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

Band Shift and Filter Position

Filters NV
  • Please log in to reply
32 replies to this topic

#26 GeezerGazer

GeezerGazer

    Apollo

  • *****
  • topic starter
  • Posts: 1,387
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 08 July 2021 - 02:20 PM

Joe, I'm not sure if you are now talking about H-a filtration or long-pass filtration.  And, without knowing which type of filter, I can't answer your question.  

A thought I had: for slower optics like f2.8, how would rear vs front mounted compare in terms of contrast and eof?

I don't consider f:2.8 a slow optic.  My test photos in post 1&2 speak for themselves regarding H-a filtration.  And as noted, I also tested my f:2.8 Newt with the filters mentioned with almost identical results compared to the camera lenses.  

 

As Butterfly explained, long-pass filters are very different; they do not present the same challenges as narrow band-pass H-a filters.  Long-pass filters can prevent the aspect of star bloat that derives from chromatic distortion.  For example, a 640nm long-pass can eliminate most chromatic distortion when used with NV in my 120mm f:5 achromat, because it is blocking most of the shorter, visible wavelengths... which is why they are used as light pollution blocking filters with NV.  


Edited by GeezerGazer, 08 July 2021 - 02:21 PM.

  • longbond likes this

#27 longbond

longbond

    Sputnik

  • *****
  • Posts: 25
  • Joined: 19 May 2018
  • Loc: Chicago

Posted 08 July 2021 - 05:24 PM

I'm pretty sure I get the gist of Joe's question. Joe, correct me if I'm wrong, but you're referring to narrowband H-a filtration. The first photos by Ray (GeezerGazer) are great examples of what can sometimes go wrong with rear-mounting filters.

 

The quick answer is to do front-mounting if:

1) Your filter is big enough for the front of your lens

2) The field of view is less than your filter's band-shift drop-off point

 

My testing shows the following drop-off points:

>>12nm filter is good up to 36° true field of view

>>6nm is good up to 25½° FOV

>>3nm is good up to 18° FOV

 

These amounts can vary by brand and by individual filter, but this is generally what you should expect. If you're okay with some minor drop-off at the edge of field, you can go a bit wider. For example, I can use a 12nm Optolong in front of my 39½° Mod-3/Envis and I barely see the drop-off.

 

About rear-mounting...when it's in-band, it's every bit as good as front-mounting. But therein lies the rub. Rear-mount band shift is stealthy and the math is more complex. Additionally, camera/C-mount lens band-shift performance varies greatly. Unlike with front-mount, you can't directly see rear-mount drop-off. Ray's photo series of the Rosette nicely shows how you can reclaim missing H-alpha by stopping down. Generally, f/2.8 is good enough...most of the time smile.gif



#28 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,720
  • Joined: 14 Jan 2008
  • Loc: Saratoga, CA

Posted 09 July 2021 - 11:19 PM

Joe, I'm not sure if you are now talking about H-a filtration or long-pass filtration.  And, without knowing which type of filter, I can't answer your question.  

I don't consider f:2.8 a slow optic.  My test photos in post 1&2 speak for themselves regarding H-a filtration.  And as noted, I also tested my f:2.8 Newt with the filters mentioned with almost identical results compared to the camera lenses.  

 

As Butterfly explained, long-pass filters are very different; they do not present the same challenges as narrow band-pass H-a filters.  Long-pass filters can prevent the aspect of star bloat that derives from chromatic distortion.  For example, a 640nm long-pass can eliminate most chromatic distortion when used with NV in my 120mm f:5 achromat, because it is blocking most of the shorter, visible wavelengths... which is why they are used as light pollution blocking filters with NV.  

I'm just thinking about Ha filtration. In your posts you show how at f1.4, rear mounted filters lose a lot of contrast compared to front mounted. However, I'm wondering about how if at f2.8 a rear mounted compares with the same telescope at f2.8 but now front mounted. I looked closely at post 1 and 2 and don't see that exact comparison. What I see is in post 1 is 1) f1.4 rear vs front compared 2) f1.4 rear center vs rear edge 3) f2.8 rear 3.5nm vs 7nm and finally f2.8 rear vs f1.4 rear and in post 2: f1.4 vs f2 vs  f2.8 vs f4 all rear. 

 

The only time you compare rear vs front is the very first post but at f1.4. How about the same comparison at f1.8 and/or 2.8 (which I have more lenses at these apertures)?



#29 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,720
  • Joined: 14 Jan 2008
  • Loc: Saratoga, CA

Posted 09 July 2021 - 11:20 PM

I'm pretty sure I get the gist of Joe's question. Joe, correct me if I'm wrong, but you're referring to narrowband H-a filtration. The first photos by Ray (GeezerGazer) are great examples of what can sometimes go wrong with rear-mounting filters.

