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

Use of barlow lenses during planetary imaging

  • Please log in to reply
21 replies to this topic

#1 jsnatale

jsnatale

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 13 Aug 2021

Posted 15 September 2021 - 11:09 AM

I have recently started planetary astrophotography, and use a Celestron CPC 800 HD EDGE telescope, with go-to. The camera I use is a ZWO ASI 224 MC. I have a general question on whether or not to use barlow lenses. 

 

In the case of Jupiter and Saturn, it seems like I get the best images when using no barlow lens at all. The planet is fairly small when capturing, compared to when I use a barlow. I have been using autostakkert and registax for processing. When I am done (using no barlow) I just enlarge the planet using MS paint, word, or another program, so I end up with a fairly large image that also has some decent detail. 

 

When using barlow lenses I get less planetary detail, and the image is obviously larger when capturing. But it is difficult keeping the planet on the screen when it is larger- there tends to be more drift. If you have an 8" SCT like mine, which do you find works best? No barlow, or a 2X or 3X barlow?  I have seen pictures from others who have had great success using barlow lenses (or focal expenders) but it seems to me I am better off without them with my set-up. What has your experience been? If any of you have a CPC 800 like mine, what have you been using?

 

Another question.  Does anyone know how to calculate the magnification I get when using my ZWO ASI 224 camera, with no barlow?  Since there is no eyepiece in the image train, I can't divide telescope focal length by the eyepiece focal length... is there a table or guide somewhere that lists magnifications of different ZWO cameras for different aperature sizes? I am really curious. Jupiter and Saturn are pretty large when capturing, so I am guessing there is some pretty significant magnification going on there.


Edited by jsnatale, 15 September 2021 - 11:12 AM.


#2 ngc7319_20

ngc7319_20

    Vanguard

  • *****
  • Posts: 2,271
  • Joined: 25 Oct 2015
  • Loc: MD

Posted 15 September 2021 - 11:51 AM

 

Another question.  Does anyone know how to calculate the magnification I get when using my ZWO ASI 224 camera, with no barlow?  Since there is no eyepiece in the image train, I can't divide telescope focal length by the eyepiece focal length... is there a table or guide somewhere that lists magnifications of different ZWO cameras for different aperature sizes? I am really curious. Jupiter and Saturn are pretty large when capturing, so I am guessing there is some pretty significant magnification going on there.

 

The concept of "magnification" usually applies only to visual observing.  For imaging people usually go by "arcseconds per pixel."

 

There are various calculator tools online.  Here is one:

 

https://astronomy.to...calculators/ccd

 

I have used a Barlow when imaging planets with my C8.  But you want a good quality Barlow (or PowerMate, etc.).  And as you say, tracking becomes more challenging.  But ultimately under the best "seeing" conditions there is a bit more detail available with the Barlow.

 

Your camera has 3.75 micron pixels, so the native C8 works out to 0.4 arcseconds per pixel.  The resolution of the C8 is about 0.5 arcseconds.  But to record the full detail available in the image, you want to over-sample the image with the pixel grid about twice as fine as the optical resolution -- hence about 0.25 arcseconds per pixel would be optimal, and around a 2x Barlow.  But this is not a hard and fast rule -- depending on many factors like atmospheric seeing, tracking, color vs mono, etc -- some other pixel scale might be better.  When all else fails -- experiment.


Edited by ngc7319_20, 15 September 2021 - 12:06 PM.


#3 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted 15 September 2021 - 12:10 PM

Your image scale in arcseconds per pixel = given by:

206.3 x (camera pixel size in microns) / (telescope focal length in mm)

