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

Estimating your fully illuminated field size

EAA NV
  • Please log in to reply
10 replies to this topic

#1 Eddgie

Eddgie

    ISS

  • *****
  • topic starter
  • Posts: 23741
  • Joined: 01 Feb 2006

Posted 26 May 2019 - 08:49 AM

When used visually, field illumination falloff or vignetting of fairly strong proportion is often easily accepted and sometimes totally escapes visual notice.  When imaging or when using image intensifiers though, field illumination falloff or vignetting is far more of a concern, and when heavy focal reduction is used, aperture loss becomes a very real concern.

 

For aperture loss, the laser projection method is probably the most effective method to use, but for a lot of people getting a calculation of vignetting or illumination falloff can be a challenge.

 

The good news is that there is a very easy way to estimate the amount of intensity falloff of virtually any kind of telescope configuration using nothing more than a bright star an the knowledge of the size of the field (sensor size, photocatode, or eyepiece field stop).   

 

An obstruction is useful for this, so even if the scope does not have one, adding one made from a cut out piece of cardboard and held over the aperture with a couple of strips of tape will be fine.  This is just useful for getting a better estimate of the amount of illumination falloff that is occurring.

 

The test is easy.   Just point the telescope at a bright star and center it in the field of view.   Now, defocus the star so that three or four rings are showing (if using an obstruction, maybe six if not, but these numbers are not at all critical).

 

Now, move the scope so that the unfocused pattern drifts toward the edge of the field and watch the outside edge of the pattern. As you move the scope, you may see that there is a smooth, ark shaped intrusion into the edge of the pattern.  When you see this, you are seeing the point where either the off axis rays from the primary are starting to fail to intercept the edges of the secondary (Newtonians) or where a baffle or focuser tube in extending into the edges of the light cone (SCTs are the worse for this, with many only starting with a 6mm to 10mm fully illuminated true field).d

 

Knowing your sensor size, you should now be able to estimate the size of your fully illuminated field.  For example, if using an APS-C size sensor or an image intensifier, if the intrusion occurs when the pattern is half way to from the center tot the edge of the field, this would indicate that the fully illuminated field is only about half the diameter of the chip or photocatode, so the fully illuminated field size would only be about 9mm.   If it happened 1/3rd of the way from the center tot he edge, the fully illuminated field would be only 6mm, and so on.

 

As you continue the drift to the edge of the field, notice that the intrusion now starts to consume more and more of the pattern, and by estimating the amount of the pattern being blanked out, one can estimate the amount of illumination falloff.  For example, if by the edge of the field, only half of the area of the pattern remains, then this indicates that the illumination has fallen by 50%.

 

As long as a subject fits entirely inside the fully illuminated area, then 40% or 50% falloff will not damage the image other than darkening the outside of the field (flats) but for very large objects that extend over most or all of the field, a considerable amount of image brightness is being lost.

 

Illumination falloff.png

 

Here is an example I would give you (and next time I have a clear night, I hope to take pictures, but I have been plagued by clouds for a month now). Sometimes I use my 6" f/2.8 in a configuration that has a 1.25" eyepiece holder behind the filter wheel.   I do this simply because it makes it easy to swap the device with a 1.25" nose to other telescopes.   If I do this though, it places the 1.25" filter 45mm in front of the photocatode.   In this configuration, the test I just describes shows that the Fresnel pattern starts to loose illumination at only about 3mm from the center of the field, indicating that my fully illuminated field is only about 6mm.  Because the vignetting source (the filter) is far ahead of the focal plane, the illumination falloff is not aggressive, but it is constant and by the time the pattern reaches the edge of the field, illumination has fallen to 40%.   Now for a galaxy at the center of the field, this is meaningless, but for very faint nebula that stretches outside of the field, this is all to often the difference between seeing it and not seeing it.

