44 replies to this topic

[Souldrop]

M57:

 

ETA: Deleted as image is under copyright. Posted briefly for illustrative purposes. Please refer to "Spectroscopic Atlas for Amateur Astronomers - A Guide to the Stellar Spectral Classes" by Richard Walker for a copy of the emission plot

Edited by Souldrop, 04 September 2024 - 09:46 PM.

[ButterFly]

And how much more h-alpha emission is there than NII in these nebulae?

 

It's much easier to see ripples from a stone tossed into a still lake, than it is to see the ripples from that same stone tossed into a raging sea during a storm.  The peak to trough of those ripples may be the same, but the raging seas are what makes it more difficult to make out those ripples during the storm.

 

In much the same way, detecting the differences in brightness depends very much on the background.  With NII, not only is there sky background to contend with, but the enormously larger h-alpha emissions as background as well.

 

I have been eagerly awaiting observational reports with SII filters and intensifiers.  They are hard to find, even though SII emissions are orders of magnitude higher than NII emissions.  For some nebulae, SII can be comparable to h-alpha, though still much, much lower.

[Souldrop]

NGC 6888:

 

ETA: Deleted as image is under copyright. Posted briefly for illustrative purposes. Please refer to "Spectroscopic Atlas for Amateur Astronomers - A Guide to the Stellar Spectral Classes" by Richard Walker for a copy of the emission plot

 

 

I did the footwork. Feel free to discuss or point out issues. Images aren't mine, so I will delete in a few hours. My argument is only that narrow filters run the risk of cutting out signal. I would assume anyone who's done narrowband imaging near h-alpha would determine that a narrow enough filter will at the very least cutout NII emissions that contribute to the signal. Target dependent of course. Although some emission spectra seem to show wider profiles around the h-alpha line. 

Edited by Souldrop, 04 September 2024 - 09:46 PM.

[sixela]

Mhh -- would you care to tell us what the resolution of these spectrometers were?

 

As I said, the FWHM of emission lines of nebulae measured with really high resolution isn't a dark secret (and there is indeed a contribution of turbulence that is different from that due to mere temperature). I posted it already, I'm not going to repeat it.

 

Of course you are indeed right that if you consider SII and NII "signal" then a very narrow H-alpha filter isn't going to catch those (although SII is bloody far from H-alpha). That's why it's not called an "emission nebula filter".

 

For a H-alpha filter H-alpha is signal and the rest is noise (by definition). Whether the H-alpha filter is the best filter for an object is another matter, so forgive me for thinking that you're just moving the goalposts slightly (especially since we were discussing your observations of M8 and M20 and not M57 and NGC 1068.)

 

A wider H-alpha filter isn't going to catch those OIII lines either, while my dualband OIII/H-alpha is (it definitely helps on some planetary nebulae, and some high surface brightness ones are even better with IDAS LPS-D3 or Astronomik L1).

 

As for NGC 1068, isn't that a galaxy? Needless to say, that's even less usually what you use narrowband filters for. Of course it could be a typo too, you paste graphs on here (and say you're going to delete them) without posting references.

Edited by sixela, 04 September 2024 - 05:33 PM.

[ButterFly]

Reading spectra takes more than just looking at them.  Applying them to observation takes much, much more.

 

Here are some representative spectra of M8 (the lagoon nebula), with a cutout around h-alpha:

 

https://stellarscene.../spectle/m8.htm

 

 

The most important question when applying this spectrum to observation is: where is the background level?

 

When theories are in discord with observation, that really only means one of two things.  The observations are all wrong, or the theories are incomplete.

[sixela]

I'd bet that spectrometer didn't have a FWHM resolution of 0.4nm, Butterfly. The real width is probably a lot narrower...from the text the FWHM of the resolution seems to be around 2.35*0.8 nm or roughly 1.8-2 nm, so the line width is dominated by the spectrometer resolution.

 

The illustration does go to show that M8 is a H-alpha object (or H-alpha/OIII) while Souldrop's graphs tell us some planetary nebulae aren't. No big surprises there.

 

H-beta is indeed added "signal" for us and is not considered as such by even OIII/H-alpha dualbands, but UHC/H-alpha filters have to have really large passbands that aren't always helpful, and the regions that radiate in H-beta also radiate in H-alpha (which is why astrophotographers haven't been clamouring for dual band H-alpha/H-beta either).