Your C8 at nominal focal length is ~2000mm, although you may not be actually imaging at this focal length. Your pixel size on the ASI224 is 3.75 microns, which is fairly large. The rule of thumb is that critical sampling occurs at a focal ratio of about 5 times the pixel size in microns, although this is seeing dependent. So critical sampling for the 224 would be at about f/19. If you are actually at your scopes nominal focal length you are imaging at f/10, which is very under-sampled and can produce quite sharp but small images. If you were to shoot at around f/19, are well collimated and in focus, you should definitely be able to pick up more details for the same seeing. However there are some caveats. Doubling your focal ratio reduces the image brightness and hence your signal to noise ratio by a factor of 4. So you would need to shoot and stack 4 times as many frames to get those details back to the same SNR, all else being equal, and it is not equal because you will almost certainly have to adjust your exposure and or gain between the 2. You can brighten the image and increase your SNR by longer exposures, but then you are convolving in more atmospheric turbulence. You can brighten the image by increasing the gain, but increasing the hardware gain above a certain point (that depends on the camera) will indeed make the image brighter but will also increase the noise faster, also reducing SNR, and once again requiring more frames. The end result is that you may need to stack an order of magnitude more frames with the Barlow than without to see those extra details, presuming you are well collimated, in focus, not plagued by at atmospheric dispersion, and your processing is good enough to pull those details out.

Edited by Borodog, Yesterday, 07:46 AM.

  • jkmccarthy likes this

#4 jkmccarthy

jkmccarthy

    Vostok 1

  • *****
  • Posts: 183
  • Joined: 10 Mar 2021
  • Loc: Los Angeles, California U.S.A.

Posted 15 September 2021 - 12:18 PM

Also, how much magnification you actually get from the barlow depends on the separation between the barlow and the camera focal plane.  When you compare the diameter of the planet (measured simply in units of imager pixels) with the barlow compared to without, is the planet's image 2X larger with the barlow, or significantly more ?   The ratio of measured diameters will tell you how much of an increase you're actually getting from using the barlow (alone, with no eyepiece, correct?) with your imager setup.

 

Elsewhere in these forums I've read that for capturing the planets, an effective focal length (EFL) of around 5-meters is thought by many to be optimum for use with typical imaging cameras ... so as a sanity check, you might multiply your telescope's focal length by the planet-diameter-increase-ratio you measure with the barlow, and see if your setup with barlow is delivering an EFL of around 5m.

 

Lastly, it's worth keeping in mind that with the barlow, the total light received from the planet will be spread out over many more pixels (how many more is given by the planet-diameter-increase-ratio squared ... 4x more pixels if the size of the planet is doubled using the barlow), so therefore the image is fainter with the barlow vs. without.  Are you using a longer exposure time (i.e., shooting with fewer frames per second of video) due to the fainter image ?  If frame rate is too low, atmospheric seeing and/or image motion could explain some of the loss of sharpness you're seeing.  If the frame rate is the same, then lower signal-to-noise ratio could be a factor (suggesting you might want to try capturing longer video sequences for doing your post-processing) [*].

 

Good luck,

 

       -- Jim

 

[*] EDIT ... or as Borodog's reply (posted while I was busy composing mine) suggests, increasing the gain provided this doesn't increase the noise even more than the signal.

 

Regarding Borodog's formula, I can shed light on where the "206.3" factor comes from:  the number of arcseconds in one radian is equal to 206,265 but since his formula uses mm for telescope focal length but microns for the pixel size (where 1 micron = 1/1000 mm), that factor of 1000 difference in the units reduces 206265 arcsec down to 206.3 ... so nothing mysterious there in the image scale calculation.

 

Regarding his "rule of thumb" for f/ratio = 5 x (pixel size in microns), note that because f/ratio = EFL / (Telescope Diameter), this is the same thing as saying the optimum EFL = 5 x (pixel size in microns) x (Telescope Diameter).   So if one then takes this EFL and plugs it back into the first formula, one finds the "rule of thumb" target image scale is:

 

Target image scale (arcsec/pixel) = 206.3 / (5 x Telescope Diameter in mm)

 

... or equivalently --

 

Target image scale = 0.2 arcsec/pixel x (206.3 mm / Telescope Diameter)

 

... so for an 8-inch telescope, the target image scale is ~ 0.2 arcsec/pixel, and coarser sampling can be targeted for smaller aperture telescopes while finer sampling should be targeted for larger aperture telescopes (to take maximum advantage of the moments of very best seeing).

