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

Experiments in extreme focal reduction

NV
  • Please log in to reply
122 replies to this topic

#101 Kevdog

Kevdog

    Surveyor 1

  • *****
  • Posts: 1799
  • Joined: 11 Jul 2012
  • Loc: Desert Hills, AZ

Posted 08 August 2018 - 03:50 PM

I haven't found an app that does the stacking on the phone for android.   Is there one?   

 

If not, maybe I need to write one! :D



#102 Jeff Morgan

Jeff Morgan

    Voyager 1

  • *****
  • Posts: 10621
  • Joined: 28 Sep 2003
  • Loc: Prescott, AZ

Posted 08 August 2018 - 06:16 PM

I haven't found an app that does the stacking on the phone for android.   Is there one?   

 

If not, maybe I need to write one! laugh.gif

 

Nite Cap is iOS only?

 

Edit - just checked their web site. iOS only.


Edited by Jeff Morgan, 08 August 2018 - 06:17 PM.


#103 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 11 September 2018 - 02:50 PM

Back in July, I indicated in this thread that I'd do more testing when I had time.  So I set up last night in my red zone yard to do some additional tests with focal reduction AND with filters.  This took me from 21:00 to 02:30, but the results were telling.  I'll include my observations with the photos this time. 

 

The first set of photos were taken with the ST 120, f:5 using a 2" .5x reducer and 2" 7nm H-a filter with only 13mm between the reducer and the C-threads on the Mod 3C NVD.  The 2" 7nm H-a filter was screwed directly to the 2" reducer and this assembly was dropped into a 2" eyepiece holder on the diagonal.  Moshen previously recommended 11-12mm separation, but 13mm was as close as I could get with current adapters (2" to TT and TT to C-mount).  Here's an image of the assembly:

 

IMG_2411.JPG

 

The first image, .5s exposure, averaged for 15s was at ISO 1250.  Mild vignetting was visible at the eyepiece; the image shows it clearly.  This did not surprise me... I have not been able to use a generic .5x reducer in any of my scopes without seeing at least mild vignetting.  There is also mild field curvature visible with slightly elongated stars at the perimeter.  Visually, at the eyepiece of the Mod 3C, I could not detect the curvature seen in the images:

IMG_2413.jpg

 

Although in the past, I have referred to the NightCap Camera app for iOS as being able to stack images in the "long exposure" mode, I have learned that it doesn't really stack images in the sense that many programs align and combine images.  Instead, NightCap averages short exposure images.  By averaging the images, most noise and scintillation is removed, because the bright pixels caused by scintillation do not show up in the same place on the sensor during different exposures.  At least that's my understanding of the process.  Anyway, now I call it averaging instead of stacking.  

 

Second photo with the same mechanical setup but 1/6s averaged 15s at ISO 4000:  

IMG_2414.jpg

 

This second image is brighter, with the higher ISO and lower exposure setting.  Typically, a higher ISO means that more noise will be visible in the photo.  But I used the ISO boost setting in NightCap and the Noise Reduction setting.  Although neither of these first two images are well focused, neither image reveals much extraneous noise, which indicates that the ISO boost and noise reduction function do play a roll in performance when taking images with higher ISO settings.  

 

For the next 2 images, I put the 2" 7nm filter on the barrel of an Orion .8x reducer and screwed the reducer directly to the T2 diagonal.  Then used a C2TT adapter to add the Mod 3C directly to the top of the diagonal; this assembly slides into the telescope focuser.  This provides quite a short light path and looks like this:

IMG_2419.JPG

 

This arrangement of hardware provided a little better image with less vignetting and almost no visible field curvature in the image (none seen in the visual image at the eyepiece).  But the more interesting improvement came with much higher ISO settings where extraneous image noise was no greater than at the lowest setting.  This image was 1/2s averaged 15s at ISO 1000:

IMG_2416.jpg

 

This image was 1/16s at ISO 8000... which is the highest possible ISO setting for NightCap:

IMG_2417.jpg

 

The comparison of the last two images is a little unfair because the second resulted in a brighter image.  But the important issue here is that if you click on these images and expand them side by side, the higher ISO image shows no greater noise.  

