromosphere in the image with the 0.6A.
For some strange reasons, the chromospheric fringe looks sharper with the 0.3A.Something related to the Doppler effect ?
The prominence is also more visible with the 0.3A, but this is just because of its evolution between these two images.

Here are the full images :
Image
Image
Cheers
Christian

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 08, 2016 2:42 am
by Bob Yoesle
Nice comparison Christian!

Since no "fringe" is visible when double stacking 0.7 A bandpass filters (even though the FWHM becomes ~ 0.5 A) the real-world performance of such a DS'd filter system appears to exceed that of a 0.3 A single filter: http://www.cloudynights.com/topic/43800 ... ted-equal/

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 08, 2016 9:22 am
by mattwastell
Thanks for the comparisons. The images are great BTW!

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 08, 2016 9:23 am
by Montana
I always find that prominences are more visible with a lower band pass.

I guessed immediately it was a rear mounted etalon as the photosphere is still visible in the 0.3A due to the leakage of the broad shoulders in the transmission curve.

Great shots though :hamster: :bow :bow

Alexandra

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 08, 2016 3:25 pm
by cshahar
Christian, this is very informative. Thank you!

-Charles

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sat Jul 09, 2016 12:00 pm
by Valery
Thanks, Christian. Very interesting comparision between two DayStar etalons.

BTW Have you applied a flat field to both etalons? Or none of them?
How much better in uniformity the PE vs SE versions?

Can you show 2-3 images taken with DayStar Ion 0,3A SE without flat field applied?

Thanks in advance.


Valery.


BTW, I have conducted much more numerous comparisions between two DayStar etalons
with 0,25A (RMS) and 0,38A (RMS) etalons and the difference is not as prominent as 0,3A vs 0,6A
but still clearly visible.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sat Jul 09, 2016 12:36 pm
by george9
Very useful for someone choosing a filter, especially the first two images. For someone purely visual, given the difference in price, it is not clear that you get that much by halving the bandwidth. I was thinking 0.5A would about right, all things considered. The eye has a good dynamic range, so the brighter photosphere in the .6 compared to the .3 is not as harmful as it would be for an image. If you really want to increase visual contrast significantly, you've just got to figure out some way to double stack.

I understand the theory that with a narrow filter, Doppler effects could filter out parts of prominences and spicules, but in practice, I just have not seen it (other than filaments that are way off band). I think the main effect is not bandwidth but brightness. If the narrower filter (single or double stack) has a lower peak transmission, then you'll definitely lose prominences and spicules. And if you can boost brightness, e.g., by dropping one of two ERFs for double stack, then prominences and spicules will be just as good as in a wider filter. (I am sure there are exceptions to this, when some feature does have a large Doppler effect, but it just doesn't seem common.)

Christian, did it seem that the brightness of the proms and spicules in the 0.6 and 0.3 were identical before processing? That is, my understanding is that you normalized the first two images so that the disks (photosphere + chromosphere) looked the same, which made the proms in the 0.6 look dimmer. But before processing, did the .6 and .3 let about the same amount of chromospheric light through? Or perhaps was the .3 dimmer than the .6?

George

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sat Jul 09, 2016 3:40 pm
by pedro
Great comparison Christian. The Daystar etalons are a different ball game

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sun Jul 10, 2016 11:06 am
by marktownley
A very informative comparison Christian, thanks :)

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sun Jul 10, 2016 9:01 pm
by christian viladrich
Thanks a lot for your comments :-)

Bob : Yes, FWHM is only part of the story. Double stacked filters take the lead because of the reduction of the continuum contribution.

Alexandra : all other things being equal, there should not be any difference between a front and a rear mounted filter. The wings have the same width for mica-spaced of air-spaced etalons of a given FWHM. Difference can come from the BF, and of course between DS and SS. Another thing may be the difference between advertised and actual FWHM.

Valery : yes, I applied flat image for both the PE 0.6 and Ion 0.3A. The PE is rather uniform, while part of the Ion is a little bit off-band. I' ll post the flat images soon.
What type of Daystar etalons did you test ? PE, SE, Ion ?