 

Lastly, it appears the 5m EFL target I quoted was most appropriate for 10-inch aperture telescopes, whereas per the calculations above starting with Borodog's formulae, a 4m EFL target might be more appropriate for 8-inch aperture telescopes ... assuming imager pixel size of ~ 4 microns in both cases.  ((Larger pixel sizes require proportionally longer EFLs to maintain the same arcsec/pixel scale, so EFL ~ 5m might be optimum for an 8-inch aperture telescope in the case of ~ 5 micron pixels....))

 

Hope this helps.


Edited by jkmccarthy, 15 September 2021 - 01:19 PM.


#5 matt_astro_tx

matt_astro_tx

    Apollo

  • -----
  • Posts: 1,439
  • Joined: 19 Jan 2021
  • Loc: Dallas, Texas

Posted 15 September 2021 - 12:20 PM

I just responded to your Neptune/Barlow thread.  The calculations I posted there are the same for Jupiter and Saturn, or any planet you want to capture, so check that post out.

 

In short, you want a 2x or a 2.5x barlow for your camera/scope combination.


Edited by matt_astro_tx, 15 September 2021 - 12:20 PM.

  • RedLionNJ likes this

#6 jkmccarthy

jkmccarthy

    Vostok 1

  • *****
  • Posts: 183
  • Joined: 10 Mar 2021
  • Loc: Los Angeles, California U.S.A.

Posted 15 September 2021 - 03:29 PM

In short, you want a 2x or a 2.5x barlow for your camera/scope combination.

Personally, I struggle a bit with these f/ratio -based rules of thumb ... maybe it's just "my problem" and I should keep it to myself, since for those just starting out, there's certainly a great deal of simplicity offered by a "rule of thumb" that relates just two quantities, pixel size and f/ratio.

 

But on the other hand, f/ratio is a somewhat abstract  concept, and moreover it really can't easily be determined directly by examining a captured image.   This is why in post #4 above I thought there might be value in re-expressing the "rule of thumb" to be instead a recommendation for optimum image scale, in arcsec/pixel.   Roughly speaking, 0.2 arcsec/pixel  (or its reciprocal, 5 pixels per arcsec) is a sensible and easily understood "sweet spot" compromise between having pixels that too coarsely sample the image (such that pixel size would limit the amount of detail captured in moments of very best seeing) and having pixels that too finely sample the image (such that the combination [mathematically speaking, the convolution] of the telescope's optical blur size and the atmospheric seeing blur size gets spread out over too many pixels and is consequently too faint to record with good signal-to-noise ratio at a fast frame rate) with a typical 8-inch telescope.  Granted, larger aperture telescopes (providing more photons, and having smaller optical blur size) can and should be used at an image scale proportionally finer than 0.2 arcsec/pixel, and the opposite is true of smaller aperture telescopes.

 

In my mind equally important, though, is that image scale is straightforward to determine directly by examining the image one has captured.  Where I think it's easy for those just starting out to be led astray using the f/ratio rule is the temptation to just multiply the telescope's nominal f/ratio by the nominal 2X or 3X number printed on the side of a barlow lens, and then assume that represents the final f/ratio of the setup.  But as I pointed out earlier, the change in effective focal length (or f/ratio) produced by a barlow depends on how far ahead of the image plane the negative lens is located.  Unfortunately for astrophotographers, the 2X or 3X number printed on the the outside of the barlow lens barrel represents to the change in EFL (or f/ratio) when the image plane is located at the place inside the barrel of the barlow where the field stop internal to the eyepiece is assumed to be when focused for visual use.   If instead we insert a nosepiece into the barrel of the barlow, maybe followed by a T-ring, and then behind that a camera unit whose focal plane is some distance inside the camera housing, and we refocus the telescope appropriately for the imager [*], the barlow lens optic is likely to be significantly further upstream from the imager's focal plane than it is from the eyepiece's internal field stop, so the nominal 2X or 3X number no longer applies.   Absent having a target for the optimum image scale we are trying to achieve, we might not realize how close-to or else far-from optimum our actual barlow-lens-projection setup might be.