 

The quick answer is to do front-mounting if:

1) Your filter is big enough for the front of your lens

2) The field of view is less than your filter's band-shift drop-off point

 

My testing shows the following drop-off points:

>>12nm filter is good up to 36° true field of view

>>6nm is good up to 25½° FOV

>>3nm is good up to 18° FOV

 

These amounts can vary by brand and by individual filter, but this is generally what you should expect. If you're okay with some minor drop-off at the edge of field, you can go a bit wider. For example, I can use a 12nm Optolong in front of my 39½° Mod-3/Envis and I barely see the drop-off.

 

About rear-mounting...when it's in-band, it's every bit as good as front-mounting. But therein lies the rub. Rear-mount band shift is stealthy and the math is more complex. Additionally, camera/C-mount lens band-shift performance varies greatly. Unlike with front-mount, you can't directly see rear-mount drop-off. Ray's photo series of the Rosette nicely shows how you can reclaim missing H-alpha by stopping down. Generally, f/2.8 is good enough...most of the time smile.gif

How do I calculate my FOV? I know the ENVIS 1x or PVS-14 is 40 degrees, but thats it. Theres calculators for FOV here: https://astronomy.tools/ but not sure what to input for eyepiece focal length or do I use imaging mode? If so, what do I enter for camera settings?


  • GeezerGazer likes this

#30 longbond

longbond

    Sputnik

  • *****
  • Posts: 25
  • Joined: 19 May 2018
  • Loc: Chicago

Posted 10 July 2021 - 12:52 AM

Joe, calculating FOV depends on how much time you have.

 

Option #1. Quick and dirty is 40° divided by your power. 1x gives you 40°, a 50mm lens gives you 20° and a 75mm gives you 13.3°.

 

Option #2. FOV (in degrees) = Sensor diameter x 57.3 / focal length. For my Mod-3 and Envis, I use 18.24mm for sensor diameter and 26.5mm for focal length. This gives me 39.4° for FOV. It's a good approximation, but not exact.

 

Option #3. The exact FOV requires a little bit of trigonometry. FOV (in degrees) = 2 x arctan (0.5 x sensor diameter / focal length) x 57.3. For my Envis/Mod-3, this gives me 38.0°

 

BTW, 57.3 ≈ 180°/PI. This is the radian to degree conversion factor.

 

Most people use Option #2. It's really simple and for small FOVs, it's nearly perfect. At 20° FOV, the error is only 1%. However, it blows up at large angles. At 40°, the error is 4% and it rapidly gets worse above that. Most AP imagers use Option #2.

 

Trig...you can't live with it and you can't live without it smile.gif. Hope this helps...class dismissed


  • GeezerGazer and joelin like this

#31 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,720
  • Joined: 14 Jan 2008
  • Loc: Saratoga, CA

Posted 10 July 2021 - 01:31 AM

Joe, calculating FOV depends on how much time you have.

 

Option #1. Quick and dirty is 40° divided by your power. 1x gives you 40°, a 50mm lens gives you 20° and a 75mm gives you 13.3°.

 

Option #2. FOV (in degrees) = Sensor diameter x 57.3 / focal length. For my Mod-3 and Envis, I use 18.24mm for sensor diameter and 26.5mm for focal length. This gives me 39.4° for FOV. It's a good approximation, but not exact.

 

Option #3. The exact FOV requires a little bit of trigonometry. FOV (in degrees) = 2 x arctan (0.5 x sensor diameter / focal length) x 57.3. For my Envis/Mod-3, this gives me 38.0°

 

BTW, 57.3 ≈ 180°/PI. This is the radian to degree conversion factor.

 

Most people use Option #2. It's really simple and for small FOVs, it's nearly perfect. At 20° FOV, the error is only 1%. However, it blows up at large angles. At 40°, the error is 4% and it rapidly gets worse above that. Most AP imagers use Option #2.

 

Trig...you can't live with it and you can't live without it smile.gif. Hope this helps...class dismissed

thanks for those numbers

i need to bookmark the sensor diameter, focal length, etc and those formulas 


  • GeezerGazer likes this

#32 GeezerGazer

GeezerGazer

    Apollo

  • *****
  • topic starter
  • Posts: 1,387
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 11 July 2021 - 03:07 AM

Joe, 

My experience with band shift using mostly 12, 7 and 3.5nm filters, suggests that the EoF darkening experienced with the 7nm filter, when used with f:1.4, f:1.8 or f:2 optics is substantially reduced at f:2.8 and at f:4 it disappears completely.  I have taken many photos in prime using the 7nm rear-mounted in f:2.8 optics that allows H-a to remain visible almost to the very EoF.  That is not true for the 3.5 filter.  Eric has described 3nm filters as a "harsh mistress."  