 

The reason I jumped to such a high ISO was to use a shorter exposure time (1/16s) to see if shorter exposure plays a part in elongated stars at the perimeter.  I found that the shorter exposure did not effect that issue, but perimeter stars with this mechanical setup showed as more round than oval anyway.  I'll try this again using a setup with obvious elongated stars... maybe tonight.  

 

... more to follow...


Edited by GeezerGazer, 11 September 2018 - 06:01 PM.

  • AllanDystrup likes this

#104 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 11 September 2018 - 03:19 PM

The next optical train I set up was using a T2 diagonal with 1.25" eyepiece holder, a 1.25" nose with 7nm H-a screwed directly to the Mod 3C... this is a very standard kind of arrangement.

 

First is a 1/2s exposure averaged 15s at ISO 2000.  At the eyepiece I saw no vignetting, but the image appears mildly vignetted.  I am pretty sure this is a result of non full-field illumination; the 1.25" nose with 1.25" filter at the end may not be providing enough clear pathway for the light cone.  I have a shorter nose I'll try tonight to see if that makes a difference. 

IMG_2420.jpg

 

This image was 1/8s at ISO 8000.  Again, it appears a little brighter with greater detail, but it really is just the result of slightly more exposure from the higher ISO.  Although the perimeter is slightly darker than most of the field, the stars are sharp to the edge... at least as sharp as I actually focused them, which wasn't nearly good enough.  I have found that if stars appear elongated on only one side of the image, then the problem is most likely a tilted phone bracket.  If I see that happening, I removed the bracket and re-apply to get the phone's camera lens perpendicular to the ocular on the Mod 3C.

IMG_2421.jpg

 

I said that tonight's test was inclusive of filters... so here's an example.  With the same optical chain, I switched H-a filters to a 12nm.  This image was 1/2s averaged 15s at ISO 1600:IMG_2422.jpg

 

This was still with the 12nm filter but 1/8s at ISO 8000:

IMG_2423.jpg

 

So, let me put the 7nm and 12nm filter images both at 1/2s @ ISO 8000 averaged 15s side by side for a more direct comparison:

7 & 12.jpg

Except for the brighter image of the 12nm, detail and nebular extent was pretty much equal.  But remember that this is for imaging.  The difference at the eyepiece between these filters was pretty significant; increased scintillation because of the difference in light that was permitted to pass made the visual image of the narrower band much different.  This is where personal preference plays a really big part in what you choose to use.  Narrower band = more scintillation visually because of photon starvation. Actually, I did the same exact test using a 5nm filter, so in the next post, I'll include a comparison of the 5nm and the 12nm. 


Edited by GeezerGazer, 11 September 2018 - 06:23 PM.

  • Rickster and AllanDystrup like this

#105 Kevdog

Kevdog

    Surveyor 1

  • *****
  • Posts: 1799
  • Joined: 11 Jul 2012
  • Loc: Desert Hills, AZ

Posted 11 September 2018 - 03:48 PM

 

Although in the past, I have referred to the NightCap Camera app for iOS as being able to stack images in the "long exposure" mode, I have learned that it doesn't really stack images in the sense that many programs align and combine images.  Instead, NightCap averages short exposure images.  By averaging the images, most noise and scintillation is removed, because the bright pixels caused by scintillation do not show up in the same place on the sensor during different exposures.  At least that's my understanding of the process.  Anyway, now I call it averaging instead of stacking.  

 

I was looking to write my own Camera software for Android as there aren't any good programs like NightCap.   I was researching stacking and the algorithms to do it, so I could program it in my phone.   All stacking is actually averaging the pictures.   Some do a straight up average of the pixels, others do it a bit fancier, but it is still basically the same thing.   So they are (for the most part) synonyms as we use them.


  • GeezerGazer likes this

#106 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 11 September 2018 - 04:54 PM

OK, so the same scope, same settings, EXCEPT the 5nm filter required a slightly slower exposure speed of 1/6s for 15s at ISO 8000... the 12 and 7nm filters were at 1/8s for 15s at 8000:12, 7, 5.jpg

L to R:  12nm, 7nm, 5nm filters.  I do see an improvement in detail as I look at the images from L to right all taken with the same scope at near the same setting.  