 

So does anyone besides me see the potential pitfalls here of recommending just a target f/ratio based on pixel size ?   Since it's probably not necessary to operate at *exactly* that target f/ratio (or equivalently, at the target image scale in arcsec per pixel) maybe it doesn't matter a great deal, but here the OP is not obtaining the expected benefits from his or her nominally-2X barlow, so the first thing I would recommend is determining how much the image scale has changed between "with barlow" and "without".

 

Good luck,

 

       -- Jim

 

[*] In the case of SCTs or MCTs where changing focus is accomplished by moving the primary mirror, this also results in changing the EFL (or equivalently changing the f/ratio), which is yet another potential obstacle to achieving the desired image scale in arcsec/pixel -- especially if the amount of refocus needed is significant, and one just assumes the nominal f/ratio number still applies.
 


Edited by jkmccarthy, 15 September 2021 - 03:46 PM.

  • diver66 likes this

#7 Tulloch

Tulloch

    Gemini

  • *****
  • Posts: 3,354
  • Joined: 02 Mar 2019
  • Loc: Melbourne, Australia

Posted 15 September 2021 - 04:40 PM

Hi Jim, if you are mathematically inclined, this "proof" might appeal as to explain the "f/num = 5x pixel size" rule of thumb.

 

Focal ratio 5x pixel size proof.jpg

 

Andrew

 

 


  • matt_astro_tx likes this

#8 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted 15 September 2021 - 10:00 PM

You can definitely take any planetary image at capture resolution and work out the exact focal ratio as long as you know the pixel size and telescope aperture. 
 

206.3 x pixel size in microns / (aperture in mm x focal ratio) = object size in arcseconds / object size in pixels

 

solve for focal ratio. You don’t even need to know anything about the camera if all you want to do is figure out how close you were to critical sampling; solve for focal ratio / pixel size in microns. 



#9 BQ Octantis

BQ Octantis

    Fly Me to the Moon

  • *****
  • Posts: 5,202
  • Joined: 29 Apr 2017
  • Loc: Red Centre, Oz

Posted Yesterday, 02:39 AM

There is no such thing as critical sampling.

 

But there is high quality glass. High quality glass has minimal air-to-glass interfaces; all air-to-glass surfaces are fully multi-coated to maximize the signal throughput from the target. It brings the image into focus on a flat plane; the image on the sensor is crisp from edge to edge. Crappy glass brings the image into focus on a curved, spherical surface; projected onto a flat sensor, the image is mostly out of focus. It also has internal reflections and glare that lower the contrast of the image. So if you're putting crappy glass between your precision-made, perfectly collimated aperture and your exquisitely selected planetary camera, you're better off not magnifying.

 

For what it's worth, in planetary imaging there is a fairly straightforward relationship between SNR and numerous parameters of the setup and target. For a given target flux FT , sensor noise flux FN , exposure time τ, pixel size p, aperture diameter D, and effective focal length f, the SNR balance formula of a single frame is given by

 

Screen Shot 2021-09-16 at 4.13.54 PM.png

 

An important thing to note is the the flux from the target FT is different by target; for planetary, you also don't typically know the noise flux of the sensor FN . What you do typically know is the optimum gain and exposure time Gopt × τ for an "optimum" histogram. So instead of fluxes, the gain and exposure time become the normalizers for per-frame SNR balance. And of course, what we really care about is the SNR of the stack, which brings in the total number of stacked frames N. So the SNR balance formula becomes

 

Screen Shot 2021-09-16 at 6.36.53 PM.png

 

If you do Canon DSLR planetary capture like me, your shutter speed and pixel size are fixed (for my 600D/T3i, 1/30sec and 4.30µm, respectively), and gain is called ISO. So the SNR balance formula collapses into a very simple, convenient calculation:

 

Screen Shot 2021-09-16 at 6.36.59 PM.png

 

