39 replies to this topic

[BYoesle]

I was wondering what, if any, benefit I may get from an H-alpha filter designed for viewing nebula emissions. It's about 7nm wide FWHM. I understand that's really wide by solar viewing standards, will there be any benefit with even say flares or prominences?

Thanks,
Mike

Many people wonder why you can't use a white-light solar filter in conjunction with a night-time Hydrogen alpha nebula filter to see solar flares and prominences. The first thing you must understand is that the sun is composed of many layers, which give off and absorb light in differing ways.

The baseline of light coming from within the sun gives the bright "continuous spectrum" of all the colors of the rainbow. This is known as the "thermal" or "blackbody" radiation, and any dense solid object will produce this radiation (because of the sun's high internal density, it acts like a solid object at normal earthlike temperatures). As the temperature rises, the body will shift from giving off infrared (heat) energy to visible energy and into the ultraviolet and beyond, in the form of a continuous spectrum. Most stars are classified on this basis of where the peak of their light output falls in the visible spectrum.

Because the sun is not a solid body, the relatively cooler gases that lie above the layers producing the continuous spectrum can absorb certain frequencies of light, which correspond the amount of energy required to shift electrons of a given element into a higher energy level, or "orbit." Thus when an electron orbiting the proton of a Hydrogen atom absorbs a discrete amount of energy to shift the electron, it removes the corresponding color, or frequency, of light from the continuous spectrum coming from below, and an "absorption line" is created in the spectrum.

If we look closely at the bright continuous spectrum of the sun, we observe many "spectral lines" corresponding to the various elements of the gases in the photosphere -- the "absorption spectrum." Relatively cool (6000 degrees C) hydrogen gas in the photosphere absorbs light coming up from below, and we see a very dark line – the Hydrogen alpha absorption line -- in the continuous spectrum. It is the darkest line, and relatively wide, because the sun is mostly hydrogen, and it is this gas that predominates the layers of the photosphere.

The sun's chromosphere lies above the photosphere, and is much hotter -- 20,000 degrees C. At these higher temperatures we see what is know as an "emission spectrum". Here light is being given off instead of absorbed, corresponding to the energy given off when the electron shifts back down to a lower energy state, and at the same frequency or "color" of light which is absorbed in the photosphere. Even though it is much hotter than the photosphere, the amount of light coming from the chromosphere -- which is far less dense than the photosphere -- is correspondingly less intense by several orders of magnitude.

In order to see the chromosphere and it's flares, prominences, etc., you must first totally eliminate all the bright light coming from the photosphere, except where the broad H alpha absorption line is. Since the Hydrogen in the photosphere is absorbing the exact wavelength of light we are interested in, all that will be left to see is the relatively weak H alpha emission coming from the chromosphere.

This is why you can't use a white light solar filter with a nebula filter and see chromospheric features like flares and prominences: The white light filter reduces the H alpha emission from the chromosphere by the same amount it reduces all the other wavelengths from the photosphere. The wide bandwidth of a nebula filter only lets you see the photospheric features from light coming from either side of the photospheric absorbtion line, and the chromosphere emission itself is reduced by the white light filter to invisibility.

[David Knisely]

Byoesle wrote:

If we look closely at the bright continuous spectrum of the sun, we observe many "spectral lines" corresponding to the various elements of the gases in the photosphere -- the "absorption spectrum." Relatively cool (6000 degrees C) hydrogen gas in the photosphere absorbs light coming up from below, and we see a very dark line – the Hydrogen alpha absorption line -- in the continuous spectrum. It is the darkest line, and relatively wide, because the sun is mostly hydrogen, and it is this gas that predominates the layers of the photosphere.

Visually in a simple spectroscope, the H-alpha line is not necessarily the "darkest". Indeed, the H and K lines of Calcium are quite broad and somewhat easier to see with a spectroscope than is H-alpha. Even the G-band, the H-Beta line, and the Sodium doublet tend to be somewhat easier to notice than H-alpha (mostly because of the eye's lower red sensitivity), although the H-alpha line is still fairly dark and deep.

The sun's chromosphere lies above the photosphere, and is much hotter -- 20,000 degrees C. At these higher temperatures we see what is know as an "emission spectrum". Here light is being given off instead of absorbed, corresponding to the energy given off when the electron shifts back down to a lower energy state, and at the same frequency or "color" of light which is absorbed in the photosphere.