Edited by sixela, 04 September 2024 - 05:38 PM.

[Souldrop]

Regarding resolution you are correct sixela.


As far as moving the goal post I disagree. More the game changed. My initial observations were on different targets that I conceded my initial musings were incorrect.

Afterwards statements were made that a narrow h-alpha filter never reduces signal from the target. That I can’t agree with (at least working from the assumption we as NV user are only after increased contrast.) Some
mechanism appears at play broadening the profile of some objects around the h-alpha line. By most arguments the emission lines would be extremely narrow and should be confined in the posted plots. Other plots show the expected peak around h-alpha.

ETA: I do actually really appreciate the constructive and informational post by other people on the matter…I wanted to point that out while I was thinking about. This thread made me think and taught me some things.

Note: Just noticed an edit. They weren’t meant to be nameless ninja plots. I had posted the source in another post that I decided to delete in favor of posting the plots directly. Apologies I forgot to add it back “Spectral Atlas for Amateur Astronomers” by Richard Clark. Apologies Alexis I don’t edit ad nauseam like you do. I would likely have less typos. :-)

Edited by Souldrop, 04 September 2024 - 06:57 PM.

[sixela]

My initial observations were on different targets that I conceded my initial musings were incorrect.

We seem to have changed targets without it having been clearly stated. 
 

 

Afterwards statements were made that a narrow h-alpha filter never reduces signal from the target.

That was still in the context of your earlier post. Of course if you change the targets to targets with signal outside of the H-alpha band then a H-alpha filter does reduce the signal. That's why it's called "a H-alpha filter". After all, the thread is called "H-alpha narrowness".

For continuum emitters like galaxies and planetary nebulae, I'd certainly agree that the views through other filters than narrowband H-alpha can be better, that's a non-controversial statement.

 

But on your examples a dual band filter would be a lot better than a wider narrowband H-alpha filter, and for some planetary nebulae a "light pollution" filter with some narrow blocking bands (rather than some narrow passbands) can be even better; at higher magnifications for the brighter ones a "visual spectrum" one can even be better than those (for PNs my filter slide tends to have an Astronomik L1, an IDAS LPS-D3 and a dualband OIII/H-alpha; that is also the filter setup for galaxies) . I also have a UHC with a red passband and I'd use that but frankly I ran out of filter slide slots ;-).
 

 

I don’t edit ad nauseam like you do

Guilty as charged. You should see me moderate another astro forum ;-).

Edited by sixela, 05 September 2024 - 04:10 AM.

[PEterW]

Sounds like someone should try making a “filter tuner” like this one https://arxiv.org/html/2406.17979v1 properly reject out of band and redshirts as required.

Peter

[ButterFly]

Dragonfly is awesome for many reasons.  But they actually get to subtract background, on top of the mitigating by filtering that we get to do.

[WheezyGod]

Appreciate all the responses although most of it is going over my head. I do understand the earlier point about SNR playing a key role and the idea of doing a comparison of two devices with very similar specs, but with different EBIs being difficult to compare fairly in practice.

Instead, let’s imagine two devices that are exactly the same. One has an EBI = 0.0 and the other has an EBI = 1.0. Both are being used to observe Ha nebula under the same sky conditions at a site with average light pollution at a temperature of 72degrees F. The two Ha filters are 3nm and 7nm.

If I’m someone who mostly cares about seeing as much Ha detail as possible, at the expense of a view that’s a bit abnormal (more muted stars, a bit more noise), then I would expect I’d prefer the view with a 3nm filter in each device.

However, let’s shift this example to a very dark site where there’s typically a preference amongst most here for a less narrow filter. With the 1.0 EBI device I’d expect the additional gain in ha detail from a 3nm filter vs. 7nm to be very limited because there’s little light pollution to cut out. The contrast reduction from EBI is probably more impactful than the light pollution. With the 0.0 EBI, device, the difference between a 3nm vs. a 7nm filter is probably still small to the point where a 7nm filter would be better, but I would expect there to be more of a difference.

Edited by WheezyGod, 05 September 2024 - 02:55 PM.

[ButterFly]

The contrast reduction from EBI is probably more impactful than the light pollution.