 

This is not to say that visual performance is the same with these filters.  There are tradeoffs.  The narrower the band, the more scintillation I would see, especially with the Mod 3C gain on full.  And with the 5nm, visually it was not pleasing with too much scintillation.  When the 5nm was on, I took a single .5s exposure (no averaging) to approximate what I could see at the eyepiece... it wasn't pretty:

IMG_2439.jpg

My L3 tube isn't the best with an EBI of 1.1, but in this case, the photo using a 5nm filter was infinitely better than the visual image at the eyepiece of the Mod 3C.

 

I changed the 1.25" nose for a very short (6mm lightpath) 1.25" to C-mount adapter and put a 1.25" .5x GSO/scopestuff reducer on with a 7nm filter in front and put that in the diagonal.  No surprises... mild vignetting, probably non full-field illumination, but at the eyepiece of the Mod 3C, it was an acceptable image with a slightly smaller scale than the native ST 120 provides.  Sorry, for some reason the planned image will not load here. 

 

Next I changed scopes, going to a ScopeStuff VersaScope 60mm @ f:3.8.  In using this tiny scope I removed the nose of the T2 diagonal and screwed the diagonal directly to to the back of the scope which has T2 threads.  In this configuration, I can keep the largest clear path for the light cone (34mm diameter) all the way to the C2TT adapter which is attached directly to the Mod 3C.  So here is a 1/10s exposure stacked 15s at ISO 8000:IMG_2435.jpg

Even trying to keep the light path open, I see mild vignetting likely from non full-field illumination.  Visually, at the Mod 3C, it looked pretty nice, but images do reveal vignetting.

 

That's it for now.  Tonight I actually get to spend time observing! smile.gif  

Ed, I think now I'm ready to buy a Boren-Simon! bow.gif


Edited by GeezerGazer, 11 September 2018 - 06:08 PM.

  • Rickster and AllanDystrup like this

#107 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 11 September 2018 - 05:57 PM

I was looking to write my own Camera software for Android as there aren't any good programs like NightCap.   I was researching stacking and the algorithms to do it, so I could program it in my phone.   All stacking is actually averaging the pictures.   Some do a straight up average of the pixels, others do it a bit fancier, but it is still basically the same thing.   So they are (for the most part) synonyms as we use them.

Thanks.  Averaged or stacked doesn't really matter to me, but the clarification is nice.  I'm hoping to have a camera soon that will take one long exposure that I can watch build over 20-30 seconds, like Gavster's Huawei P20 Pro!  My 6+ iPhone is getting long in the tooth!  

 

I haven't a clue how to program.  Another skill I never learned!  Good luck with that project. 


Edited by GeezerGazer, 11 September 2018 - 06:32 PM.


#108 Jeff Morgan

Jeff Morgan

    Voyager 1

  • *****
  • Posts: 10621
  • Joined: 28 Sep 2003
  • Loc: Prescott, AZ

Posted 11 September 2018 - 07:34 PM

is not to say that visual performance is the same with these filters.  There are tradeoffs.  The narrower the band, the more scintillation I would see, especially with the Mod 3C gain on full.  And with the 5nm, visually it was not pleasing with too much scintillation.  When the 5nm was on, I took a single .5s exposure (no averaging) to approximate what I could see at the eyepiece... it wasn't pretty:

attachicon.gif IMG_2439.jpg

My L3 tube isn't the best with an EBI of 1.1, but in this case, the photo using a 5nm filter was infinitely better than the visual image at the eyepiece of the Mod 3C.

 

As I am figuring out NiteCap, it occurred to me that the biggest variable will be the manual gain setting on the tube.



#109 AllanDystrup

AllanDystrup

    Apollo

  • *****
  • Posts: 1407
  • Joined: 27 Sep 2012
  • Loc: Denmark

Posted 12 September 2018 - 10:41 AM

I was looking to write my own Camera software for Android as there aren't any good programs like NightCap.   I was researching stacking and the algorithms to do it, so I could program it in my phone.   All stacking is actually averaging the pictures.   Some do a straight up average of the pixels, others do it a bit fancier, but it is still basically the same thing.   So they are (for the most part) synonyms as we use them.