Actually, light is still being absorbed in the chromosphere (otherwise, the hydrogen atoms in the chromosphere would remain in the ground state and not have excited electrons to decay back to the ground state and release an H-alpha photon). Indeed, many of the features seen against the solar disk in sub-angstrom solar H-alpha filters do appear dark as the gas they contain absorbs light from the sun. What happens is that the chromospheric feature absorbs the intense light from directly beneath the feature (over a somewhat limited solid angle) but then re-radiates it in all directions, with only a small amount of the re-radiated energy heading directly on a line to the observer. This is why filaments and fibrils appear dark on the solar disk. On the limb, we finally see the weaker emission component of the filament against the black background of space, so the filament becomes a brighter "prominence". Unlike the photosphere's surface, the chromosphere does not really have a well-defined temperature, as it reaches a minimum at its base (somewhat lower than the photospheric temperature), and then rises both as altitude increases and as you get over areas of the magnetic network where it can also be higher. What can be said is that the temperature is probably somewhere between 3500 K and the 1,000,000 K of the Corona, although I have seen figures of 8000 K to 10,000 K for some portions of its upper edge.

This is why you can't use a white light solar filter with a nebula filter and see chromospheric features like flares and prominences: The white light filter reduces the H alpha emission from the chromosphere by the same amount it reduces all the other wavelengths from the photosphere. The wide bandwidth of a nebula filter only lets you see the photospheric features from light coming from either side of the photospheric absorbtion line, and the chromosphere emission itself is reduced by the white light filter to invisibility.

It might be a little more general to say that the reason broader "nebula" H-alpha filters won't show the chromospheric detail is that they are letting in a little too much off-band light from the photosphere. That light is basically "drowning out" the weaker chromosphere. If you use one of those really narrow night H-alpha filters (say, under 70 Angstroms FWHM) and an occulting disk to cover the photosphere, you might still occasionally photograph some of the brighter prominences on the limb, but something narrower would still be better (4 angstroms or less). With a regular white-light solar filter, you would, of course, dim the chromospheric emission, but the H-alpha filter would still be letting in too much light from wavelengths well away from the H-alpha centerline, and that is the crux of the problem. Clear skies to you.

[BYoesle]

Thanks David, you are of course correct in many details.

Given the frequency of those new to solar observation and asking this same question of using nebula/white light filters to view the sun, I was trying to keep the concepts simplified, as the details may tend to bewilder a person new to solar filters, and unfamiliar with solar physics and spectroscopy.

I might disagree about the visibility of the Ca H & K lines, as these are in a part of the spectrum the eye is even less sensitive too than the H alpha line, as you note.

It might be a little more general to say that the reason broader "nebula" H-alpha filters won't show the chromospheric detail is that they are letting in a little too much off-band light from the photosphere. That light is basically "drowning out" the weaker chromosphere... With a regular white-light solar filter, you would, of course, dim the chromospheric emission, but the H-alpha filter would still be letting in too much light from wavelengths well away from the H-alpha centerline, and that is the crux of the problem.

The off-band contamination of a broad-band H alpha nebula filter used with a white light filter is irrelevant and only adds insult to injury: chromospheric details and prominences will be invisible with a white light filter because they are rendered 100,000 to 1,000,000 times fainter than they otherwise would be. That's why during the totality phase of a solar eclipse you do not see anything through a telescope equiped with a white light filter, where there is no side band contamination whatsoever.

This is easy to verify, since a narrow-band H alpha filter passes ~ 95% + of the H alpha emission -- just place a white light filter in front of your PST or other H alpha solar filtered telescope -- there will be no image to observe, even without photosphereic side band contamination.

[Ralph Marantino]

If it worked why would any of us spend so much on our Solar equipment?

Attached Thumbnails

• 

[blackhaz]

Just wanted to add that the standard 7nm Baader H-Alpha deep sky filter does not even work as a Coronado's block filter. Too wide bandpass, no details are seen.

[colinsk]

It is very important the a blocking filter not only has the right bandwidth to block the extra peaks from the etalon, but it is equally that the off band performance exceeds the Optical Density to make vewing safe. ODs of 6 to 8 are common in blocking filter systems.

[David Knisely]

BYoesle posted:

The off-band contamination of a broad-band H alpha nebula filter used with a white light filter is irrelevant and only adds insult to injury: chromospheric details and prominences will be invisible with a white light filter because they are rendered 100,000 to 1,000,000 times fainter than they otherwise would be.

Well, this does cover those who might want to try a H-alpha nebula filter with a white light filter, but does *not* cover those who might try the H-alpha filter "bare bones". This is a danger which I felt needed to be mentioned (not to mention that it wouldn't work either). As I noted, the white light filter with the H-alpha nebula filter would dim the H-alpha detail considerably, but you still might need to mention up-front that the H-alpha nebula filter will not work *period* due to it being too broad and would be risky if it were tried solo. That would cover all the bases.