I've actually measured the ISOopt for everything from Mercury to Saturn in Live View for various magnifications over prime M through my f/15 Mak. Here is the exposure value (EV) chart with those results:

 

600D_T3I.jpg

 

Each vertical ISO line is a line of equal SNR, so the required stack size is the same. Note that because of its brightness, I cannot image Venus with Live View below M = 3.1 ( f /46.5) without it clipping. And because of the camera's ISO limit of 12800, I cannot reach optimal exposure on Saturn above M = 3.3 ( f /49.4). I find I get optimal results on Jupiter at M = 3.3 ( f /49.4 @ ISO 2260) because any larger and I can't easily keep the disk on the Live View screen—especially in gusty winds or seeing with a large random walk. While I get good luminosity results on Saturn at high magnifications in good seeing, in poor seeing I get some ugly color shifts starting at M=2.2 ( f /32.2 at ISO4800).

 

Lastly, in poor seeing you need all the SNR you can get. Less magnification buys you less integration time N×τ needed for the final stack, so you can get away with shorter frames and/or throw away more of the frames with crappy seeing.

 

Hope that helps.

 

BQ


Edited by BQ Octantis, Yesterday, 08:02 AM.

  • PKDfan and Borodog like this

#10 spaceju

spaceju

    Lift Off

  • -----
  • Posts: 11
  • Joined: 25 Mar 2020

Posted Yesterday, 03:23 AM

A 2x barlow for your setup should be fine (this is the combination I use).
But as mentionned : 2x (or less) the details, 4x (or more) the issues to be solved : good seeing conditions, good collimation, winjupos derotation eventually to compensate for the lower snr.
Also, I wonder if using a barlow on low planets makes the use of an atmospheric dispersion corrector even more critical ?

Edited by spaceju, Yesterday, 03:27 AM.


#11 jsnatale

jsnatale

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 13 Aug 2021

Posted Yesterday, 05:53 AM

Thank you all so much for all the information! I am really amazed at the great responses I got. I am new at this, so there is a lot to digest here. It is kind of like a short course in astrophotography! This forum has really been great- I have learned a lot. Again I appreciate your time and all the great info.


  • matt_astro_tx likes this

#12 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted Yesterday, 07:45 AM

BQ, what a fantastic post; you have corrected an error in my intuition; if you insert a 2X Barlow and change nothing else, you don’t need 4 times the frames to retain the same SNR; you need 16.

Edited by Borodog, Yesterday, 07:50 AM.


#13 BQ Octantis

BQ Octantis

    Fly Me to the Moon

  • *****
  • Posts: 5,202
  • Joined: 29 Apr 2017
  • Loc: Red Centre, Oz

Posted Yesterday, 08:16 AM

…if you insert a 2X Barlow and change nothing else, you don’t need 4 times the frames to retain the same SNR; you need 16.

Yep. This is borne out by my Moon and Mercury data compared to my Jupiter data. I've found 1024 frames yields reasonable results at 100% sensor scale on Jupiter at ~f/56. But I can get good results on the Moon and Mercury at ~f/56 at 100% scale with ~60-80 frames (so I shoot 100 for 80 or so). At that focal ratio Jupiter is 2 EV stops higher than both the Moon and Mercury, which puts the total number of frames for them at 1024/(22)2 = 1024/16 = 64.

 

Don't take my word for the EVs. I found a fantastic tool for AP exposures made back in 1998:

 

Starry_Room_Exposure_Calc_800.jpg

[Source]

 

Looks like photonics hasn't changed much in 20 years…

 

BQ

 

 


  • Borodog likes this

#14 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted Yesterday, 08:41 AM

You have got me totally rethinking how I shoot the Moon with the a6000. However, I feel like I'm sort of trapped by the ISO for minimum read noise, which I'm pretty confident is 800. My shutter speeds have to be pretty darn fast to shoot the Moon at f/5.3 and 800 ISO, and I definitely see the noise in the individual frames. I need to experiment with longer exposures at 400, 200, and 100 ISO and see exactly what is going on, though. If the camera were truly ISO independent then it shouldn't even matter; tG should be a constant. But it isn't, so I need to see what combination of t and G actually maximize the SNR per frame, but I think it has to be 800; that's what it means to have the minimum read noise.