 

Now since my reducer/corrector requires 65mm of spacing, if I want to more fully illuminate the field, I can remove the 1.25" eyepiece holder and attach the NV device directly to the  rear of the filter wheel using a T2 adapter.  This puts the filter only about 20mm in front of the focal plane.  I now have to put a spacer between the reducer and the front of the filter wheel, so this setup does not allow for easy movement of the devices between scopes, but now, I an move the defocused pattern to the edge to the edge of what I estimate to be a 15mm image circle before the intrusion starts, and while the falloff is more aggressive (the amount of pattern lost for each millimeter of movement) by the edge of the circle, I estimate that I still have 80% illumination.  This one change produces an extremely large increase in the size of the fully illuminated field, and the amount of vignetting at the edge remains very small.   By going to 2" filter I could fully illuminate the entire 18mm circle, but the cost vs the tiny drop in illumination at the edge of the field is substantial, so I am happy to accept this tiny amount of illumination loss as a good compromise in terms of cost for performance.   The point here is that I know the behavior of my system so I can make this decision knowing exactly what my money will or won't get me. 

 

The reason I posted this is because I see a great number of cases where people are doing extreme amounts of compression, and I often wonder if they have ever devoted any effort to actually measuring the loss of focal speed over the entire field.  Sometimes, it can be quite extreme, making a less aggressive reduction producing almost the same brightness over the average of the field size.  For example, a Celestron C6 only starts with a fully illuminated field that is about 5mm in diameter.  Compressing this to .33 percent of its original size means that maybe only 1.5mm of the image is fully illuminated, and if the configuration induces aperture loss (easy to do in an SCT without attention to the configuration) the center of the field could be working at well under the "nominal" f/3.3.    Of course this number is already false because with transmission loss and secondary shading, the nominal .33 reduction is really only yielding maybe f/3.8 brightness, but that is a different story for a different time.   

 

I have used this method to estimate field illumination for over a decade now.  Most calculators do not allow for the modifications induced by focal reduction, so for this case, the method I provided allows for an easy way to make a reasonable estimate of how big the fully illuminated image circle is, and how much illumination loss occurs at the edge of the field.

 

I hope someone found this post to be useful. 


Edited by Eddgie, 26 May 2019 - 09:23 AM.

  • ccs_hello, GlennLeDrew and Earthbound1 like this

#2 Rickster

Rickster

    Viking 1

  • *****
  • Posts: 706
  • Joined: 09 Jun 2008
  • Loc: NC Kansas Bortle 3 SQM 21.8+

Posted 29 May 2019 - 02:27 AM

More along this line.

 

https://www.peterson.../vignetting.htm



#3 Eddgie

Eddgie

    ISS

  • *****
  • topic starter
  • Posts: 23741
  • Joined: 01 Feb 2006

Posted 29 May 2019 - 01:58 PM

More along this line.

 

https://www.peterson.../vignetting.htm

Well, the above link is good in general concept and points out the effect of light path restrictions, but the the author of the above page is quite a bit off on his fully illumined field figures. 

 

Here are the figures. 

 

 

The C8 only has a 100% illuminated circle about 8mm in diameter.  With the f/6.3 focal reducer, only the very center of the field is fully illuminated (with 100mm of back focus)

 

The C 9.25 likewise only has a fully illuminated circle of about 8mm and with the f/6.3 reducer It is a bit smaller than this. The C925 achieves this by using a considerably larger secondary mirror and baffle.

 

The C11 has a 30mm fully illuminated image circle at 100mm of back focus, and this drops to 26mm with the reducer.  This is the biggest fully illuminated field of the 100mm back focus designs and well short of the 50mm that the author above suggests.  Still, this is the standard SCT that would be best suited to aggressive reduction. 

 

The C14 has about a 16mm fully illuminated image circle at 150mm of back focus (this scope was designed for 2" eyepiece use). With th f/6.3 the redcuer at 100mm of back focus, the fully illuminated field of the C14 is only a couple of millimeters.

 

This is based on the recommended spacing for the Celestron f/6.3, which is (as I recall) 96mm or something.    Of course if less back space were used behind the reducer, the reduction would be less.

 

Withe the spacing required for these multiple reducers to give their stated reduction power, it is likely that many SCTs are in a reduced aperture state. 


Edited by Eddgie, 29 May 2019 - 02:47 PM.


#4 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 538
  • Joined: 07 Mar 2014

Posted 29 May 2019 - 02:12 PM

Well, the above link is good in general concept and points out the effect of light path restrictions, but the the author of the above page is quite a bit off on his fully illumined field figures. 

Here are the figures. 

 

 

The C8 only has a 100% illuminated circle about 8mm in diameter.  With the f/6.3 focal reducer, only the very center of the field is fully illuminated (with 100mm of back focus)

 

The C 9.25 likewise only has a fully illuminated circle of about 8mm and with the f/6.3 reducer It is a bit smaller than this. The C925 achieves this by using a considerably larger secondary mirror and baffle.

 

The C11 has a 30mm fully illuminated image circle at 100mm of back focus, and this drops to 26mm with the reducer.  This is the biggest fully illuminated field of the 100mm back focus designs and well short of the 50mm that the author above suggests.  Still, this is the standard SCT that would be best suited to aggressive reduction. 

 

The C14 has about a 16mm fully illuminated image circle at 150mm of back focus (this scope was designed for 2" eyepiece use). With th f/6.3 the redcuer at 100mm of back focus, the fully illuminated field of the C14 is only a couple of millimeters.

 

This is based on the recommended spacing for the Celestron f/6.3, which is (as I recall) 96mm or something.    Of course if less back space were used behind the reducer, the reduction would be less.

 

Withe the spacing required for these multiple reducers to give their stated reduction power, it is likely that many SCTs are in a reduced aperture state. 

Do you know what the equivalent figures are for the edge scts with the specialist 0.7 edge reducer?



#5 Eddgie

Eddgie

    ISS

  • *****
  • topic starter
  • Posts: 23741
  • Joined: 01 Feb 2006

Posted 29 May 2019 - 08:28 PM

Do you know what the equivalent figures are for the edge scts with the specialist 0.7 edge reducer?

I am sorry but these figures I have are from the SCT Vignetting Analysis that was done by Tom Richardson (?).  It does not cover the EdgeHD.

 

Now Celestron says that the EdgeHD provides a 42mm field, but they do not say that this field is fully illuminated and the drawing provided in the white paper suggests that fully illuminated circle is about 30mm (Figure 13 on page 12 of the white paper).  According to the same document, the C11 and C14 miantain the 42mm image circle when the reducer is used, but they do not say what happens to the fully illuminated circle.   My guess (just a guess based on the math) is that the EdgeHD 11 and 14, when used with the reducer, will fully illuminate an APS-C size circle.  Now again, Celestron gives no figures, but the drawing I referenced above clearly suggests that the full 42mm image circle is not fully illuminated, and the 30mm figure I use was simply derived by estimation from the image, and by applying the ray trace for the standard C11.

 

Again, this was why I posted this.  Often there is not really any data about field illumination, but using the method I provided, one can make a reasonable estimate of the actual performance of the system.  While it might not be super accurate, it is at least a fair approximation.

 

For people with software, I think their programs can estimate it, but for NV people this method might be very useful.


  • Gavster likes this

#6 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 538
  • Joined: 07 Mar 2014

Posted 30 May 2019 - 01:33 AM

I am sorry but these figures I have are from the SCT Vignetting Analysis that was done by Tom Richardson (?).  It does not cover the EdgeHD.

 

Now Celestron says that the EdgeHD provides a 42mm field, but they do not say that this field is fully illuminated and the drawing provided in the white paper suggests that fully illuminated circle is about 30mm (Figure 13 on page 12 of the white paper).  According to the same document, the C11 and C14 miantain the 42mm image circle when the reducer is used, but they do not say what happens to the fully illuminated circle.   My guess (just a guess based on the math) is that the EdgeHD 11 and 14, when used with the reducer, will fully illuminate an APS-C size circle.  Now again, Celestron gives no figures, but the drawing I referenced above clearly suggests that the full 42mm image circle is not fully illuminated, and the 30mm figure I use was simply derived by estimation from the image, and by applying the ray trace for the standard C11.

 

Again, this was why I posted this.  Often there is not really any data about field illumination, but using the method I provided, one can make a reasonable estimate of the actual performance of the system.  While it might not be super accurate, it is at least a fair approximation.

 

For people with software, I think their programs can estimate it, but for NV people this method might be very useful.

Thanks Eddgie. It seems I accidentally lucked out by getting a c11 since it has the largest fully illuminated image circle of any sct.

I have been using my standard c11 with AP 0.75x reducer and 55mm plossl afocally with my nv monoculars for about a year. It has surprised me how well it works with good performance up to the edge with only minimal vignetting. I use a 2 inch diagonal so I’m guessing I go beyond the 100mm back focus you quote which may have an impact but the results look good to me.

I’ve recently purchased a c11 edge with the specialist 0.7x edge reducer with the aim of getting tighter stars at the edge of the fov. I’ve had good success here with my limited testing but am not sure whether I’m getting more vignetting etc. I will continue to compare with my standard c11.

 

Please could you also expand on your comments regarding transmission loss and secondary shading having an impact on the f ratio of the system?


Edited by Gavster, 30 May 2019 - 01:34 AM.


#7 Eddgie

Eddgie

    ISS

  • *****
  • topic starter
  • Posts: 23741
  • Joined: 01 Feb 2006

Posted 30 May 2019 - 11:56 AM

It does not necessarily change the "Focal ratio" (thought if the system is loosing aperture as is most likely when using extreme focal reduction) but it does alter the brightness so that it is not really as bright as the focal ratio number would suggest.

 

Let's take the case of the C9.25 as an example but let's start with something that all modern Celestron scopes have in common, and that is system transmission.  The data I am going to use here can be found on the web and I will link it but here are the important points.   First, while modern night vision gear does provide some gain at the normal visual frequencies, the real gain for Gen III is found in the near IR range, and specifically from about 600nm to 900nm.    Also, many NV users that observe under light polluted skies use IR long pass filters, and this will of course reject most of the light that is above the cutoff, so let's say 650nm to about 850nm, and past this, the GaAs photocatode starts to loose sensitivity rapidly.   Ok, so a lot of NV is done over the 650nm  to 875nm.

 

Now, if you look at the Celestron transmission charts for the XLT coatings (and this is as good as it gets) the transmission at 650nm is only 85% for the system, but that is only at 650nm.   By the time you get to 750nm, the system transmission has fallen to 75%.  While the Celestron chart stops at 750nm, assuming that the drop was constant, by the time one reached 875nm, the transmission would be probably in the 55% to 65% range.   This matters when doing observing with IR Pass filters, but for H-a, lets say that the system transmission (as indicated by Celestron chart) is about 84% (and this of course does not include the diagonal).   For non H-a then, the transmission is quite a bit lower than it would be at visual frequencies, and this means the image is dimmer than it would otherwise be.   Again, the focal ratio is has not changed but the brightness is not what you would get from a "perfect" f/10 system.

 

https://www.celestro...ptical-coatings

 

Next is secondary shading.   While it is a small factor, it is a factor, and remember that the effect of this is not included in the above transmission chart.  The secondary of the C9.25 is about 3.38" and that equates to an area of about 5.3".  Since the area of the primary is 68.7 inches, this means that only about 61.9 square inches of area is unshaded, and again, this is what the transmission acts con, so 83% (best case) of your 61.9" of collection means that your brightness would only be that of a perfect (no obstruction, perfect transmission) 8" aperture.  Now that is not bad, but in terms of focal ratio, it would mean that you are really working at a focal ratio brightness equivalent  of 2800 + 203 =) f/13 (for H-alpha, but when running a 650nm long pass filter, it would actually be less than this due to the dropping transmission in near IR).

 

And then there is the size of the fully illuminated circle. The C9.25 only has a fully illuminated image circle of about 8mm.  This means that illumiantion falloff stars just outside of this circle, and by the edge of an 18mm circle, the illumination has fallen to about 90%. If you reduce the C9.25 using the standard f/6.3 reducer, the fully illuminated circle is only about 2mm in diameter, and at the edge of the 18mm field, the illumination has fallen to about 80%

 

Now I used the C9.25 as a worse case, but the C8 and C14 behave somewhat similarly (espeially with the focal reducer).  If you were to average the brightness of the field over the area of the field, it would be maybe 95% without the reducer(again, of the available light, which has already been reduced by transmission losses and secondary shading) and maybe 90% with the reducer (this is optimistic but most people would only care about the center of the field. though I would say that using a system with a fully illuminated field produces a far better view for nebula that extent outside of the field of view).

 

Field illumination.png

 

So, what this means is that while the focal ratio of the scope is f/10, the actual brightness that is delivered is more on par with a system that is f/13 (no compression) and probably f/7.5 to f/8 when used with a reducer.

 

My 6" f/2.8 has a wide band coating but probably does not extend down to 900nm.   It has a 40% obstruction.  I calculate that while it is nominally f/2.8, the effective brightness is probably more like f/3.7. 

 

In direct comparisons of my Comet Catcher (which is f/3.6, but has old relatively low transmission optics) to the 6" f/2.8, it is easy to see that the 6" f/2.8 (with an effective brightness of perhps f/3.7) is much brighter than the Comet Catcher was (I sold the Comet Catcher) and what really sealed the deal for me was my experience using a high quality 80mm Apo with broad band coatings.  Even at f/6, the view in the 80mm was not nearly as different as the disparity in "published" focal ratio would suggest.   The Comet Catcher has a large obstruction, and old coating technology, so probably 65% to 70% total system transmission, and when you do the math, the effective brigtness was only a little better than the f/6 Apo.

 

The 6" f/2.8 has a fully illuminated field that is 16mm in diameter and only a tiny amount of falloff outside of this circle when using a filter, and zero falloff when running unfiltered, with stars that are dead sharp to the edge.   It made using the Comet Catcher less and less enjoyable, and  I sold it.   It is being replaced with a 130mm f/5 imaging reflector that I hope will work with the ASA.  This has a large secondary and modern coatings (and no corrector and it was the corrector on the CC was only 90% in visual and less in IR). 

 

Again, unless the aperture is being reduced, the focal ratio does not change, but when you factor in all of the properties of these systems, the brightness is going to in reality be from a half to a full stop less than the raw focal ratio suggests it would be, and if the aperture is being reduced (a possible consequence of stacked reduction) the loss could be more than a stop even at the center of the field, but far greater at the edge of an APS-C chip, which is a good match for the size of a Gen 3 photocatode.

 

To anyone that has read my posts, I am a huge fan of using imaging Newtonians with NV.  They generally are designed to deliver very large fully illuminated fields, and with modern coatings, even with transmission loss and shading, can still produce brighter images than similar sized but slower refractors (with no reduction).

 

Sorry for the long post.   It is something that most people don't really delve into, but I think important if the goal is to get the best possible efficiency out of the system).

 

(For afocal, there is also transmission loss of eyepiece.  Many eyepieces are less than 90% in near IR and some much less). 


Edited by Eddgie, 30 May 2019 - 11:57 AM.


#8 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 538
  • Joined: 07 Mar 2014

Posted 30 May 2019 - 12:31 PM

Many thanks Eddgie for the detailed reply.

In terms of a brightness comparison, my direct experience is that my Tak Epsilon is much brighter when operating at f2 with my 41mm panoptic compared with my Baader 95mm app refractor operating at f2.8 with a 0.75x reducer and the same 41mm panoptic. So many different variables going on here I guess. 

 

After quite a bit of experimenting with nv on different kit, I’m impressed by how well my c11 delivers on nv with a reducer and 55mm plossl used afocally.


Edited by Gavster, 30 May 2019 - 12:53 PM.


#9 Eddgie

Eddgie

    ISS

  • *****
  • topic starter
  • Posts: 23741
  • Joined: 01 Feb 2006

Posted 30 May 2019 - 04:30 PM

Many thanks Eddgie for the detailed reply.

In terms of a brightness comparison, my direct experience is that my Tak Epsilon is much brighter when operating at f2 with my 41mm panoptic compared with my Baader 95mm app refractor operating at f2.8 with a 0.75x reducer and the same 41mm panoptic. So many different variables going on here I guess. 

 

After quite a bit of experimenting with nv on different kit, I’m impressed by how well my c11 delivers on nv with a reducer and 55mm plossl used afocally.

The Tak is the kind of scope that I recommend for NV in that it is a fast imaging scope with broad band coatings.   Some of the brightness may be a field illumination issue, and once again, this is why I posted this to start with.   Not every scope will provide even a fully illuminated field with a reducer.

 

Most reading this might not remember the big flub that affected Stellarvue when they first started.   When their first product was shipped (an 80mm Achormat) one of the testers felt that the image was a little dim.   The culprit turned out to be that when racked in fully to accomdate a 2" focuser, the front of the focuser tube cut into the light cone, reducing the aperture.  Now Vic stepped up quickly and not only changed the design, but made it right for existing buyers, but I use this to illustrate the problem.

With many fast refractors, if the focuser is not reasonably oversized, when fully racked in (as is the case when focal reducers are used), the focuser can vignette the off axis rays greatly reducing the size of the fully illuminated circle.   While the view might be full brightness at the very center, a few millimeters off axis, this might change.  

This is why I suggest that people do this kind of test.  It lets you directly see if your configuration is not working optimally and if not, one might be able to figure out how to tweak the configuration to improve it.

 

I am a big fan of the imaging Newtonians though. They give a fully illuminated field even with compressed and I for the last couple of years, I have been kind of trying to convince people of why this is important.   I am not at all surprised to hear how well the Tak does because it was a scope that was designed to give an outstanding result with a very fast focal ratio.   I am sure the coatings are broad band too, which is not common in visual use reflectors, where designers don't care about performace longer than 600nm. 



#10 Starman81

Starman81

    Soyuz

  • *****
  • Moderators
  • Posts: 3620
  • Joined: 06 Mar 2008
  • Loc: Metro Detroit, MI, USA

Posted 02 June 2019 - 07:09 PM

Many thanks Eddgie for the detailed reply.
In terms of a brightness comparison, my direct experience is that my Tak Epsilon is much brighter when operating at f2 with my 41mm panoptic compared with my Baader 95mm app refractor operating at f2.8 with a 0.75x reducer and the same 41mm panoptic. So many different variables going on here I guess.

After quite a bit of experimenting with nv on different kit, I’m impressed by how well my c11 delivers on nv with a reducer and 55mm plossl used afocally.

f/2 and f/2.8 sound close but that's a full f-stop right there! f/2 will be twice as bright, all other things held equal.

About the standard C11, who knows, it might be a Goldilocks scope for NV. Thankfully, they are in plentiful supply.

#11 Starman81

Starman81

    Soyuz

  • *****
  • Moderators
  • Posts: 3620
  • Joined: 06 Mar 2008
  • Loc: Metro Detroit, MI, USA

Posted 02 June 2019 - 07:17 PM

Eddgie, I think your work in the area of fully illuminated fields and focal reduction will long outlive you! Thanks.


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