As for the visual spectrum, when I first made a simple diffraction grating spectroscope in the 1970's, I had a hard time identifying any but a few of the more prominent lines. I quickly picked up the Sodium D-doublet and H-Beta, along with the G-band, but with the bright sun, my eye was immediately drawn to the dark broad features both in the violet (the Calcium lines) and in the deep-red (the Telluric lines of Oxygen). H-alpha took a while to isolate (and only after some effort with a chart of the solar spectrum). It is obvious to me now, but compared with the others, it isn't exactly the darkest or most prominent absorption line in the visible spectrum. Clear skies to you.

[MikeBu]

If it worked why would any of us spend so much on our Solar equipment?

I'm not sure I follow your logic. I notice you have a 152mm Astrophysics refractor, does that mean the $270 6" DOB from Orion can't be used for night time star gazing?

[AKP]

Has anyone tried using a night time H alpha or any other filter with a Herschel wedge? If the image was too dim, has a long exposure CCD cam been used with this setup? Thanks.

[David Knisely]

Has anyone tried using a night time H alpha or any other filter with a Herschel wedge? If the image was too dim, has a long exposure CCD cam been used with this setup? Thanks.

Again, you are throwing a lot of light away with a Herschel wedge, so even if the filter was narrow enough to fully screen out the off-band light, you probably would not see anything. Imaging with long exposures with the Herschel wedge and the nighttime "nebula" H-alpha filters would only show white-light detail because these filters are just plain far too broad to kill off enough of the light from the photosphere to let you see much of the chromosphere. The narrowest I know of are the Baader H-alpha CCD with a Full-width at Half Maxima of 65 angstroms and the Custom Scientific H-alpha filter with a FWHM of about 41 Angstroms. Now it *might* be possible with an occulting disk to image the brightest prominences with these two filters at a good high-altitude site, but I would not risk it visually. Even the older Baader Coronagraph used an 11 Angstrom passband filter along with occulting disks. To see prominences alone without the use of an occulting disk requires a bandwidth that is about 1.5 angstroms or narrower (as is found with the fine Thousand Oaks prominence filters). Seeing H-alpha chromospheric disk detail requires a bandwidth of under one angstrom, so the H-alpha nebula filters just are way too broad to work for solar use. People who try the night H-alpha filters with regular white-light solar filters will just see a pretty dim red sun with more or less the same features that are seen with the white light filter alone. Again, for the lowest entry cost in solar H-alpha viewing, the Coronado PST is probably the best way to go. Clear skies to you.

[AKP]

David ,
Thank you for your very knowledgeble and detailed answer. We are all fortunate to have experts like you chime in.

[MikeBu]

I was curious so I crunched some numbers on the solar spectrum.

Bandwidth………......………...…..………Power………………..........Spectrum
200nM - 10,000nM……………...…...…1.37 kW/sq.M.............(IR - UV)
380nM – 750nM……………………..........531 W/sq.M.............(Visible)
655.8nM – 656.8nM………...……...……1.12 W/sq.M.............(1.0 nM Red around H-alpha)
656.2050nM – 656.3550nM...….…..75.2mW/sq.M.............(1.50 Angstrom, H-alpha)
656.2300nM – 656.3300nM...….…..35.7mW/sq.M.............(1.00 Angstrom, H-alpha)
656.2425nM – 656.3175nM...….…..22.9mW/sq.M.............(0.75 Angstrom, H-alpha)
656.2455nM – 656.3050nM...….…..14.1mW/sq.M.............(0.50 Angstrom, H-alpha)

The minimum at H-alpha is approx. 16.04% the local maximum (i.e. Plank curve). At 0.75 and 0.50 Angstroms the ratio of bandwidths is close to the ratio of the powers indicating the spectral density is getting pretty flat.

Comparing the 1nM (10 Angstrom) band to the 1.00 Angstrom band the 1nM band has 31.4 times the light, more than enough to mask the H-alpha emission spectrum. Not that I didn’t believe everyone here, but just to get some perspective.

For a 100mm scope aperture with no white filter, the 1nM filter will need to dissipate 10.8 Watts. The level through just the 1nM filter would still be 8.8mW, brighter than looking into a red laser pointer. Even if the level of light through a nighttime filter was eye safe without a white filter it probably couldn’t dissipate the power and would fail in a very unpleasant way.

[colinsk]

Mike, I love this data. How did you calculate it? Does assume the sun is a blackbody emitter? Or does it account for the spectral lines as well.

[BYoesle]

Comparing the 1nM (10 Angstrom) band to the 1.00 Angstrom band the 1nM band has 31.4 times the light, more than enough to mask the H-alpha emission spectrum.

This doesn't seem to be enough of a difference to account for the absorbtion line in the smaller bandwidth, so it's probably the theoretical blackbody (Plank) curve's intensity. The absorbtion line would result in a far greater a difference, and it is the absorbtion line which provides the "window" to see the relatively faint glow of the chromosphere.