 

Night before last, I was at my green zone observing site and I tested the 3.5nm filter using my f:7  AT/EDL 100mm refractor with TV 55/67 eyepiece in afocal.  My limited experience using afocal suggested that it was more akin to using a front mounted filter than a rear mounted filter in prime, because I usually saw EoF darkening depending on which optics and filter I was using.  Rear mounting the filter in prime spreads the issue of band shift across the FoV, reducing contrast to some degree.  Front mounted filters in prime isolates the EoF darkening.  In prime focus, using a front mounted 3.5nm filter reveals substantial EoF darkening visible with optics from f:1.3 to f:2.8, less at f:4 and negligible at f:5.6.  I chose the 3.5nm filter to test in afocal because the 7nm filter would not show EoF darkening at f:2.8.  

 

Although the photo results show EoF darkening that begins at about 50% of the radius... it doesn't become onerous until it gets closer to the EoF.  It is an even drop-off of the H-a contrast due to band shift... very similar to this 3.5nm filter being front mounted in prime.  

 

These photos were taken at exactly the same settings about a minute apart.  Exposure in NightCap was 1 second, averaged 20 seconds, at ISO 160, taken in manual mode in my iPhone 12 Pro Max.  I made sure they were bright enough so the FoV perimeter can be seen to judge the drop-off point in H-a contrast.  

 

IMG_2824.jpeg

 

IMG_2826.jpeg

 

IMG_2825.jpeg

 

Like many other NV issues, the severity of this problem shows in photos much more prominently than when used visually.  The night before I took these photos, I used the 3.5nm in afocal with the same scope and eyepiece and was not really bothered by the degree of band shift that was present.  But if you look carefully at these photos, you can see that the outer 50% of the FoV is effected by band shift. 

 

 

 


  • longbond and AllStarez like this

#33 GeezerGazer

GeezerGazer

    Apollo

  • *****
  • topic starter
  • Posts: 1,387
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 13 July 2021 - 12:17 PM

Joe, 

My experience with band shift using mostly 12, 7 and 3.5nm filters, suggests that the EoF darkening experienced with the 7nm filter, when used with f:1.4, f:1.8 or f:2 optics is substantially reduced at f:2.8 and at f:4 it disappears completely.  I have taken many photos in prime using the 7nm rear-mounted in f:2.8 optics that allows H-a to remain visible almost to the very EoF.  That is not true for the 3.5 filter.  Eric has described 3nm filters as a "harsh mistress."  

 

Night before last, I was at my green zone observing site and I tested the 3.5nm filter using my f:7  AT/EDL 100mm refractor with TV 55/67 eyepiece in afocal.  My limited experience using afocal suggested that it was more akin to using a front mounted filter than a rear mounted filter in prime, because I usually saw EoF darkening depending on which optics and filter I was using.  Rear mounting the filter in prime spreads the issue of band shift across the FoV, reducing contrast to some degree.  Front mounted filters in prime isolates the EoF darkening.  In prime focus, using a front  mounted 3.5nm filter reveals substantial EoF darkening visible with optics from f:1.3 to f:2.8, less at f:4 and negligible at f:5.6.  I chose the 3.5nm filter to test in afocal because the 7nm filter would not show EoF darkening at f:2.8.  

 

Although the photo results show EoF darkening that begins at about 50% of the radius... it doesn't become onerous until it gets closer to the EoF.  It is an even drop-off of the H-a contrast due to band shift... very similar to this 3.5nm filter being front mounted in prime.  

 

These photos were taken at exactly the same settings about a minute apart.  Exposure in NightCap was 1 second, averaged 20 seconds, at ISO 160, taken in manual mode in my iPhone 12 Pro Max.  I made sure they were bright enough so the FoV perimeter can be seen to judge the drop-off point in H-a contrast.  

 

attachicon.gifIMG_2824.jpeg

 

attachicon.gifIMG_2826.jpeg

 

attachicon.gifIMG_2825.jpeg

 

Like many other NV issues, the severity of this problem shows in photos much more prominently than when used visually.  The night before I took these photos, I used the 3.5nm in afocal with the same scope and eyepiece and was not really bothered by the degree of band shift that was present.  But if you look carefully at these photos, you can see that the outer 50% of the FoV is effected by band shift. 

Joe, 

My apologies, the sentence above that I've highlighted in red is incorrect.  Changing the aperture when using a front mounted filter has no effect on the amount of band shift that is visible in the image; it only darkens the overall brightness of the image.  I was thinking about results with rear-mounted filters which is how I normally mount them.  


Edited by GeezerGazer, 13 July 2021 - 06:50 PM.



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





Also tagged with one or more of these keywords: Filters, NV



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics