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

Extreme Focal Ratios and Narrowband Filter

  • Please log in to reply
39 replies to this topic

#1 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 03:45 AM

I just got an f/1.4 Sigma 105mm lens.

 

I'm exploring the possibility of using a narrowband Ha filter with it.

 

What is the lastest on using narrowband filters with extreme focal ratios like this?

 

Anybody make a filter that won't shift the bandpass at f/1.4?

 

Thanks,

 

Jerry



#2 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7484
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 23 August 2019 - 04:53 AM

When placed between the lens and the sensor all narrow-band, thin-film filters will have some shift in the bandpass. So, most (all?) that are designed for fast f-ratios just offset the bandpass so that when they are used the expected shift will place the bandpass back into the needed range. I think some manufacturers also widen the bandpass so that even if there is a shift there will still be transmission at the desired wavelengths.

 

I'm not sure whether it is even possible to make a really narrow bandpass filter using any technique that doesn't suffer from this issue as most of these filters rely on interference effects that depend upon a given path length through a thin film (and that path length changes based upon the angle of incidence of the incoming light).

 

I suspect that the only reason why these filters are more expensive than standard narrow-band filters is that they are made in smaller quantities because of lower market demand.


Edited by james7ca, 23 August 2019 - 04:56 AM.


#3 cnoct

cnoct

    Viking 1

  • *****
  • Posts: 752
  • Joined: 02 Sep 2010
  • Loc: Hawai'i

Posted 23 August 2019 - 05:24 AM

The Astronomik 12nm works exceptionally well with fast optics.

Tried the Baader f/2 Highspeed to compare, didn't even come close to the Astronomik 12nm. My Astrodon 5nm worked better than the Baadder.

For reference, this is the configuration I most use it in...

 

gallery_139776_8407_429959.jpg

 

300mm f/1.4
 



#4 lucam

lucam

    Viking 1

  • -----
  • Posts: 614
  • Joined: 09 Oct 2017
  • Loc: Upstate NY, USA

Posted 23 August 2019 - 05:49 AM

I have used Astrodon 3nm and 5nm filters with f1.4 to f2.8 lenses (sometimes stopped down with step-down rings) and have always obtained very usable data with CMOS cameras (ASI183 and ASI1600). It is difficult to know exactly how much signal is lost off-axis but there were no gradients or weird artifacts difficult to correct. Things may be different with a full-frame sensor, which certainly pushes things a lot farther off axis than the small sensors I have used.



#5 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 05:51 AM

When placed between the lens and the sensor all narrow-band, thin-film filters will have some shift in the bandpass. So, most (all?) that are designed for fast f-ratios just offset the bandpass so that when they are used the expected shift will place the bandpass back into the needed range. I think some manufacturers also widen the bandpass so that even if there is a shift there will still be transmission at the desired wavelengths.

 

I'm not sure whether it is even possible to make a really narrow bandpass filter using any technique that doesn't suffer from this issue as most of these filters rely on interference effects that depend upon a given path length through a thin film (and that path length changes based upon the angle of incidence of the incoming light).

 

I suspect that the only reason why these filters are more expensive than standard narrow-band filters is that they are made in smaller quantities because of lower market demand.

Is there any way to calculate the shift based on the f/number?

 

I have used inteference filters between the lens and the camera for a 16mm f/2, and it seemed to work ok, but it wasn't a narrrowband filter.  I do know there there would be no way to use an inteference filter in front of that wide of a lens.

 

So I was hoping the light cone would be more parallel in the 105mm.  

 

I certainly can't afford a narrowband filter that would be needed for the 105mm front-filter diameter of the lens I'm using.

 

Jerry



#6 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 05:52 AM

The Astronomik 12nm works exceptionally well with fast optics.

Tried the Baader f/2 Highspeed to compare, didn't even come close to the Astronomik 12nm. My Astrodon 5nm worked better than the Baadder.

For reference, this is the configuration I most use it in...

 

gallery_139776_8407_429959.jpg

 

300mm f/1.4
 

That is awesome!  Where did you get that lens?

 

So your experience is that the AstroDon 5nm Ha filter worked the best at f/1.4?

 

Thanks,

 

Jerry



#7 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 05:54 AM

I have used Astrodon 3nm and 5nm filters with f1.4 to f2.8 lenses (sometimes stopped down with step-down rings) and have always obtained very usable data with CMOS cameras (ASI183 and ASI1600). It is difficult to know exactly how much signal is lost off-axis but there were no gradients or weird artifacts difficult to correct. Things may be different with a full-frame sensor, which certainly pushes things a lot farther off axis than the small sensors I have used.

Hi Luca,

 

Thanks, I PM'd you about this too.  

 

Jerry



#8 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7484
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 23 August 2019 - 06:03 AM

Is there any way to calculate the shift based on the f/number?

...

Jerry

The link I gave earlier has a formula to calculate the shift, but you'd have to know the index of refraction for each layer in the filter and only the manufacturer has that kind of information. In any case, the wider the original bandpass the less likely you will have problems.



#9 cnoct

cnoct

    Viking 1

  • *****
  • Posts: 752
  • Joined: 02 Sep 2010
  • Loc: Hawai'i

Posted 23 August 2019 - 06:07 AM

That is awesome!  Where did you get that lens?

 

So your experience is that the AstroDon 5nm Ha filter worked the best at f/1.4?

 

Thanks,

 

Jerry

 

The Astronomik 12nm is better, less band shift than the AstroDon 5nm.  The Astronomik 12nm handily outperforms the AstroDon 5nm @ f/1.4

 

Was using the AstroDon 5nm results to convey how poorly the specialized Baader "highspeed" performed. 

 

The lens came from a DLA auction, the FBI bought a mess of them over the years.  



#10 AtmosFearIC

AtmosFearIC

    Apollo

  • -----
  • Posts: 1020
  • Joined: 10 Dec 2015
  • Loc: Melbourne

Posted 23 August 2019 - 06:43 AM

The Astronomik 12nm is better, less band shift than the AstroDon 5nm. The Astronomik 12nm handily outperforms the AstroDon 5nm @ f/1.4

Was using the AstroDon 5nm results to convey how poorly the specialized Baader "highspeed" performed.

The lens came from a DLA auction, the FBI bought a mess of them over the years.


What’s the correction like with such a beast?

#11 Churmey

Churmey

    Vostok 1

  • *****
  • Posts: 129
  • Joined: 05 Feb 2017

Posted 23 August 2019 - 07:06 AM

Use step down rings from the front to reduce your incoming aperture to accommodate 2" filters.  This should eliminate the shift issue but it would decrease your aperture to around F2.2 when using your 105.


Edited by Churmey, 23 August 2019 - 07:09 AM.

  • kingjamez likes this

#12 Zebenelgenubi

Zebenelgenubi

    Ranger 4

  • *****
  • Posts: 393
  • Joined: 06 Apr 2012
  • Loc: Texas, USA

Posted 23 August 2019 - 08:01 AM

I may be oversimplifying things but the designed passband frequency should be proportional to the layer thickness "d".  Parts of the light cone will enter the filter at and angle "theta" will have a longer path length through the filter of "d x cos(theta)".  That seems to indicate that the passband for those light rays should shift by ( 1-cos(theta) ) relative to the designed passband.  The shift would seem to be significant for very short focal ratios which have a wide light cone.  The light arriving nearly normal to the filter is pretty much unaffected but the passband will be shifted for the extremes of the light come at small f-numbers.  You could probably measure the shift experimentally with a white light source, narrow slit and diffraction grating as you rotate the filter relative to the normal of the light path.  I can see the color shift when I hold my narrowband filter in my hand and turn the filter relative to angle of the incident light.



#13 fetoma

fetoma

    Gemini

  • *****
  • Posts: 3002
  • Joined: 26 Sep 2006
  • Loc: Southern NJ

Posted 23 August 2019 - 09:24 AM

The Astronomik 12nm is better, less band shift than the AstroDon 5nm.  The Astronomik 12nm handily outperforms the AstroDon 5nm @ f/1.4

 

Was using the AstroDon 5nm results to convey how poorly the specialized Baader "highspeed" performed. 

 

The lens came from a DLA auction, the FBI bought a mess of them over the years.  

I don't think Jerry is wanting that large of a bandpass. Just a hunch though.



#14 kathyastro

kathyastro

    Viking 1

  • *****
  • Posts: 883
  • Joined: 23 Dec 2016
  • Loc: Nova Scotia

Posted 23 August 2019 - 09:38 AM

The wavelength shift varies over the field of view.  Light coming from the centre of the scope's light cone hits the filter at 90 degrees,and passes the designed wavelength.  Light coming from the edges of the scope's light cone hits the filter at an oblique angle, and gets its wavelength shifted.  That is why you need a wider bandpass, not just a different wavelength, for fast scopes/lenses.


  • Zebenelgenubi likes this

#15 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 03:15 PM

Use step down rings from the front to reduce your incoming aperture to accommodate 2" filters.  This should eliminate the shift issue but it would decrease your aperture to around F2.2 when using your 105.

I'm certain that would vignette, and the vignetting is bad enough already.

 

That was my experience using step down rings on my 180mm f/2.8. 

 

And I didn't spend $1700 to not use $1,000 worth of the glass. :-)

 

Jerry



#16 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 03:17 PM

I don't think Jerry is wanting that large of a bandpass. Just a hunch though.

If the bandpass is going to shift a lot because of the focal ratio, I would need a wider bandpass.

 

Ideally if the bandpass shift wasn't a problem, you would be right, I would want to go as narrow as possible.

 

Jerry


  • kingjamez likes this

#17 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2160
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 23 August 2019 - 03:35 PM

If the bandpass is going to shift a lot because of the focal ratio, I would need a wider bandpass.

 

Ideally if the bandpass shift wasn't a problem, you would be right, I would want to go as narrow as possible.

 

Jerry

Exactly, since the bandpass shifts progressively as the light cone steepens, the bandpass must widen and not simply shift because you don't want the on axis light to be blocked. The faster the focal ratio the wider the bandpass.

I'm certain that would vignette, and the vignetting is bad enough already.

 

That was my experience using step down rings on my 180mm f/2.8. 

 

And I didn't spend $1700 to not use $1,000 worth of the glass. :-)

 

Jerry

There is a careful balance to be played here. You can put an extremely narrow pass filter on the front and loose the outer portion of the lens (effectively just increasing the focal ratio) or you can have a much wider bandpass filter at the other end.

 

Which is better: F/1.4 with a ~20nm wide filter or F3.0 with a 3nm wide filter? I'd wager the later will produce a better result in light polluted skies.

 

-Jim


  • Churmey likes this

#18 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 05:04 PM

Exactly, since the bandpass shifts progressively as the light cone steepens, the bandpass must widen and not simply shift because you don't want the on axis light to be blocked. The faster the focal ratio the wider the bandpass.

There is a careful balance to be played here. You can put an extremely narrow pass filter on the front and loose the outer portion of the lens (effectively just increasing the focal ratio) or you can have a much wider bandpass filter at the other end.

 

Which is better: F/1.4 with a ~20nm wide filter or F3.0 with a 3nm wide filter? I'd wager the later will produce a better result in light polluted skies.

Hi Jim,

 

It's probably a moot point because of the vignetting.

 

You realize that the entrance pupil that defines the aperture is inside the lens, right?

 

Jerry



#19 AtmosFearIC

AtmosFearIC

    Apollo

  • -----
  • Posts: 1020
  • Joined: 10 Dec 2015
  • Loc: Melbourne

Posted 23 August 2019 - 05:05 PM

Exactly, since the bandpass shifts progressively as the light cone steepens, the bandpass must widen and not simply shift because you don't want the on axis light to be blocked. The faster the focal ratio the wider the bandpass.
There is a careful balance to be played here. You can put an extremely narrow pass filter on the front and loose the outer portion of the lens (effectively just increasing the focal ratio) or you can have a much wider bandpass filter at the other end.

Which is better: F/1.4 with a ~20nm wide filter or F3.0 with a 3nm wide filter? I'd wager the later will produce a better result in light polluted skies.

-Jim


If you’re shooting at F/3 you’d probably want to go to 5nm pass band.
I’ve got some 3nm Astrodons and I have for a long time been contemplating moving to 5nm on a chip larger than an ASI1600. Plus at 3nm you get a 14% intensity drop at F/3 as opposed to F/3.5.

#20 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8893
  • Joined: 12 Oct 2007

Posted 23 August 2019 - 05:47 PM

Hi Jim,

 

It's probably a moot point because of the vignetting.

 

You realize that the entrance pupil that defines the aperture is inside the lens, right?

 

Jerry

Hi Jerry-

 

People here are talking about vignetting - but there is likely very little vignetting caused by increasing field angle because it's the exit pupil - and its location - that matter.  In dslr lenses the exit pupil is pushed far away from the sensor optically - so there is much less tilt in field angle than people expect.

 

Instead of vignetting there will mainly just be a loss of aperture across the field - and that won't show in the image at all if you expose long enough.  That's why so many people try fast lenses with narrow filters and say they 'work.'  Well - they certainly should work, but at reduced aperture.

 

For some dslr lenses you can find out where the exit pupil is located and use that to estimate how much actual vignetting there would be.  But even in the center of the field, where no vignetting is possible, there will be reduced aperture and throughput at f/1.4.  And to make things worse - that reduced aperture only affects the nebulosity signal you are after, while the sky background still comes through at full power.

 

I created a tool to show how vignetting behaves with different exit pupil distances.  Note that the size of the exit pupil could be anything - including infinity - even with a very slow lens.  That's because its size and location are determined by the optical design - and there are benefits to pushing the exit pupil far from the sensor - particularly for wide field lenses.

 

Here is the thread and tool that somewhat explains this - and you may be able to find the parameters for your lens:

 

https://www.cloudyni...tting-with-app/

 

Note that hyperstar is an example that appears to have the exit pupil at infinity so there is no increased tilt as you move away from the center - even though it is f/2.

 

Frank


  • kingjamez likes this

#21 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2160
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 23 August 2019 - 06:27 PM

If you’re shooting at F/3 you’d probably want to go to 5nm pass band.
I’ve got some 3nm Astrodons and I have for a long time been contemplating moving to 5nm on a chip larger than an ASI1600. Plus at 3nm you get a 14% intensity drop at F/3 as opposed to F/3.5.

If you put the filter in front of the objective (since it's presumably smaller, this would then define the objective size) then the incoming light would be parallel and a very narrow filter could be used. The Baader UFC has adapters specifically designed to allow for putting the filter in front of the lens objective.

-Jim


Edited by kingjamez, 23 August 2019 - 07:34 PM.


#22 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 23 August 2019 - 08:31 PM

If you put the filter in front of the objective (since it's presumably smaller, this would then define the objective size) then the incoming light would be parallel and a very narrow filter could be used. The Baader UFC has adapters specifically designed to allow for putting the filter in front of the lens objective.

Hi Jim,

 

Actually, it depends on the lens.

 

Try stopping down a wide-angle lens like that and see what happens.

 

Jerry



#23 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8893
  • Joined: 12 Oct 2007

Posted 23 August 2019 - 08:46 PM

Hi Jim,

 

Actually, it depends on the lens.

 

Try stopping down a wide-angle lens like that and see what happens.

 

Jerry

For your 105mm f/1.4 lens I can't tell where the exit pupil is but it may be closer than 105mm or farther.  Here is a table with some examples:

 

https://www.lensrent...comment-page-1/

 

For Sigma the 50mm has a pupil at 97.75mm distance, while the 85mm is at 58mm.

 

But your 105mm has an entrance pupil diameter of 75mm and is long enough focal length that it could probably be stopped down with an aperture in front and not get too much aberration.  I stop down my Canon 135mm f/2 that way to f/4 to avoid diffraction spikes.

 

For your lens I think the best results would come with a 50mm 3nm filter in front - if you can afford it.  That would reduce the aperture but with a well defined and narrow emission passband - with a fully utilized f/2.1 and no detriment from the combination of speed and narrowband since the filter is in front - as kingjamez suggests.

 

Frank


  • kingjamez likes this

#24 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 24 August 2019 - 04:53 AM

If you’re shooting at F/3 you’d probably want to go to 5nm pass band.
I’ve got some 3nm Astrodons and I have for a long time been contemplating moving to 5nm on a chip larger than an ASI1600. Plus at 3nm you get a 14% intensity drop at F/3 as opposed to F/3.5.

Hi AtmosFearlC,

 

Where are you finding the data on the amount of intensity drop at different focal ratios?

 

Thanks,

 

Jerry



#25 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • topic starter
  • Posts: 7262
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 24 August 2019 - 05:50 PM

For your 105mm f/1.4 lens I can't tell where the exit pupil is but it may be closer than 105mm or farther.  Here is a table with some examples:

 

https://www.lensrent...comment-page-1/

 

For Sigma the 50mm has a pupil at 97.75mm distance, while the 85mm is at 58mm.

 

But your 105mm has an entrance pupil diameter of 75mm and is long enough focal length that it could probably be stopped down with an aperture in front and not get too much aberration.  I stop down my Canon 135mm f/2 that way to f/4 to avoid diffraction spikes.

 

For your lens I think the best results would come with a 50mm 3nm filter in front - if you can afford it.  That would reduce the aperture but with a well defined and narrow emission passband - with a fully utilized f/2.1 and no detriment from the combination of speed and narrowband since the filter is in front - as kingjamez suggests.

Hi Frank,

 

Well, I don't know where the exit pupil is for the 105mm, but I did some tests and my suspicions were confirmed.

 

A front filter aperture stop increases the vignetting. See the animated GIF below.

 

This was with a 52mm stop that made the scope f/2 roughly.

 

Jerry

 

ApertureStopVignetting2.gif


Edited by Jerry Lodriguss, 24 August 2019 - 05:53 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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics