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

f-ratio and Ha filter wavelength

  • Please log in to reply
27 replies to this topic

#1 StrStrck

StrStrck

    Apollo

  • *****
  • topic starter
  • Posts: 1,171
  • Joined: 19 Mar 2017
  • Loc: Denmark

Posted 28 September 2017 - 08:35 AM

I see Baader has a special Ha filter for f/2. My Newt is not as extreme with f/3.8. I also see a line of Ha filters from Baader with differing wavelengths. I don't know the connection between "scope speed" and filter wavelength.

 

I have my eye on their 7nm, would that do? My Newt is 8" using coma corrector. I would put a ZWO 1600 mm cool behind it.


Edited by StrStrck, 28 September 2017 - 08:37 AM.


#2 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 08:49 AM

I image at f/4 with a set of 3nm AstroDon filters. I have not had any problems.

 

The problem with very fast scopes (faster than f/3) is spectral shift. The bandpass of the filter starts to shift, which can result in the band being passed at a lower transmission rate, or not being passed at all. Some filters are explicitly designed to handle this to a degree...either with a wider bandpass, or by placing the band in the bandpass such that as the pass starts to shift, the band must first pass over the peak transmission of the pass. This allows some filters to handle a small amount of spectral shift without any loss at all. 



#3 StrStrck

StrStrck

    Apollo

  • *****
  • topic starter
  • Posts: 1,171
  • Joined: 19 Mar 2017
  • Loc: Denmark

Posted 28 September 2017 - 08:58 AM

Thanks Jon. From 3 to 7 nm, is that a lot of difference?



#4 ManuelJ

ManuelJ

    Surveyor 1

  • -----
  • Posts: 1,528
  • Joined: 19 Dec 2005
  • Loc: Madrid, Spain

Posted 28 September 2017 - 10:09 AM

F/3.8 and 7nm, no problem. Even 5nm would be perfect.



#5 terry59

terry59

    Cosmos

  • *****
  • Posts: 9,577
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 28 September 2017 - 10:36 AM

Thanks Jon. From 3 to 7 nm, is that a lot of difference?

Since the purpose of narrower filters is to increase contrast, what is acceptable to you? I believe that for most imagers 7nm in Ha is narrow enough. For OIII I would prefer a more narrow filter than the 8.5 nm Baader I have but I compensate by being careful of the conditions when I use it


Edited by terry59, 28 September 2017 - 10:36 AM.

  • Goofi likes this

#6 Goofi

Goofi

    Cosmos

  • *****
  • In Memoriam
  • Posts: 8,137
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 28 September 2017 - 11:22 AM

+1 to what Terry said



#7 StrStrck

StrStrck

    Apollo

  • *****
  • topic starter
  • Posts: 1,171
  • Joined: 19 Mar 2017
  • Loc: Denmark

Posted 28 September 2017 - 11:22 AM

Yes, I have to start gaining my own experiences. So if 7nm will do, then I´ll see if I can get that and the camera, I´m looking for a good deal on the 1600 mm-c hoping to make the purchase this fall.

 

Thanks everyone!waytogo.gif



#8 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 12:11 PM

How much the bandpass matters is going to depend a bit on where you are imaging. At a dark site, a 7-8nm is fine, you'll get high contrast regardless. In a red or white zone, you might want to consider narrow bandpasses, at least for OIII and SII. I use 3nm filters myself, for all bands, in a red/white zone border. Even with the narrow bandpass, I've noticed that my background skies and data is not as contrasty as people using 5nm filters around an orange/yellow zone, and I often have more noise. So don't underestimate the value of using narrower filters...however, keep an eye on cost as well. A 3nm filter is quite expensive, and may not really be necessary unless you are a stickler for detail.
  • Dean J. likes this

#9 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • Posts: 7,388
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 28 September 2017 - 12:13 PM

I image at f/4 with a set of 3nm AstroDon filters. I have not had any problems.

 

The problem with very fast scopes (faster than f/3) is spectral shift. The bandpass of the filter starts to shift, which can result in the band being passed at a lower transmission rate, or not being passed at all. Some filters are explicitly designed to handle this to a degree...either with a wider bandpass, or by placing the band in the bandpass such that as the pass starts to shift, the band must first pass over the peak transmission of the pass. This allows some filters to handle a small amount of spectral shift without any loss at all. 

 

Jon,

You've got the reason right but without being able to analyze the filter design, it's hard to know if you are getting all of the signal that you think you are getting--particularly with such a narrow bandpass filter.  At F/4, the angle of the marginal rays at the very edge of the pupil are tilted by 7.13 degrees and that may be approaching (and even exceeding) the angle where you may be blocking light.  Remember that the field angle adds directly to that angle so some of the beam will be tilted even more at the edge of the field.  Using really narrowband filters on a fast system will always work; but, it may not always pass as much light as it should.  In that case, you won't know if there's a problem!

 

I'll be at AIC so I'll try to talk to Don Goldman about this issue.  He'll know the angular sensitivity and I'm interested to know how sensitive his 3 nm filters are to tilt.

 

John


  • WesC likes this

#10 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 04:19 PM

 

I image at f/4 with a set of 3nm AstroDon filters. I have not had any problems.

 

The problem with very fast scopes (faster than f/3) is spectral shift. The bandpass of the filter starts to shift, which can result in the band being passed at a lower transmission rate, or not being passed at all. Some filters are explicitly designed to handle this to a degree...either with a wider bandpass, or by placing the band in the bandpass such that as the pass starts to shift, the band must first pass over the peak transmission of the pass. This allows some filters to handle a small amount of spectral shift without any loss at all. 

 

Jon,

You've got the reason right but without being able to analyze the filter design, it's hard to know if you are getting all of the signal that you think you are getting--particularly with such a narrow bandpass filter.  At F/4, the angle of the marginal rays at the very edge of the pupil are tilted by 7.13 degrees and that may be approaching (and even exceeding) the angle where you may be blocking light.  Remember that the field angle adds directly to that angle so some of the beam will be tilted even more at the edge of the field.  Using really narrowband filters on a fast system will always work; but, it may not always pass as much light as it should.  In that case, you won't know if there's a problem!

 

I'll be at AIC so I'll try to talk to Don Goldman about this issue.  He'll know the angular sensitivity and I'm interested to know how sensitive his 3 nm filters are to tilt.

 

John

 

Interesting. If any light was blocked, would that exhibit in some way...say as increased vignetting?

 

My NB sub fields are very flat. There is veyr slight vignetting in the corners, however I get that with every filter, and I actually think it is due to the lens hood on my 600mm f/4 lens. But otherwise, I usually calibrate my nb subs sans flats, only with a master dark, and they come out pretty darn flat. Flattest native lights I get. The broader band LRGB subs have significantly more vignetting. 

 

Anyway, would be interested in hearing what Don has to say. I mentioned f/3, because I read on the AstroDon site that his 3nm filters are good down to f/3. 



#11 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 05:32 PM

There will be variability in the exact central passband of each filter, and even some variability of that passband across the x,y location of the filter.  If you just image a single object you wouldn't be able to tell if any drop off of signal is due to the true signal or an artifact of the filter passband.

 

So one test would be to image a nebula in the center of the field - and then offset by 1/2 field.  After flattening each image the overlap region of the nebulosity should look identical.  If it isn't then you know you are losing light from the edge of the pupil - and it will have a field dependence.  This would be especially bad for mosaics - but for a single image - again you may not notice or realize it at all.

 

Another problem for Ha is that you might have the Nii signal at the edge of the passband, so that it is included in the center of the field but diminishes as you move away - even with the Ha coming in strong.  So for a planetary nebula it might look very different in the center of the field vs. at the edge.

 

At the same time, the sky background will increase linearly with the passband - but the sky background shot noise, which is what matters, will increase with the square root.  So going from 3nm to 5nm will only have a sky background noise increase of 29%.  This isn't very much by itself - and if you are actually losing nebulosity signal, you could have a net loss of SNR.  Going from 7nm to 12nm is a 31% increase - again not huge.

 

The Baader filters are designed not only with a wider passband, but it is shifted so the rays coming in at an angle will still be in the passband.  Those filters are designed to work at f/2 - and the passband is I think around 12nm or more.  I think you can be confident they won't lose signal at f/2, and for other narrower filters they will probably "work" at f/3 or f/4, but there may be problems and loss of signal that are happening but you don't notice in a single image.

 

Frank



#12 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,322
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 28 September 2017 - 05:39 PM

I'm using Astrodon 3nm filters @f/3 due to extreme light pollution where I live. I had an issue with my first OIII filter, that I eventually replaced. It was a sort of weird dark banding that I could not get rid of.

 

I would love to be able to use a wider bandpass and always image from a dark site, but that's just not practical, so I make due.

 

I'm very curious to hear what Don says about his 3nm filters, though I doubt it will be anything but a glowing report. ;)



#13 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 06:01 PM

Another test is to put an annular mask on the front of the 'scope to reduce its diameter and increase the f/ratio.  The corresponding nebulosity signal should decrease according to 1/FNum^2.  If it doesn't decrease as much as expected, it means the outer part of the pupil isn't contributing signal - and that would be bad.

 

Make sure to subtract off the sky background signal when measuring the nebulosity signal.  A small planetary with a dark sky background would be good for this - and you could try it in the center and on the edge as a test of both extreme field angle and pupil angle.

 

You don't even need to apply a flat and calibrate for this test.  Just measure the nebulosity signal at the edge of the field with and without the annular mask - using the raw exposure - relative to a sky background measurement next to the nebula.

 

Frank


Edited by freestar8n, 28 September 2017 - 06:06 PM.


#14 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2,289
  • Joined: 08 Apr 2015

Posted 28 September 2017 - 06:12 PM

Jon Rista, on 28 Sept 2017 - 5:19 PM, said:

 

jhayes_tucson, on 28 Sept 2017 - 1:13 PM, said:

 

Jon Rista, on 28 Sept 2017 - 09:49 AM, said:

I image at f/4 with a set of 3nm AstroDon filters. I have not had any problems.

 

The problem with very fast scopes (faster than f/3) is spectral shift. The bandpass of the filter starts to shift, which can result in the band being passed at a lower transmission rate, or not being passed at all. Some filters are explicitly designed to handle this to a degree...either with a wider bandpass, or by placing the band in the bandpass such that as the pass starts to shift, the band must first pass over the peak transmission of the pass. This allows some filters to handle a small amount of spectral shift without any loss at all. 

 

Jon,

You've got the reason right but without being able to analyze the filter design, it's hard to know if you are getting all of the signal that you think you are getting--particularly with such a narrow bandpass filter.  At F/4, the angle of the marginal rays at the very edge of the pupil are tilted by 7.13 degrees and that may be approaching (and even exceeding) the angle where you may be blocking light.  Remember that the field angle adds directly to that angle so some of the beam will be tilted even more at the edge of the field.  Using really narrowband filters on a fast system will always work; but, it may not always pass as much light as it should.  In that case, you won't know if there's a problem!

 

I'll be at AIC so I'll try to talk to Don Goldman about this issue.  He'll know the angular sensitivity and I'm interested to know how sensitive his 3 nm filters are to tilt.

 

John

 

Interesting. If any light was blocked, would that exhibit in some way...say as increased vignetting?

 

My NB sub fields are very flat. There is veyr slight vignetting in the corners, however I get that with every filter, and I actually think it is due to the lens hood on my 600mm f/4 lens. But otherwise, I usually calibrate my nb subs sans flats, only with a master dark, and they come out pretty darn flat. Flattest native lights I get. The broader band LRGB subs have significantly more vignetting. 

 

Anyway, would be interested in hearing what Don has to say. I mentioned f/3, because I read on the AstroDon site that his 3nm filters are good down to f/3. 

 

Central wavelength shift can be calculated from:

 

CWS = target_wavelength[sqrt(n^2 - sin^2(theta))]/n

 

Where n is the refractive index of the filter, theta is the marginal ray angle, and target_wavelength is the wavelength that is shifted. For an f/4 scope, theta is 7.13. Using Ha as the target_wavelength (656nm) and assuming a refractive index of about 1.6, you'll get a CWS of about 1 nm.  Assuming that the 3nm filter is centered about the Ha wavelength, you should see no loss in transmission.

 

Tim


  • Jon Rista likes this

#15 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 06:17 PM

Central wavelength shift can be calculated from:
 
CWS = target_wavelength[sqrt(n^2 - sin^2(theta))]/n
 
Where n is the refractive index of the filter, theta is the marginal ray angle, and target_wavelength is the wavelength that is shifted. For an f/4 scope, theta is 7.13. Using Ha as the target_wavelength (656nm) and assuming a refractive index of about 1.6, you'll get a CWS of about 1 nm.  Assuming that the 3nm filter is centered about the Ha wavelength, you should see no loss in transmission.
 
Tim


This ignores the field angle, corresponding to the size of the sensor and the edge of the field. It also assumes the central wavelength is exact - but we have no specs on how accurately that value is maintained.  So not only is it hard to know the min f/ratio for a given sensor (especially since sensor size also plays a role) - but you can't confidently use one person's results to predict your own - if the central wavelength is shifted by +0.5nm in one case, and -0.5 nm in another case, for example.

 

Frank



#16 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2,289
  • Joined: 08 Apr 2015

Posted 28 September 2017 - 06:39 PM

freestar8n, on 28 Sept 2017 - 7:17 PM, said:

 

spokeshave, on 28 Sept 2017 - 7:12 PM, said:

Central wavelength shift can be calculated from:
 
CWS = target_wavelength[sqrt(n^2 - sin^2(theta))]/n
 
Where n is the refractive index of the filter, theta is the marginal ray angle, and target_wavelength is the wavelength that is shifted. For an f/4 scope, theta is 7.13. Using Ha as the target_wavelength (656nm) and assuming a refractive index of about 1.6, you'll get a CWS of about 1 nm.  Assuming that the 3nm filter is centered about the Ha wavelength, you should see no loss in transmission.
 
Tim


This ignores the field angle, corresponding to the size of the sensor and the edge of the field. It also assumes the central wavelength is exact - but we have no specs on how accurately that value is maintained.  So not only is it hard to know the min f/ratio for a given sensor (especially since sensor size also plays a role) - but you can't confidently use one person's results to predict your own - if the central wavelength is shifted by +0.5nm in one case, and -0.5 nm in another case, for example.

 

Frank

 

It ignores what you call the field angle because it is irrelevant. The wavelength is shifted by the filter and that has nothing to do with sensor size. Additionally, for a filter with a positive refractive index (all of them), the central wavelength shift is always toward a shorter wavelength. You can, quite confidently, predict the maximum CWS with this equation.

 

Having said that, you need to apply the equation correctly. I must have keyed something incorrectly. For Jon's f/4 system at 656nm, the central wavelength is shifted to 654nm, not 655. So the CWS is 2nm, not 1. So he may be getting some transmission loss. All of this is predicated on the correct value for the refractive index since the CWS is quite sensitive to small differences in refractive index.

 

Tim



#17 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 07:36 PM

?  A wide field system will have a ray from the left side of the pupil making a much steeper angle to reach the right side of the sensor.  Of course the field angle matters...

 

And because the field angle matters, if you are close to the edge of the bandwidth in the center of the field, the increasing loss of signal as you go out will be insidious and not directly noticeable - because the sky background is coming in at full strength - and flats won't correct for the loss of nebulosity signal.

 

I did not check the math in your equation - but it does need to include the field angle - especially for a fast and wide field system like hyperstar or rasa with a large sensor.

 

The main thing is - I don't see Astrodon providing a spec. on how accurately centered the passband is - and it should be spec'd for the entire filter area and not just the center.  A typical spec. might be +/- 10% of the fwhm - which for 3nm would be 0.3nm.  For 3nm fwhm, only the very center is at full transmission - so if you are already 0.3nm off center there would be some loss.  And as you continue away from that the loss would increase much earlier than expected, compared to a perfectly centered passband and a rectangular rather than semi-Gaussian profile - and not including the field angle.

 

These things are hard to make accurately when the passband is narrow - and even harder with a large and square 50mm filter - if you want it consistent across the whole filter.  That's why they cost a lot.  So they are not expected to be an exact match to the spec - and more info is needed to know how well they would actually perform with a fast and wide field system.

 

Frank

 

Addendum - for a 280mm f/2.2 RASA with 43mm corrected image diameter, I get a shift of 7.2 nm in the center of the field and 9.4 nm at the edge of field.  So - assuming a perfectly centered filter, 15nm bandpass would have full transmission in the center of the field but possibly significantly reduced transmission at the edge of the field, due to one side of the pupil shifting out of the passband.

 

But the Baader fast filters are designed with the nebula wavelength intentionally shifted to the right in the passband - which leaves more room on the shorter wavelength side to accommodate these steeper rays.  So 15nm may be fine for RASA across the full field.


Edited by freestar8n, 28 September 2017 - 08:31 PM.


#18 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 08:43 PM

This is what I was originally referencing when I bought my AstroDon 3nm filters:

 

http://astrodon.com/...rrowbandfaq.pdf

 

He notes loss up to 20% at f/3, but doesn't get down to field angle or anything like that. 



#19 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 09:04 PM

This is what I was originally referencing when I bought my AstroDon 3nm filters:

 

http://astrodon.com/...rrowbandfaq.pdf

 

He notes loss up to 20% at f/3, but doesn't get down to field angle or anything like that. 

There are several things missing in that write up.  First he doesn't say that the background noise reduction involves a square root - so the improvement from 5 to 3nm is less than you might expect.

 

Second, he assumes the filter is centered perfectly - and he doesn't provide a spec on how well centered they are guaranteed to be.  If there is some variation in the central wavelength then 20% at f/3 won't be the result everyone gets.

 

Third, he doesn't mention the dependence on field angle.  As far as I know I'm the only one to point this out.  It will play a role in how much loss of signal there will be at the edge of the image.  It will also play a role in how much Nii signal you get across the field - again as a function of f/ratio and field angle.

 

I have seen people using very narrow filters with wide and fast systems - and they show an image of a nebula and say, "It works."  But you can't tell from the image how much loss you have - and how much it varies across the image.

 

Frank



#20 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 09:11 PM

I believe I have an email from Don where he says he places the band at the far end of the bandpass. So the band isn't dead centered, he purposely places it off to the side so that as spectral shift occurs, the band shifts through the range of maximum transmission before transmission loss begins to occur. 


Edited by Jon Rista, 28 September 2017 - 09:11 PM.


#21 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 09:20 PM

I believe I have an email from Don where he says he places the band at the far end of the bandpass. So the band isn't dead centered, he purposely places it off to the side so that as spectral shift occurs, the band shifts through the range of maximum transmission before transmission loss begins to occur. 

Well - the plot in his write up shows Ha pretty well centered - but maybe a tiny bit to the left of center in the passband - which is good.  But one of the Nii lines is right at the left edge of the passband - so it will pull in Nii in a field-dependent way.  This doesn't matter if you don't have Nii in the first place - but it's another thing missing from his write up.

 

It's just really hard to predict the loss with f/ratio and field angle if you don't know exactly what the central wavelength is, and how much variation to expect from filter to filter.  But it should be easy to measure as I describe above.

 

Frank

 

Addendum-  M27 and M57 have very bright sections dominated by Nii - so you could get a sense of how much Nii is getting through by placing those parts in the middle or edge of the image - and/or adding an annular mask.  See how the change in Nii differs from the change in Ha.


Edited by freestar8n, 28 September 2017 - 09:31 PM.


#22 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 09:32 PM

I believe I have an email from Don where he says he places the band at the far end of the bandpass. So the band isn't dead centered, he purposely places it off to the side so that as spectral shift occurs, the band shifts through the range of maximum transmission before transmission loss begins to occur.

Well - the plot in his write up shows Ha pretty well centered - but maybe a tiny bit to the left of center in the passband - which is good.  But one of the Nii lines is right at the left edge of the passband - so it will pull in Nii in a field-dependent way.  This doesn't matter if you don't have Nii in the first place - but it's another thing missing from his write up.
 
It's just really hard to predict the loss with f/ratio and field angle if you don't know exactly what the central wavelength is, and how much variation to expect from filter to filter.  But it should be easy to measure as I describe above.
 
Frank


Are you looking at the blue plot or the black plot? The black plot is 6nm, includes both the NII emissions, with the brighter one right at the end of the 6nm peak pass. The 3nm one puts both NII emissions right at the tails of the pass, so transmission is very low, and as the pass shifts, it shifts away from the brighter emission and over the much weaker emission. So even if there is a little bit of NII passed during spectral shift, it's of the weaker of the two NII emissions.

#23 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 09:37 PM

Yes that's the plot I'm referring to and yes it is a weaker Nii line.  But with a small shift to the left you could end up with almost no Ha and 100% of that weaker Nii line.  Which means for M57 and M27, the only signal you would get is from those really bright Nii patches.

 

Of course for rays on-axis in the middle of the field you would still be getting Ha - but when you consider the variation of the passband across the pupil, along with the variation across the image - and with lines near each other and getting blocked here and boosted there - it's a pretty complicated scenario.  And you can't tell what is going on from a single image that appears to show a perfectly normal looking nebula.

 

Frank



#24 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 9,611
  • Joined: 12 Oct 2007

Posted 28 September 2017 - 09:54 PM

For the OP's question involving 210mm f/3.8 newtonian with ASI1600 and 22mm diagonal, the total shift in the corners of the image would be about 3nm.  So that would point to at least 7nm bandwidth and probably more - if no loss is desired.

 

If you compared background noise in 7nm to perhaps 12nm - there is only a 30% increase, so 12 would certainly be safer and with not much extra noise.

 

But if you aren't doing mosaics and aren't imaging large objects with Nii, then 5nm may be ok - if the central wavelength happens to be offset properly.

 

Frank



#25 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,324
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 September 2017 - 11:04 PM

Yes that's the plot I'm referring to and yes it is a weaker Nii line.  But with a small shift to the left you could end up with almost no Ha and 100% of that weaker Nii line.  Which means for M57 and M27, the only signal you would get is from those really bright Nii patches.
 
Of course for rays on-axis in the middle of the field you would still be getting Ha - but when you consider the variation of the passband across the pupil, along with the variation across the image - and with lines near each other and getting blocked here and boosted there - it's a pretty complicated scenario.  And you can't tell what is going on from a single image that appears to show a perfectly normal looking nebula.
 
Frank


Gocha. So certainly a more complex problem than simply a shifting bandpass.


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