And what are you basing that on?  EBI means effective background illumination.  Contrast means, for example, (signal - background) / background.  What about that other tube is giving it a higher EBI?  One can have the same contrast with that higher EBI tube as with that lower EBI tube.

 

Contrast to noise is different from contrast.

 

Again, it's a matter of taste for given ability:

 

 

 

This is a matter of taste.  Even though there will always be lower contrast to noise ratio with decreasing bandpass, for a given device in a given setup, whether that matters for you depends solely on you.  Some don't mid the fizzles as much as other do.  Some can "see through" the fizzle better than others can.  If you don't like the results of a given bandpass with your device, when used in some particular setup, don't use it.  If you find yourself never using it, get rid of it.

 

[WheezyGod]

And what are you basing that on? EBI means effective background illumination. Contrast means, for example, (signal - background) / background. What about that other tube is giving it a higher EBI? One can have the same contrast with that higher EBI tube as with that lower EBI tube.

Contrast to noise is different from contrast.

Again, it's a matter of taste for given ability:

More of a guess that at a very dark site a 1.0 difference in EBI is more impactful than the light pollution at such a site.

In a tube with a higher EBI, wouldn’t the sky background be less black and more grey compared to a device with a lower EBI? If yes, then whether it’s “technically” contrast or contrast to noise it’s still a reduction in what can be seen.

I agree with your last point, that seeing more details vs. the negative impact more narrow filters can have is preference. I was thinking about if someone like myself or Gavster who prefers more narrow filters (even at darker sites) would have less or the same preference for them if we used a device with a higher EBI. How dark a site is more important than EBI for preferences on ha filter narrowness, but I think it could be the combination of how dark a site is and EBI.

It would be much easier to get a clearer answer if we could quantify this. Let’s say a very dark site has a light pollution value of 100, and an EBI of 1.0 equates to a value of 120. A 3nm filter might cut out 99% of the LP while the 7nm filter cuts out 97%. Here the difference between the two filters on ha detail might be imperceptible so most if not all would prefer the 7nm.
At a moderate LP site that has a value of 10,000, the difference in LP blocked between a 3nm and a 7nm is larger than the 120 EBI value and the difference in ha detail is more apparent. The device with the 1.0 EBI has some of the increase in ha detail lessened making the 3nm filter less incrementally impactful compared to the 7nm. Very easy to criticize any/all of the above values, but does the thinking make sense directionally?

[Thierry Legault]

 

My argument is only that narrow filters run the risk of cutting out signal. I would assume anyone who's done narrowband imaging near h-alpha would determine that a narrow enough filter will at the very least cutout NII emissions that contribute to the signal.

 

this discussion is a bit confusing. First, you said that a loss of signal could come from Doppler effect (target speed or molecular speed), and several of us replied that it's negligible with deepsky narrowband filters (not with solar Ha filters which are typically 50 to 100 times narrower, but it's another story).

If now you say that a narrow Ha filter (eg 3nm) would cancel NII (2nm away from Ha) or SII (16nm from Ha) lines, yes of course. The remaining question is: how intense (or faint!) are those lines compared to Ha for most targets? 

[sixela]

In a tube with a higher EBI, wouldn’t the sky background be less black and more grey compared to a device with a lower EBI?

The EBI "background" is neither uniform nor static (just look at a capped NVD in a dark room to see it). In practice it's fairly easy to recognize when it starts having an effect (at high temperature, with very narrow filters and at higher effective f/ratios). Unfortunately that doesn't depend only on the numeric value of the EBI but also its spatial and temporal characteristics.

 

Once it's there --and that's where it's hard to quantify-- sometimes raising both the background and the object surface brightness (which is at the cost of inherent contrast) can indeed subjectively "quieten" down the noise in the background somewhat and make objects easier to see again despite the lower contrast. But then so can playing a bit with the gain.

 

I've given up trying to quantify it, but I know my rules of thumb° derived from experience; in summer above f/4-5 I tend to switch over to a wider filter (at a dark site a dual-band 4 nm and at a less dark site a 6 nm H-alpha) and above f/8 an even wider one (a 12 nm dual band).

Some of it is also due to the objects you tend to look at at these f/ratios; I'd tend to filters at a high f/ratio and thus the NBZ for fairly small but bright planetary nebulae (at a dark site the filter slide setup for these is usually IDAS LPS-D3 - 12 nm dual band - 4 nm dual band, and I use an H-alpha on the eyepiece on the open filter slide slot to compare if necessary). These tend to have higher inherent contrast with the sky background, so really narrow filters are less necessary to preserve contrast.

 

In winter the cutoffs tend to shift somewhat --narrow filters to slightly higher f/ratios--. A really well dark-adapted eye also does in the opposite direction (but if I use the well dark adapted eye I'll set the gain in such a way the background is never brighter than the naked eye sky background with my other eye first).

 

If I know I'm going to look at strong more pure H-alpha targets (e.g. fainter DWBs) then the filter selection is all H-alpha 12-6-4 nm (the 12 nm H-alpha is an Astronomik ProPlanet that nukes the OIII band of an NBZ).

 

For galaxies I tend to have a stack that is Astronomik L1 - IDAS LPS-D3 - 12 nm dual band (the latter serves mainly to highlight HII regions without completely nuking stars).

 

--

°I still don't trust my rules of thumb, BTW...a filter slide means you don't have to. The rule of thumb just documents what I end up with.

Edited by sixela, 06 September 2024 - 03:41 AM.

[Souldrop]

this discussion is a bit confusing. First, you said that a loss of signal could come from Doppler effect (target speed or molecular speed), and several of us replied that it's negligible with deepsky narrowband filters (not with solar Ha filters which are typically 50 to 100 times narrower, but it's another story).
If now you say that a narrow Ha filter (eg 3nm) would cancel NII (2nm away from Ha) or SII (16nm from Ha) lines, yes of course. The remaining question is: how intense (or faint!) are those lines compared to Ha for most targets?

I was skeptical that there are no instances where broadening or doppler shift come into play for extended objects. (Still somewhat am honestly) The claim that started that dialigue was that I preferred the wider filter at 1x for a lower EBI unit because I felt the narrower unit took away too much “signal”. (At that time I didn’t specify it was h-alpha although thats what I assumed). I also don’t think OP ever explicitly stated he was only interest in h-alpha emissions at that point. (If arguments want to now be made about using “h alpha” filters to look at only h-alpha thats fine. It’s a fair point; at the time I didn’t take the discussion to be limited. I learned what I could after this discussion. :-))

Then there were claims made that a narrow filter never reduces signal only background. No qualifiers accompanied those statements . It was easy to forget that was the start of the more spirited dialogue. My changing gears was always in the context of is that claim really true? Again keeping the assumption that I as NV user don’t care where that target signal comes from (h-alpha or NII, admittedly I didn’t have NII in mind during the beginning. I thought other emission lines of note were further away)

The broadening of h-alpha appears to be a red herring for emission nebula. Those were musings and more of a stance against “never”. Again conceded that point before your initial post.

At that point I was thinking well what does the emission spectrum for some sample targets look like? I found instances where it would appear more a narrow filter would cut off target signal and posted some example spectrum.

M57 and NGC6888 are targets that appeared to have wider profiles in the h-alpha region if Richard Walkers data is to be trusted and found sufficient.(ETA: they also show fairly intense NII emission, so maybe that’s secret ingredient for what looks to be h-alpha?) I’m scared to speculate more on it given the general response to my initial speculations (or feelings) lol

At that point WG jumped in clarifying what he was hoping to find. A quantitative comparison between ebi, sky background, and filter. That’s definitely a tougher ask hence my current silence. :-).

ETA: I tried to do some quantitative comparisons a couple years ago with ebi vs sky background. I concluded it was easier to just buy another filter. :-).


Apologies I pulled a sixela with this post. :-p.

Edited by Souldrop, 06 September 2024 - 09:41 AM.

[sixela]

M57 and NGC6888 are targets that appeared to have wider profiles in the h-alpha region

There is some NII and the spectrometer used is not very high resolution (which makes sense for these kinds of targets. The one I used to measure filter would need a lot of aperture to be usable on M57).

 

There is indeed some NII but there is also a lot of OIII; here' s a higher resolution plot from wikimedia:

See https://commons.wiki...spectrum-ru.png

Edited by sixela, 06 September 2024 - 11:00 AM.

[Souldrop]

There is some NII and the spectrometer used is not very high resolution (which makes sense for these kinds of targets. The one I used to measure filter would need a lot of aperture to be usable on M57).

There is indeed some NII but there is also a lot of OIII; here' s a higher resolution plot from wikimedia:

See https://commons.wiki...spectrum-ru.png

That plots looks even coarser resolution than the one referenced above.

At this point I feel the discussion is better shifted towards user perception and preference or in pursuit of WGs request for a more quantitative way to compare two setups.

I think a fair conclusion fore my unabashed tangent is yes a narrow h-alpha filter can cut some valuable NII signal for some extended targets, but those targets are in the minority. H-alpha transmission for most (if not all) targets of interest is largely unaffected. Saying that the great discussions by you and Butterfly on contrast and preference are likely far more applicable to most users.


Eta: yeah agreed on the OIII line. If anything this exercise has piqued my interest to perhaps explore other filters at some point.

Edited by Souldrop, 06 September 2024 - 11:13 AM.

[sixela]

The triband filters like the Altair Astro 34nm/14nm width OIII/H-beta and H-alpha/NII can indeed be useful at a dark site oin planetary nebulae, but often a light pollution filter like an IDAS LPS-D3 or something more narrow in visual but more broadband in H-alpha (an Astronomik UHC, with red passband) are also good on those objects. An Astronomik UHC is surprisingly useful at many sites because there is less light pollution in its red passband.

 

The most fun thing about these wider filters (including visual UHC ones) is that you get to keep some stars.

 

My NBZ (wide enough to pick up NII) also works well but it doesn't add H-beta to the H-alpha, so you tend to lose 20% of the H-alpha+H-beta.

 

I don't think I've ever seen an object where SII is really that significant. Sure, people use SII filters but that's to highlight differences in the regions that emit it vs. those that do not, not really because it can make anything significantly brighter, AFAIK. So I think if you use a quad-band H-beta/OIII+H-alpha/SII I think that's mainly to get more stars, and the good thing about the extra passband to do so is that in the interval [H-alpha,SII] there isn't that much light pollution.

Edited by sixela, 06 September 2024 - 11:20 AM.

[WheezyGod]

The EBI "background" is neither uniform nor static (just look at a capped NVD in a dark room to see it). In practice it's fairly easy to recognize when it starts having an effect (at high temperature, with very narrow filters and at higher effective f/ratios). Unfortunately that doesn't depend only on the numeric value of the EBI but also its spatial and temporal characteristics.

Once it's there --and that's where it's hard to quantify-- sometimes raising both the background and the object surface brightness (which is at the cost of inherent contrast) can indeed subjectively "quieten" down the noise in the background somewhat and make objects easier to see again despite the lower contrast. But then so can playing a bit with the gain.

I've given up trying to quantify it, but I know my rules of thumb° derived from experience; in summer above f/4-5 I tend to switch over to a wider filter (at a dark site a dual-band 4 nm and at a less dark site a 6 nm H-alpha) and above f/8 an even wider one (a 12 nm dual band).

Some of it is also due to the objects you tend to look at at these f/ratios; I'd tend to filters at a high f/ratio and thus the NBZ for fairly small but bright planetary nebulae (at a dark site the filter slide setup for these is usually IDAS LPS-D3 - 12 nm dual band - 4 nm dual band, and I use an H-alpha on the eyepiece on the open filter slide slot to compare if necessary). These tend to have higher inherent contrast with the sky background, so really narrow filters are less necessary to preserve contrast.

In winter the cutoffs tend to shift somewhat --narrow filters to slightly higher f/ratios--. A really well dark-adapted eye also does in the opposite direction (but if I use the well dark adapted eye I'll set the gain in such a way the background is never brighter than the naked eye sky background with my other eye first).

If I know I'm going to look at strong more pure H-alpha targets (e.g. fainter DWBs) then the filter selection is all H-alpha 12-6-4 nm (the 12 nm H-alpha is an Astronomik ProPlanet that nukes the OIII band of an NBZ).

For galaxies I tend to have a stack that is Astronomik L1 - IDAS LPS-D3 - 12 nm dual band (the latter serves mainly to highlight HII regions without completely nuking stars).

--
°I still don't trust my rules of thumb, BTW...a filter slide means you don't have to. The rule of thumb just documents what I end up with.

Thanks your comment about wanting to use a wider filter during the summer tells me EBI does influence ha filter preference. In this case, it’s the warmer temperature’s impact on EBI.