| Planetary Photo Techniques (page 2, Updated 3/24/2008) (Skynyx Cameras, LRGB imaging & Lucam Recorder) |
I'll offer a few observations concerning the Lumenera Skynyx cameras, controlled by Lucam
Recorder, for LRGB (and IR/UV) planetary imaging. These comments are most useful for
new Skynyx/Lucam Recorder users, or those that are considering the acquisition of these
products:
The Seeing is more important than any of the equipment you may buy!
This is one of the astrophotography regimes were those of us operating with limited funds
can compete on an equal footing with those deep pocketed astrophotographers (that
generally focus their attention, and $50-100K setups, on long exposure subjects).
Why choose the Lumenera?
1) Go with the Lumenera 2-0M monochrome camera (7.4um pixel) w/ a filter wheel for
highest resolution (and highest frame rates), if you can create the longer focal lengths (FL)
required for critical sampling in blue. Consider the other, smaller pixel versions (2-1 or 2-2
with 4.7/4.4um pixels), if you are not able to reasonably create the longer FL. The larger
pixels in the 2-0 will offer the potential for a higher signal to noise, and greater dynamic
range, useful in bringing out minimum noise on planetary detail for Mars and Jupiter, and
catching planetary bodies and moons in the same exposure. The 2-0 also offers higher
frame rates w/o ROI or subframing. If you use ROI or subframing to increase the frame rate
with the 2-1, or 2-2, you'll end up with no greater FOV than with the 2-0, but still have the
lower quality pixels. The increase in FOV offered by the 2-1 and 2-2, in full frame, and much
slower frame rates, is about 1/3 for the 2-1, and 1/2 for the 2-2. Remember, it's not just the
number of pixels, it's the imaging area, closely related to the size of the chip that dictates
the FOV for a given scope. You will also pay a lot more for the 2-1 or 2-2 for LITTLE GAINED,
and a bit lost. It's your money, you decide how to spend it (see page 4 for more discussion).
The DMK cams, while 1/2 the price of the Lumenera, didn't offer some of the features that
interested me (I understand that their firmware was updated in 2007, and they now offer
some of the preferrable features, though still fall short on their ROI features' highest frame
rate capability). You get what you pay for; it seems that one must pay X2 the price in order
to get 10% performance improvement on astronomy related products (I wonder about this
sometimes).
What focal length should you use?
How much FL do you need with the 2-0M (7.4um pixel)? You need to create a focal ratio of
f/38 to critically sample the airy disc in blue (400nm) light. This means a FL of 9500mm for my
250mm Mewlon. I need an effective X3.2 multiplication. This will oversample longer
wavelengths. Is critical sampling the airy disc at blue enough, or too much? Some people
suggest oversampling the FWHM of the seeing disc, not the airy disc, which results in a
lower focal ratio, shorter focal length than airy disc sampling. In case the pixel size is
different than the Skynyx 2-0, just scale to get the required focal ratio. My earlier Atik had
5.6 um pix, so it would require a focal ratio of 29, from this, the Mewlon would be using a FL
of 7250mm, a multiplication of X2.4 was required. There is something to be said about a
system without ANY refractive optics, like barlows. If you have an f/25 scope, your best
options are to go with a camera using a smaller (lower quality) pixel, or find a good X1.5 to
X1.7 multiplier. If you have an f/4 scope, save it for deep space imaging, where the speed is
a real advantage. Buy something in the f/10 to f/20 range for planetary work.
In the (high frequency) electronics arena, we typically sample 15-30% higher than nyquist
due to imperfect sampling apertures, filters, and a number of other reasons. I know what
you're saying, ...my audio system is based on samples taken at 192Ksps, much higher than
the nyquist rate for audio frequencies. There is a BIG difference between highly over
sampled sigma-delta conversion, as used in todays audio systems, and simultaneous, or
parallel sampling of a optical (spatial) scene. I'm confident that I'll loose NOTHING by
running f/40 or f/45. I'm also more confident that I'll have the lum layer properly focused.
The exposures will be shorter, and the resulting frame rates will be higher at f/40 vs f/50 or
more, hence I'll have a better chance of reducing seeing induced blur. If I want a larger
presentation, I use a good resampling routine, such as Genuine Fractals, or even the
bicubic (sharpener) in Photoshop to upscale the image by 40% (presenting an image at an
effective f/56). If your seeing is 9, or better, on the Pickering scale (ie. perfect), go ahead
and crank up the capture focal length (f/50 or more). Otherwise, forget the big BLURRY
exposures, and set your focal length (and resulting focal ratio) at a reasonable value, you'll
be much happier with the results.
How long should you record frames?