To carry the analogy further: outside of this window, opaque "shutters" are needed to block the side band energy - this is what narrow band solar H alpha filters do - completely block the side band energy from the photosphere and allow us to see into the window where a candle is burning. Broad band filters let a portion of this side band energy through.

Using a white light filter to reduce the side band energy is like pulling a dark shade over the side of the house - it greatly reduces the side band energy, but also prevents us from viewing through the window to see the candle.

[MikeBu]

This site has good data on the solar spectrum, with resolution to 0.002 Angstroms in the region of H-alpha. The data is given as a spectral density, total power in a bandwidth is easy to calculate by numerical integration of the data.

http://bass2000.obsp...pect.php?step=1

This data shows the absorbtion line for H-alpha has a spectral density of about 16.04% the density of the local maximum, the local maximum is approx. the black body emission according to Plank's Law. So the intensity dips a little less than an order of magnitude. For the Calcium lines the dip was deeper, somewhere between 4-5% if I remember correctly.

When you say the dip isn't deep enough, I think the data is showing the H-alpha emission spectrum of the excited hydrogen in the chronosphere in conjunction with the emission dip from absorbtion and not just the absorbtion. The photosphere emission is blocked to a much greater degree and then scattered with the result in the data being that portion of the energy scattered in our direction.

[colinsk]

Thank you Mike! That is great!

[colinsk]

Figure 3. in this link show a nice graphical overlay of the dip in power at h-alpha with a .25A Lyot filter transmission curve overlayed. Note the bottom of a Fabry-Perot etalon would be much wider at this same passband.

http://dot.astro.uu.....6269E..12B.pdf

[DesertRat]

The table of data Mike provided shows clearly what many solar observers know already from experience. Hopefully others will read this thread and think about it before asking the question once again or worse, trying something potentially dangerous. The use of a white light filter for an ERF is just ignoring the data presented above.

The only way a wide H-alpha bandpass filter (with a red glass filter as an ERF) could reveal any prominences would be with a coronograph type design with a stop at the focus to shield the solar disk. For disk details it would be futile as well as dangerous. Many H-alpha nebula filters may be to wide even for this unless you are blessed with exceedingly deep blue skies.

I think the best thing to recommend to new experimentors is to refer them to people who have designed and built solar telescopes from relatively inexpensive materials such as Dave Groski. We can also recommend saving up for an entry level solar telescope or sharing the experience with a local solar observer.

Glenn

[colinsk]

Here is another great diagram of the H-alpha dip in power compared to different passband widths.

http://www.brayebroo...t-up/c-line.jpg

[Mark Strollo]

For the sake of argument, can we go the other way, and ask if a Ha etalon and/or blocking filter would be useful for night time viewing of Ha nebula emissions with a telescope?

Thanks for all the explanation in this thread. Very useful.

Mark Strollo
Lunt DS60T on order

[colinsk]

I can think of two reasons that it may not be as good as a nighttime filter for nebula. One is the transmission of a h-alpha solar filter is theoretically between 5% and 85% of the H-alpha line. I have not measured any but I would be surprised if any were over 50%. Although if money were no object one could be designed.

Also, even fast moving solar materials are blue shifted out of the narrow passband so if there is any significant relative motion it could easily be blue or red shifted from view. A spec quoted for solar viewing is you will find h-alpha detail 2A from center in the blue and .5 from center in the red.

I could be totally wrong about this, but when I got my "nebula filter" from Lumicon, 15 years ago or so, the idea was it removed the lines caused by streetlights leaving some important emission lines. I think OIII and H-alpha were among them. And with my filter I have noticed no benefit when in a dark sky location however in my backyard in the middle of a small city I find a large benefit.

Now I have read of astronomers using monochromators to study important aspects of distant objects. Many of these monochromators are etalon based so it could be that there is a good use for a h-alpha filter if you had a very large amount of aperture.

[moron392]

what about 'tuning' the filters as well as using an energy rejection filter???

[colinsk]

The red continuim can be a very good place to look for granulation on the photosphere. This is not h-alpha viewing which looks at the chromosphere! The set up is a normal whitelight filter with a 7nm or so H-alpha filter. See this thread:

http://www.cloudynig...5/o/all/fpart/1

Very nice work!

[mischief]

As a prospective solar observer with a Lunt LSTB600 on order, I want to thank all of you for the terrific information on solar observing. I know a little about the electromagnetic spectrum but these posts really added to my knowledge and greatly needed by me and helpful to me. Thanks very much.

[nytecam]

There's a lot of data in this thread but the unlaying principle is that narrow-band deepsky filters = very broad-band filters in solar terms and thus protentially dangerous for solar viewing. This point can't be overstressed - someone may loose their eyesight through confusion and ignorance :o