Georges : I agree with you. I have not kept the raw images, but I kind of remember the 0.3A is about 2 to 3X darker on prominences (with same exposure time). I am not 100% sure. I have to check it again ... I think (just as you do) this is more an issue related to the difference of peak transmission in the BF and the polarising filters, rather than a question of band pass.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 6:55 am
by mdwmark
Hi Christian,
I liked your images. But they where all in the red wing. They are good examples of what .3-.5Ang red looks like. When you see the tips of the surface detail with the higher contrast. Then you know you are past the center line. It will look softer when you are in the center of the line.
Also, if you where at F/30, then you where not at .3Ang HW. You would really need to be around F/45 to take advantage of the narrower HW. The .6 at F/30 would show less effect of boarding then the narrower filter.(it looked like you where using the F/27 system, if I read the image right. Then the .3HW would be closer to .6Ang and the .6 would be closer to .8Ang)
Just a friendly comment
Mark W.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 6:24 pm
by Valery
christian viladrich wrote:
Valery :
What type of Daystar etalons did you test ? PE, SE, Ion ?
Christian,

These are two Quarks Chromosphere - one with a 0,38A (RMS) and the second is 0,25A RMS.


Valery

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 6:28 pm
by Valery
mdwmark wrote: if I read the image right. Then the .3HW would be closer to .6Ang and the .6 would be closer to .8Ang)
Just a friendly comment
Mark W.
Hi Mark,

Do you have some photos taken with your narrowest filters (say, 0,2A or even narrower) and at slow enough F/D to show
all the advantages of these narrow filters?

The most contrast images I ever saw (except professionals) were taken through DS 0,3A etalons.

Thanks in advance.


Vaalery.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 6:49 pm
by christian viladrich
Hello Mark, thanks a lot for having had a look there :D

I've made some calculations about the impact of the F/D ratio on the actual bandpass for mica-spaced F-P. I've got a curve which is a bit different than the one provided by Daystar :
Image
I was wondering if you could have a quick look at some of the calculations there. I know it is OK for air-spaced F-P since I've reached the same results as for a space mission project. But maybe I've got it wrong somewhere for the mica-spaced F-P:
http://www.astrosurf.com/viladrich/astr ... lar/FP.htm

I've tested various positions (to the red and blue) with the PE and the Ion filters. That ones were giving the darkest images, so possibily at the center of the Ha line ?

Ghost : I've made some comments/ questions to your post.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 7:45 pm
by christian viladrich
Valery : Thanks, this is very interesting. How did you measured the FWHM ? Do you have some kind of spectrometer ? If so, did you find a large (or small) variability between the different Quark you have tested ?

I have also have this set of comparison images taken recently with the C8 Edge-HD at F/27.
Daystar PE 0.6 A on left, Ion 0.3A on right. The processing and visualisation thresholds are exactly the same for for images :
Image

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 8:00 pm
by marktownley
christian viladrich wrote:Valery : Thanks, this is very interesting. How did you measured the FWHM ? Do you have some kind of spectrometer ? If so, did you find a large (or small) variability between the different Quark you have tested ?

I have also have this set of comparison images taken recently with the C8 Edge-HD at F/27.
Daystar PE 0.6 A on left, Ion 0.3A on right. The processing and visualisation thresholds are exactly the same for for images :
Image
You can see the granulation hiding very well in the image on the left! :)

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Mon Jul 11, 2016 9:20 pm
by christian viladrich
Yes, it is clearly there Mark ;-)

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Tue Jul 12, 2016 7:31 am
by jp-brahic
PE 0.6 have a big problem :) I saw exactly the same thing on another PE 0.6

jp

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Tue Jul 12, 2016 8:18 am
by Derek Klepp
Very informative thanks.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Tue Jul 12, 2016 11:05 am
by george9
Yes, looking at those latter images, you may have convinced me about the 0.3A. I.e., that if you are going single stacked, then it is probably worth the money for the 0.3A. Certainly for imaging. For visual, I would have to compare 0.5A to 0.3A to make sure the 0.3A was not too dim visually. I know a 0.25A (I think) Solarspectrum filter at NEAF is certainly bright enough every year.