Edited by Borodog, Yesterday, 08:48 AM.


#15 BQ Octantis

BQ Octantis

    Fly Me to the Moon

  • *****
  • Posts: 5,202
  • Joined: 29 Apr 2017
  • Loc: Red Centre, Oz

Posted Yesterday, 04:17 PM

You have got me totally rethinking how I shoot the Moon with the a6000. However, I feel like I'm sort of trapped by the ISO for minimum read noise, which I'm pretty confident is 800. My shutter speeds have to be pretty darn fast to shoot the Moon at f/5.3 and 800 ISO, and I definitely see the noise in the individual frames. I need to experiment with longer exposures at 400, 200, and 100 ISO and see exactly what is going on, though. If the camera were truly ISO independent then it shouldn't even matter; tG should be a constant. But it isn't, so I need to see what combination of t and G actually maximize the SNR per frame, but I think it has to be 800; that's what it means to have the minimum read noise.

 

Read noise is never minimized—if the camera is "ISO-less", the read noise is a linear scalar of the ISO. But the signal from the moon is so large, the read noise is swamped, regardless of whether the read noise has become linear. If you force yourself to go with the ISO where this happens, the shorter exposures actually make things worse by making the frames shot noise limited (so there is a minimum sqrt(N) just to fill in all the holes in the data). Just shoot at your lowest ISO for a "good" histogram at the M you want to use to compose the scene and the Ts you need for the seeing. If you're neurotic about the noise, use bias frames. The exposures are short enough that the bias and dark frames are pretty much the same.

 

BQ


Edited by BQ Octantis, Today, 02:05 AM.

  • Borodog likes this

#16 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted Yesterday, 04:27 PM

Read noise is never minimized—if the camera is "ISO-less", the read noise is a linear scalar of the ISO. But the signal from the moon is so large, the read noise is swamped, regardless of whether the read noise has become linear. If you force yourself to go with the ISO where this happens, the shorter exposures actually make things worse by making the frames shot noise limited (so there is a minimum sqrt(N) just to fill in all the holes in the data). Just shoot at your lowest ISO for a "good" histogram at the M you want to use to compose the scene and the Ts you need for the seeing. If you're neurotic about the noise, use bias frames. The exposures are short enough that the bias and dark frames are pretty much the same.

 

BQ

 

BQ

You just transformed the way I'm going to image the Moon with this camera. Thank you, very, very, very much. The good news is that it's only been ~10 months of subpar Moon images before I learned this.



#17 Tom Glenn

Tom Glenn

    Soyuz

  • -----
  • Posts: 3,636
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted Yesterday, 10:44 PM

But the signal from the moon is so large, the read noise is swamped, regardless of whether the read noise has become linear.

100% true.  The Moon always benefits from shooting at low gain/ISO because dynamic range is maximized and the SNR from individual frames is also highest.  The only issue that sometimes comes up with folks shooting DSLRs is that vibrations induced by either mirror slap or the mechanical shutter can require faster shutter speeds at long focal lengths.  But if you have an electronic shutter then this is not an issue, and unnecessarily short exposures are not helpful.  


  • Borodog likes this

#18 BQ Octantis

BQ Octantis

    Fly Me to the Moon

  • *****
  • Posts: 5,202
  • Joined: 29 Apr 2017
  • Loc: Red Centre, Oz

Posted Today, 02:08 AM

The only issue that sometimes comes up with folks shooting DSLRs is that vibrations induced by either mirror slap or the mechanical shutter can require faster shutter speeds at long focal lengths.  But if you have an electronic shutter then this is not an issue, and unnecessarily short exposures are not helpful.  

 

If I shoot from Live View and leave enough time between exposures, mirror slap is a non-issue. And with an intervalometer, I can go have dinner while it's shooting.

 

BQ
 



#19 BQ Octantis

BQ Octantis

    Fly Me to the Moon

  • *****
  • Posts: 5,202
  • Joined: 29 Apr 2017
  • Loc: Red Centre, Oz

Posted Today, 03:29 AM

I found a fantastic tool for AP exposures made back in 1998…

 

After evaluating the tool, I think it's my most consequential planetary find for 2021. It let me complete my "optimal exposure" EV chart for my 600D based on the "brightness" value alone:

 

Screen Shot 2021-09-17 at 5.44.57 PM.png

 

If ISO3200 is my highest desirable ISO for Live View, then I should shoot Uranus at prime. Unfortunately, at ISO12800 I can't fully expose Neptune at prime (f/15)—the maximum f/D for Live View exposure at ISO12800 is f/14.1. To hit ISO 3200, I'd need a 0.5× focal reducer.

 

BQ

 

P.S. Being Copyright © 1998 (in retrospect, perhaps at the peak of human civilization), surely this quite basic photometric information comes standard with every planetary camera, no?


Edited by BQ Octantis, Today, 05:56 AM.


#20 diver66

diver66

    Lift Off

  • -----
  • Posts: 12
  • Joined: 09 Jan 2021

Posted Today, 11:20 AM



Also, how much magnification you actually get from the barlow depends on the separation between the barlow and the camera focal plane.  When you compare the diameter of the planet (measured simply in units of imager pixels) with the barlow compared to without, is the planet's image 2X larger with the barlow, or significantly more ?   The ratio of measured diameters will tell you how much of an increase you're actually getting from using the barlow (alone, with no eyepiece, correct?) with your imager setup.

 

Elsewhere in these forums I've read that for capturing the planets, an effective focal length (EFL) of around 5-meters is thought by many to be optimum for use with typical imaging cameras ... so as a sanity check, you might multiply your telescope's focal length by the planet-diameter-increase-ratio you measure with the barlow, and see if your setup with barlow is delivering an EFL of around 5m.

 

Lastly, it's worth keeping in mind that with the barlow, the total light received from the planet will be spread out over many more pixels (how many more is given by the planet-diameter-increase-ratio squared ... 4x more pixels if the size of the planet is doubled using the barlow), so therefore the image is fainter with the barlow vs. without.  Are you using a longer exposure time (i.e., shooting with fewer frames per second of video) due to the fainter image ?  If frame rate is too low, atmospheric seeing and/or image motion could explain some of the loss of sharpness you're seeing.  If the frame rate is the same, then lower signal-to-noise ratio could be a factor (suggesting you might want to try capturing longer video sequences for doing your post-processing) [*].

 

Good luck,

 

       -- Jim

 

[*] EDIT ... or as Borodog's reply (posted while I was busy composing mine) suggests, increasing the gain provided this doesn't increase the noise even more than the signal.

 

Regarding Borodog's formula, I can shed light on where the "206.3" factor comes from:  the number of arcseconds in one radian is equal to 206,265 but since his formula uses mm for telescope focal length but microns for the pixel size (where 1 micron = 1/1000 mm), that factor of 1000 difference in the units reduces 206265 arcsec down to 206.3 ... so nothing mysterious there in the image scale calculation.

 

Regarding his "rule of thumb" for f/ratio = 5 x (pixel size in microns), note that because f/ratio = EFL / (Telescope Diameter), this is the same thing as saying the optimum EFL = 5 x (pixel size in microns) x (Telescope Diameter).   So if one then takes this EFL and plugs it back into the first formula, one finds the "rule of thumb" target image scale is:

 

Target image scale (arcsec/pixel) = 206.3 / (5 x Telescope Diameter in mm)

 

... or equivalently --

 

Target image scale = 0.2 arcsec/pixel x (206.3 mm / Telescope Diameter)

 

... so for an 8-inch telescope, the target image scale is ~ 0.2 arcsec/pixel, and coarser sampling can be targeted for smaller aperture telescopes while finer sampling should be targeted for larger aperture telescopes (to take maximum advantage of the moments of very best seeing).