Kevdog,

   

     Well yes, stacking is a form of averaging, but the reverse is not true. The difference being that stacking corrects for tracking inaccuracies (in the simple case by using “virtual” guide stars) while averaging does not.

 

     That is a significant difference. For live video, my R2 cam can do averaging (like my iPhone), but my Infinity cam can do stacking. Oh yes, there is a difference :-).

     

     But then again: I’m using a simple classic (Zeiss) manual motorized mount without fancy 3-star alignment, go-to, PEC etc. So maybe (probably) with a modern hi-tec auto-guiding mount averaging can indeed be close to stacking. And you can argue that with an exposure time ~5s, it will be hard to tell the diff anyways... (Hmm, painting myself into a corner here? wink.gif ). But with 30s or more...

 

Allan


Edited by AllanDystrup, 12 September 2018 - 11:16 AM.

  • GeezerGazer, john hraba and Kevdog like this

#110 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 12 September 2018 - 03:11 PM

As I am figuring out NiteCap, it occurred to me that the biggest variable will be the manual gain setting on the tube.

I don't think gain is the "biggest" variable, but it is certainly important to me, especially in regards to visual use.  For images with the phone and for other cameras that average or stack images, the best place for gain seems to be full on as it provides the best light enhancement for capture.  The ability to track a target for visual NV use is not terribly important, but for images, it is critical.  Sensitivity of camera sensors and the size of the sensor is also important.  For visual NV use, I think the speed and size of the lens gathering the light is the most critical or "biggest" variable.  I wish all of my NV lenses were f:1.2 and that each was at least 8" in diameter !!!  If wishes were pennies...

 

But there are technical advancements on the horizon and many of them will mitigate the performance of slower optics.  The SiOnyx Aurora camera has a very low light sensor... it's a bigger 1" sensor with bigger individual pixels for incredible daylight or extremely low light video performance; the technology "could" become a rival to our expensive NV devices if it is developed.  Night Vision is the only method to date that actually enhances available light.  But camera sensors are getting so good at detecting low levels of light and showing them on a screen... well, who knows for sure.  

 

I did observe again last night but I also took a few photos.  I also completed two additional tests using my TEC 140 @ f:7 instead of the f:5 ST 120.  I tried using a shorter 1.25" nose on the Mod 3C to mitigate non-full field illumination but it did not help.  I also tried eliminating the nose altogether, using a C2TT adapter between the T2 diagonal and the Mod 3C, placing a .5x 2" reducer ahead of the C2TT adapter, but still saw mild vignetting.  The TEC 140 has a 3.5" focuser, so this vignetting is not coming from a smaller 2" focuser.  The .5x reducer also caused field curvature in images I took.  Trying the .8x reducer eliminated or nearly eliminated the field curvature, but still resulted in very mild vignetting.  I may keep the .8x reducer though because visually, I did not notice the vignetting and it was a slightly smaller scale image that is easily inserted into the light path making it fairly convenient in use. 


  • AllanDystrup likes this

#111 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 12 September 2018 - 06:12 PM

Here are some of my thoughts about "Experiments in extreme focal reduction"

 

First, it has been an eye opener for me.  Having tried using NV in prime AND in afocal with my refractors, I much prefer prime.  And as such, I am limited to the use of reducers... I chose to experiment with generic reducers, 1.25" .5x, 2" .5x and a 2" .8x.  In my ST-120 and my TEC-140, afocal results were marginal for my tolerance level of field curvature, astigmatism and vignetting; when I added a focal reducer to afocal, the results with my scopes were beyond my tolerance level.  But I also found that results which exceeded my threshold for images, could sometimes be satisfactory visually... looking through the NVD.  The distortions that seem to make no difference visually are exacerbated in imaging, probably because of the difference in how I look through an eyepiece as opposed to looking at a photo.  Use of my 3 reducers also revealed that the less reduction they were required to create, the better they performed... moving them closer to the focal plane produced fewer aberrations but also produced less focal ratio reduction.  Even so, I could not manage to use a reducer with my scopes without seeing at least mild vignetting.  