George

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Tue Jul 12, 2016 2:51 pm
by eroel
Chris.
A very interesting theme, thanks for sharing it.
Regards,
Eric.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Wed Jul 13, 2016 5:33 am
by marktownley
That Ion would work well with a PST etalon behind it. Lots of contrast then! ;)

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Wed Jul 13, 2016 10:29 am
by marktownley
I thought I might throw this into the mix as serves as comparison for some of the focus of the discussion that has evolved.

The following shot was taken with my HaT and a Quark that was quoted as being 0.4a that has been double stacked with a PST etalon. Notice lack of continuum leakage due to wings being cut.

Images5353-ha-bw by Mark Townley, on Flickr

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Thu Jul 14, 2016 9:00 am
by christian viladrich
Very interesting Mark !
Could you post of the sun an image with / without the PST double stack ? And an other of the set up ?
Thanks !

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Thu Jul 14, 2016 7:40 pm
by Valery
jp-brahic wrote:PE 0.6 have a big problem :) I saw exactly the same thing on another PE 0.6

jp
I am sure there is no any problem with a DayStar 0,6A PE. They just too wide and hence a lower contrast.
0,3A is MUCH narrower and contrasty!


Valery.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 15, 2016 7:42 am
by jp-brahic
It's completely abnormal to see the photosphere ( solar granulation ) by the transparency with such a filter!!!! Filter Daystar has generally a problem of transparency of the chromosphéric material, how to explain that a standard Etalon PST ( at 1A° ) shows much more chromosphéric material than a 0.3A ° Daystar?


Image



JP

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Fri Jul 15, 2016 9:47 am
by Valery
jp-brahic wrote: how to explain that a standard Etalon PST ( at 1A° ) shows much more chromosphéric material than a 0.3A ° Daystar?

PST etalons are VERY VARY in their bandwidth! I saw 0,9A - 1,0A and I saw as narrrow as <0,3A.

Similar variability in their bandwidth one can find in Quarks, Lunt (LS35 and LS50) and in Coronado etalons.

Your PST etalon sample is at least as narrow as <0,3A - hence your excellent images in H-a.

BTW. Christian confirmed by the math how much is a photospheric part in the tolal energy transmission in such a
filter like 0,6A.

Valery.

Re: Comparing 0.3A and 0.6 A bandpass in Ha

Posted: Sat Jul 16, 2016 6:15 pm
by christian viladrich
Hello Jean-Pierre,

You raised at interesting question. Here is my present understanding of the situation. It is based on some calculations, some tests and observations. Of course, I may be wrong somewhere :

For a given FWHM :
1) the transmission profile is exactly the same for an air-spaced Fabry-Perot (Coronado, Lunt, etc) and for a mica-spaced F-P (Solar spectrum, Daystar). There is no doubt about this ;-)
The bandpass at 10% transmission is 3 x FWHM, the band pass at 1% transmission is 9.95 x FWHM.
2) With a frontal step up (or in collimated light), the diameter of the sweet spot is smaller with an air-spaced F-P compared to a mica-spaced F-P.
3) In a telecentric set-up, an air-spaced F-P required a larger F/D ratio to reach the same resulting FWHM than a mica-spaced .

Another important point, is than the "contrast" of an Ha filter (ie its capacity to cut photosphere light, and let only chromosphere light through) is very badly defined by its FWHM. It would be more relevant to use the band pass at 10% transmission. This explains why stacking F-P gives very good results.

So ... back to the PE 0.6 A versus PST 1A. I guess the answer has been given by Valery ;)

I have looked through some PE0.6 or 0.5A, and also through some PST. There is a lot of variability on PST (and also on Quark). Some are very bad, a lot are average, a very few of them are excellent. There is less variability on PE.

For sure, a "good" PST doesn't have 1 A FWHM, and far from all PST have FWHM 0.3A.

I think Valery told us he tried about 10 (?) PST to nail an excellent one. I have no statictics there. Maybe, we can launch a pool on "how many PST you have to test before finding an excellent one" ;-)

This is not a critic to Coronado, since they adverstise 1 A FWHM. Indeed, we should be happy to find out a few PST with very narrow FWHM. The same holds for Quark.

There is also the question of the BF. All are not equal ...

Just my five cents on this interesting subject.