 

Lastly, it appears the 5m EFL target I quoted was most appropriate for 10-inch aperture telescopes, whereas per the calculations above starting with Borodog's formulae, a 4m EFL target might be more appropriate for 8-inch aperture telescopes ... assuming imager pixel size of ~ 4 microns in both cases.  ((Larger pixel sizes require proportionally longer EFLs to maintain the same arcsec/pixel scale, so EFL ~ 5m might be optimum for an 8-inch aperture telescope in the case of ~ 5 micron pixels....))

 

Hope this helps.


Thank you very much for this post.  Some of this is starting to clear up for me. I do have a question.

  • My C9.25 has an aperture of 235mm and a focal length of 2350mm. 

  • My Celestron Neximage 10 camera has a teeny pixel of 1.67um.

  • Based on "The rule of thumb is that critical sampling occurs at a focal ratio of about 5 times the pixel size in microns" critical sampling would occur at a focal ratio of f/8.35.

  • My Current Image Scale at nominal focal length is (206.3 x 1.67) / 2350 = 0.1466 arcsec/pixel

  • My Target image Scale is 0.2 x (206.3 / 235) = 0.1755 arcsec/pixel

  • if I understand this correctly, get to my target scale I need to decrease the denominator (focal length) by 0.835, or increase the numerator (pixel size) by ~1.2.  

Does binning effectively increase pixel size or am I misunderstanding its use?  I guess I'm asking if I need to buy a focal reducer to make this work?

FWIW, having just learned all of this for the first time I completely agree that image scale makes more sense than f-ratio to understand the "why".



#21 Tulloch

Tulloch

    Gemini

  • *****
  • Posts: 3,354
  • Joined: 02 Mar 2019
  • Loc: Melbourne, Australia

Posted Today, 05:07 PM

Thank you very much for this post.  Some of this is starting to clear up for me. I do have a question.

  • My C9.25 has an aperture of 235mm and a focal length of 2350mm. 
  • My Celestron Neximage 10 camera has a teeny pixel of 1.67um.
  • Based on "The rule of thumb is that critical sampling occurs at a focal ratio of about 5 times the pixel size in microns" critical sampling would occur at a focal ratio of f/8.35.
  • My Current Image Scale at nominal focal length is (206.3 x 1.67) / 2350 = 0.1466 arcsec/pixel
  • My Target image Scale is 0.2 x (206.3 / 235) = 0.1755 arcsec/pixel
  • if I understand this correctly, get to my target scale I need to decrease the denominator (focal length) by 0.835, or increase the numerator (pixel size) by ~1.2.  

Does binning effectively increase pixel size or am I misunderstanding its use?  I guess I'm asking if I need to buy a focal reducer to make this work?

FWIW, having just learned all of this for the first time I completely agree that image scale makes more sense than f-ratio to understand the "why".

No need for a reducer, just image at your scope's native resolution (f/10). The "5x" rule is approximate anyway, you'll be fine where you are.

 

Interestingly, I also found out recently that when I insert my camera directly into the visual back on my C9.25" SCT, the native f/num is actually only f/8.78 instead of f/10 - so you might be closer than you think :)

https://www.cloudyni...ing/?p=11210977

 

Andrew



#22 Borodog

Borodog

    Vanguard

  • -----
  • Posts: 2,339
  • Joined: 26 Oct 2020

Posted Today, 05:30 PM

100% true.  The Moon always benefits from shooting at low gain/ISO because dynamic range is maximized and the SNR from individual frames is also highest.  The only issue that sometimes comes up with folks shooting DSLRs is that vibrations induced by either mirror slap or the mechanical shutter can require faster shutter speeds at long focal lengths.  But if you have an electronic shutter then this is not an issue, and unnecessarily short exposures are not helpful.  

This is a consideration I have run into before; the a6000 does not allow you to turn off the mechanical shutter, so vibration can indeed be an issue in burst mode for longer exposures. I need to test it in the low speed continuous mode and see at what shutter speed vibration might start to cause problems.




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