 

For those of us using NV in prime mode with refractors, it seems much more productive to scale up with a barlow than to scale down with a generic reducer.  Reducers seem to impose more optical limitations than barlows making it more difficult to obtain excellent optical results.  The exception is probably the dedicated reducers/flatteners that are made specifically for intended telescopes.  But these cost a lot more money and, of course, I'd have to own the intended scope.  Fast Dobs/Newts typically require a corrector/flattener/Paracor before effective alteration of focal ratios.

 

I also have found that if I want a different scale, it is almost as easy to use a different scope as to implement a reducer.  On my tracking mount, I can easily remove my TEC-140 and replace it with the ST-120, the VS60, or even just the Mod 3C with a 50mm or 105mm camera lens.  AND, the cost of an 80-100mm f:5 achromat refractor, which perform very well with NV, is not much more than a generic reducer!  So as a step between my 60mm and 120mm, it might be a more logical choice. 

 

Does anyone want to purchase a really nice trio of barely used reducers?  lol.gif

 


  • AllanDystrup likes this

#112 Jeff Morgan

Jeff Morgan

    Voyager 1

  • *****
  • Posts: 10621
  • Joined: 28 Sep 2003
  • Loc: Prescott, AZ

Posted 12 September 2018 - 09:24 PM

Thanks for doing the legwork on this, I have enjoyed the thread.

 

Yes, there is no Free Lunch.

 

Fortunately, visual use is a bit more forgiving.


  • GeezerGazer likes this

#113 GeezerGazer

GeezerGazer

    Viking 1

  • *****
  • Posts: 892
  • Joined: 06 Jan 2005
  • Loc: Modesto, CA

Posted 13 September 2018 - 01:45 AM

Thanks for doing the legwork on this, I have enjoyed the thread.

 

Yes, there is no Free Lunch.

 

Fortunately, visual use is a bit more forgiving.

 

Yes, it is more forgiving.  When looking into my Mod 3C ocular, I'm usually concentrating on or near the center FoV.  When I look at a photo or electronic image, I may initially look at the main subject, but then I scan the entire image, so aberrations show up more readily.  The reason I see photo or electronic image vignetting is because the illumination of an image is not even; when a sensor is gathering light across the sensor and the perimeter receives fewer photons because it was not fully illuminated, it shows up.  Our eye/brain doesn't gather and process light that way, so at the ocular visually, we often can't see it.  

 

Jeff, I know you are going to create some great images with your Epsilon and your Newt.  And yes, after the learning curve, NightCap is pretty simple.  The application of smart phone, NVD and telescope is a fast and fun way to achieve very decent images.  



#114 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 539
  • Joined: 07 Mar 2014

Posted 21 May 2019 - 07:56 AM

I’ve just been re-reading this very interesting thread. I note that the afocal comparisons were all done with Televue plossls and with the relatively fast scope focal ratios being used (f7 and below), the plossls do struggle with edge of field distortion (astigmatism and fc). However, with much better corrected eyepieces for fast setups such as the 41mm panoptic, the afocal approach will give much better performance across the fov. I can also use the 41mm panoptic well with a 0.75x reducer. 

So my view is that if afocal views disappoint with the plossl then its definitely worth trying some panoptics. 


  • Starman81 likes this

#115 Starman81

Starman81

    Soyuz

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

Posted 21 May 2019 - 09:59 AM

I’ve just been re-reading this very interesting thread. I note that the afocal comparisons were all done with Televue plossls and with the relatively fast scope focal ratios being used (f7 and below), the plossls do struggle with edge of field distortion (astigmatism and fc). However, with much better corrected eyepieces for fast setups such as the 41mm panoptic, the afocal approach will give much better performance across the fov. I can also use the 41mm panoptic well with a 0.75x reducer. 

So my view is that if afocal views disappoint with the plossl then its definitely worth trying some panoptics. 

Gavster, coincidentally, I had to do another read-through myself in the past week to get re-familiarized with the findings. 

 

The one thing that had me backing off the Panoptics (or any eyepiece with greater than 40 degree AFOV) was the concern over vignetting...  In retrospect, I think I was thinking about it the wrong way... Is it that the NV device is only taking the 40 degrees AFOV from whatever available total AFOV of the eyepiece being used afocally and then it presents this with the sharp field stop of the NV device itself... Is this correct? 

 

Further, since Panoptics are well-corrected for the vast majority of their 68° FOV, displaying a well corrected 40 degree AFOV should then indeed be a piece of cake, even with a mild reducer, like the 0.75x you're using. 



#116 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 539
  • Joined: 07 Mar 2014

Posted 21 May 2019 - 11:13 AM

Gavster, coincidentally, I had to do another read-through myself in the past week to get re-familiarized with the findings. 

 

The one thing that had me backing off the Panoptics (or any eyepiece with greater than 40 degree AFOV) was the concern over vignetting...  In retrospect, I think I was thinking about it the wrong way... Is it that the NV device is only taking the 40 degrees AFOV from whatever available total AFOV of the eyepiece being used afocally and then it presents this with the sharp field stop of the NV device itself... Is this correct? 

 

Further, since Panoptics are well-corrected for the vast majority of the FOV, displaying a well corrected 40 degree AFOV should then indeed be a piece of cake, even with a mild reducer, like the 0.75x you're using. 

Yes and yes!

The views with my 41mm panoptic are definitely more aesthetically pleasing than my 55mm plossl. Obviously there is less reduction unfortunately hence why i also want to use a reducer in conjunction with the 41mm pan to aim to get to a similar 0.5x reduction that I get with the .55mm plossl on its own. It’s a shame that the 41mm doesn’t have the out focus position that the 55mm plossl has since I think that limits it to 0.75x rather than the 0.5x reducer. I do still use the 55mm plossl with my c11 edge and 0.7x reducer since the edge of field distortion is less compared with my refractors.

My f3.3 Epsilon does push the capabilities of the 41mm pan to the limit with the edge stars showing a bit of distortion but worth it to get to f2. The 55mm plossl gets the Epsilon down to f1.6 but the edge stars become rather linear smile.gif But very nice and bright nebulae in the centre - I need to do some phone shots to compare.

I’m going abroad this summer and am still not sure what travel refractor to take. They vary in terms of f ratio from f5.9 to f7 - I need to confirm that the f5.9 with the 41mm pan and reducer is better than the f7 with a 55mm plossl and reducer. Hopefully I can check this out later tonight.


Edited by Gavster, 21 May 2019 - 11:14 AM.

  • Starman81 likes this

#117 Eddgie

Eddgie

    ISS

  • *****
  • Posts: 23744
  • Joined: 01 Feb 2006

Posted 21 May 2019 - 05:07 PM

For prime focus, the limit here is the 22mm opening in front of the photocathode.

While the interface is 1", if you measure the inside of the 1.25" nose piece, you will find that they are about 22mm on the ID.  This means that you have a 22mm hole that is place about 15mm in front of the photocatode. This varies to some small extent, but the distance to the window is generally about 12-13mm, and the window is 1.8mm thick). (Also, the official specs actually place the size of the photocathode as between 17.5mm and 17.7mm, but my guess is that it is 18mm and they are allowing for the vignetting of the edge of the photocathode window itself, which will partly shade a very small area around the very outside of the photocatode, but I have ignored that here and gone with 18mm).

 

This means that if you start with a 22mm light cone at f/2, the cone will converge to a point at 44mm behind the opening.  If you look at the convergence 15mm behind the 22mm restriction, the cone will have traveled about 29% of the distance to its focal point and hence the fully illuminated circle will have been shrunk down to about 15.2mm, so even at f/2, the illumination fall-off has already started. This is just a basic geometry problem. Any circular constriction placed in front of the focal plane sets the limit to the size of the light circle that can get though it, and the distance traveled after that light cone passes through that constriction, the focal ratio of the light cone, and the distance it travels will dictate the size of the fully illuminated circle when it gets to the focal plane. 

 

The use of a 1.25" nose on the device makes this situation much worse because now the focal plane is very far back from the circular obstruction.  This means that this configuration will only yield a fully illuminated circle about 5mm in diameter.

 

It is just geometry.  If one uses a 1.25" diagonal with with filters, I recommend the short nose on the device, and the shorter, the better.  Direct connection to a 1.25" diagonal is better still, and a 2" nose on a Baader T2 is best (down to about f/5 if using reducers).  Of course this means that a 2" filter has to be used, so a filter wheel or filter drawer is a good compromise at f/2.5.  This is going to give a pretty well sized fully illuminated circle.

 

Also, a  standard 1.25" diagonal will almost always cause vignetting.  It is very soft, but there none the less.  If the ID of the 1.25" nose is 27mm and the light path is 70mm (about the light path of a 1.25" Prism) added to the 15mm flange to focal plane distance that is 85mm.   At f/7, a 27mm light cone would travel 189mm before it converged so over the 85mm to the focal plane, it will have been reduced in size from 27mm to about 11mm.   And of course this gets far worse as the scope gets faster.

 

Take the case for a Baader 1.25" with 34mm opening.  Light path is about 40mm for the body and 25mm for the eyepiece holder and 15mm for the flange to focal plane. (Assuming a 2" nose is used, the front aperture becomes the limit).

 

This means that at f/5, a 34mm wide circle would converge in 170mm.  A 34mm restriction would shrink the light cone down to 47% of its starting size, so this is down to 15.8mm.   And this is the native f/5 light cone.   That means that if you put a focal reducer in front of the diagonal, the focal reducer should be mounted directly to the body (using 48mm to T2 connector) and the device should be connected directly to the top of the diagonal.  

 

Of course the problem with 2" diagonals is that they don't work with a focal reducer in front of them, and they don't have enough back focus to work with a reducer behind them.

 

This is really a geometry problem. Light cones converge, and reducers not only reduce the size of the full illuminated cone with respect to the amplification (a 20mm light cone at the focal plane without a reducer becomes a 10mm light image circle with a .5x reducer) but also requires very large openings for anything any distance behind the reducer (a 2" to C mount nose would not be the optimal way to use a reducer with a T2 diagonal unless the device were connected directly to the top of the diagonal. 

 

It is important to say that for visual use, some vignetting or illumination falloff is almost always present, but we don't easily see it (as is the case with normal eyepieces, where even 50% illumination falloff is not always seen by many observers) but the camera does not allow you to hide from this kind of condition.  The longer the shutter is open, the more it asserts itself. 

 

I hope this was not painfully boring.  I know a lot of people don't think about stuff like this, but I give it a great deal of thought when selecting equipment. 


Edited by Eddgie, 21 May 2019 - 05:42 PM.

  • Starman81, Vondragonnoggin and chemisted like this

#118 moshen

moshen

    Apollo

  • *****
  • Posts: 1252
  • Joined: 17 May 2006
  • Loc: San Francisco, CA

Posted 21 May 2019 - 06:51 PM

My current solution is I mount the 2" 0.5x reducer on top of my filter wheel using a T2-48mm adapter. Then I had someone machine a 48mm to C mount adapter that goes between the reducer and Mod3. The length of this adapter is custom and enough to give me focal reduction with minimal (but still some) vignetting.

I don't get the full 0.5x reduction with this distance. More like 0.65x, but good enough for me.
 
The filter wheel attaches directly to the T2 diagonal with a Baader T2 quick change ring.
 
I'm actually not too sure where the light is being vignetted. I would think it's in the 1.25" filter wheel holes but that's not it. When I move the filter ring, it does not 'unvignette' part of the view as the holes move to the edge of field. It could be the path between the focal reducer and Mod3. Or my diagonal - larger than 1.25" but smaller than a full 2" diagonal with 34mm clear aperture.
 
Either way this setup still works very well for me even with slight vignetting.
 
IMG_0809.jpg

 


  • Starman81, Vondragonnoggin, Gavster and 1 other like this

#119 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 539
  • Joined: 07 Mar 2014

Posted 21 May 2019 - 07:04 PM

Ok so I did test my three refractors with the 55mm plossl and 0.75x reducer combo vs the 41mm panoptic and 0.75x reducer.

The visual views were pretty similar across all three scopes with the 55mm Plossl,  showing some fc at f7 (x0.75 for reducer) as well as at f5.9 (x0.75 for reducer. The f7 scope was slightly less affected by the fc but it was still there.

The first image of the North America (20 sec exposure) shows the edge star issue quite well.

All scopes liked the 41mm pan plus 0.75 reducer combo. See other two pics of North America and crescent nebula (10 sec exposure) with my f5.9 refractor. Not great pics but I was viewing them from my 18 sqm London back garden. 

My conclusion is that the f5.9 refractor together with 41mm pan and 0.75x reducer is the best combo - it reduces the system f ratio to 2.8 and provides nice views across the full fov. 

Attached Thumbnails

  • 2BF9F80C-EAE2-4D8C-A278-977A2103E0CE.jpeg
  • 494D3A5B-54D9-465C-B13F-A1705154CB3A.jpeg
  • 2372DE75-5311-4FD5-8615-D5BA19E862CB.jpeg

  • moshen, Starman81, Vondragonnoggin and 1 other like this

#120 Vondragonnoggin

Vondragonnoggin

    Cosmos

  • *****
  • Posts: 8400
  • Joined: 21 Feb 2010
  • Loc: Southern CA, USA

Posted 22 May 2019 - 12:07 AM

 

My current solution is I mount the 2" 0.5x reducer on top of my filter wheel using a T2-48mm adapter. Then I had someone machine a 48mm to C mount adapter that goes between the reducer and Mod3. The length of this adapter is custom and enough to give me focal reduction with minimal (but still some) vignetting.

I don't get the full 0.5x reduction with this distance. More like 0.65x, but good enough for me.
 
The filter wheel attaches directly to the T2 diagonal with a Baader T2 quick change ring.
 
I'm actually not too sure where the light is being vignetted. I would think it's in the 1.25" filter wheel holes but that's not it. When I move the filter ring, it does not 'unvignette' part of the view as the holes move to the edge of field. It could be the path between the focal reducer and Mod3. Or my diagonal - larger than 1.25" but smaller than a full 2" diagonal with 34mm clear aperture.
 
Either way this setup still works very well for me even with slight vignetting.

 

It also looks very drool-worthy!  drool5.gif

 

Very nice setup.

 

cool.gif


  • moshen likes this

#121 Starman81

Starman81

    Soyuz

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

Posted 23 May 2019 - 03:41 PM

Ok so I did test my three refractors with the 55mm plossl and 0.75x reducer combo vs the 41mm panoptic and 0.75x reducer.

The visual views were pretty similar across all three scopes with the 55mm Plossl,  showing some fc at f7 (x0.75 for reducer) as well as at f5.9 (x0.75 for reducer. The f7 scope was slightly less affected by the fc but it was still there.

The first image of the North America (20 sec exposure) shows the edge star issue quite well.

All scopes liked the 41mm pan plus 0.75 reducer combo. See other two pics of North America and crescent nebula (10 sec exposure) with my f5.9 refractor. Not great pics but I was viewing them from my 18 sqm London back garden. 

My conclusion is that the f5.9 refractor together with 41mm pan and 0.75x reducer is the best combo - it reduces the system f ratio to 2.8 and provides nice views across the full fov. 

Nice results, Gavin. :waytogo:

 

I don't have the Pan 41 but I do have the Pan 35 that I can try with. 



#122 Jeff Morgan

Jeff Morgan

    Voyager 1

  • *****
  • Posts: 10621
  • Joined: 28 Sep 2003
  • Loc: Prescott, AZ

Posted 23 May 2019 - 05:36 PM

Worth a look. A 41 Panoptic and 0.7x reducer should be about the same reduction as a 55 Plossl.



#123 Gavster

Gavster

    Viking 1

  • -----
  • Posts: 539
  • Joined: 07 Mar 2014

Posted 24 May 2019 - 01:15 AM

Worth a look. A 41 Panoptic and 0.7x reducer should be about the same reduction as a 55 Plossl.

Yes from my eyepiece comparisons virtually equivalent.

However, sometimes I just can’t resist going for the 0.75x reducer and 55mm plossl combo to act as a 0.35x reducer. The inner 3/4 of the fov is fine and from light polluted areas it does bring out the nebulosity more. Running at f3 is nice but running at f2 or less is nicer smile.gif (particularly with narrow 3 or 5nm ha filters, not so needed when using longpass filters)


Edited by Gavster, 24 May 2019 - 01:16 AM.



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