There is another item to consider when deciding on the focal length you will use. The
planets rotate, a feature on the surface moves. If you just start running an image
sequence, and go long enough, you'll find that the object will be located in a different
position on later frames in your sequence. You need to allocate time for LRGB sequences
such that the total time for all is less than the time for the object on the planet surface to
move a noticeable amount. How long do we have before an object on the planet's surface
moves 1" as seen from your telescope? Mars allows at least 1160sec, Saturn allows 590sec,
and Jupiter allows 230sec (assumes closest planetary approach). This is for 1", not the
angular extent that a pixel sees when you are using f/40 on a 10 inch scope. For this case,
we are using 10000mm focal length, the image scale (which is what really counts) for the
Skynyx 2-0 is 0.15"/pixel (slightly oversampled in blue light, even greater oversampling in
red). If we want to be critical, one would want to keep smearing at, or below one pixel
(0.15"). This means that we have only 180sec for the entire LRGB sequence for Mars, 90sec
for Saturn, and 34sec for Jupiter. No problem for Mars, but for Saturn, and especially
Jupiter, this is a problem. The only solution is to let the smearing go to 1 1/2 pixel, which
allows us to take the LRGB sequence over a period of 135sec for Saturn and 51sec for
Jupiter. No problem on Saturn, but this is not much time for many frames in a complete
LRGB set on Jupiter, therefore, you may want to consider using the LR(sG)B color layering.
Instead of four filters to run through, you need only three. The color is not as good, but the
increased number of frames for those colors collected, decreases the noise, smoothing
out the image.
If you want examples of both good and bad focal length selections, take a look at the yahoo
groups that focus on planetary images and/or ALPO websites. You'll see a few nice sharp
images, some upsampled post capture, taken at focal lengths that the seeing was able to
support. You'll also see a lot of oversized, blurry photos. Some were captured at excessive
focal lengths, and then made worse by upsampling. I might also add that it may help you to
take a look at several of the planetary imaging sites to see what results are being
delivered with various systems, operated in various ways. You may start to see some
patterns, some methods just yield better results.
If you are willing to accept a lower resolution, but want to simplify your life, go with the
color version Skynyx cams. The problem is that the bayer filter matrix, over the CCD, means
that the RGB capture sampling (alternating GRGR/BGBG) is not representative of the post
interpolation effective sampling, and trying to select the appropriate FL for critical
sampling will be questionable. The Bayer filtering /interpolation results in a scene
dependent resolution, or effective pixel size. I guess that you will choose a FL based on
the pixel size, then multiple that by 20-30%% (looks like you may need f/50 for this case), to
be sure, or just accept the potential undersampling. You may also want to consider the
(potentially) lower effective RGB resolution on these cameras. Their resolution,
interpolated from the raw RGB camera data, is scene dependent. Discussions with a couple
camera manufacturers indicate that the RGB resolution on these one-shot cameras CAN be
as much as 20-25% lower than what would be achieved with a monochrome cam through
filters. If you're looking for "ease of use" go with the color cam. If you want "best possible
results" go with the monochrome camera and filter set.
An alternative to the use of filters is to do as I did with the usb1 cams. Use a monochrome
for lum, and a RGB for chroma capture (both equal size pixels). You'll need a turret so that
the cams can be brought into the optical path quickly. Take a look at the previous page for
a more complete description of two cam use. The two cam approach works well.
Recorder, for LRGB (and IR/UV) planetary imaging. These comments are most useful for
new Skynyx/Lucam Recorder users, or those that are considering the acquisition of these
products:
The Seeing is more important than any of the equipment you may buy!
This is one of the astrophotography regimes were those of us operating with limited funds
can compete on an equal footing with those deep pocketed astrophotographers (that
generally focus their attention, and $50-100K setups, on long exposure subjects).
Why choose the Lumenera?
1) Go with the Lumenera 2-0M monochrome camera (7.4um pixel) w/ a filter wheel for
highest resolution (and highest frame rates), if you can create the longer focal lengths (FL)
required for critical sampling in blue. Consider the other, smaller pixel versions (2-1 or 2-2
with 4.7/4.4um pixels), if you are not able to reasonably create the longer FL. The larger
pixels in the 2-0 will offer the potential for a higher signal to noise, and greater dynamic
range, useful in bringing out minimum noise on planetary detail for Mars and Jupiter, and
catching planetary bodies and moons in the same exposure. The 2-0 also offers higher
frame rates w/o ROI or subframing. If you use ROI or subframing to increase the frame rate
with the 2-1, or 2-2, you'll end up with no greater FOV than with the 2-0, but still have the
lower quality pixels. The increase in FOV offered by the 2-1 and 2-2, in full frame, and much
slower frame rates, is about 1/3 for the 2-1, and 1/2 for the 2-2. Remember, it's not just the
number of pixels, it's the imaging area, closely related to the size of the chip that dictates
the FOV for a given scope. You will also pay a lot more for the 2-1 or 2-2 for LITTLE GAINED,
and a bit lost. It's your money, you decide how to spend it (see page 4 for more discussion).
The DMK cams, while 1/2 the price of the Lumenera, didn't offer some of the features that
interested me (I understand that their firmware was updated in 2007, and they now offer
some of the preferrable features, though still fall short on their ROI features' highest frame
rate capability). You get what you pay for; it seems that one must pay X2 the price in order
to get 10% performance improvement on astronomy related products (I wonder about this
sometimes).
What focal length should you use?
How much FL do you need with the 2-0M (7.4um pixel)? You need to create a focal ratio of
f/38 to critically sample the airy disc in blue (400nm) light. This means a FL of 9500mm for my
250mm Mewlon. I need an effective X3.2 multiplication. This will oversample longer
wavelengths. Is critical sampling the airy disc at blue enough, or too much? Some people
suggest oversampling the FWHM of the seeing disc, not the airy disc, which results in a
lower focal ratio, shorter focal length than airy disc sampling. In case the pixel size is
different than the Skynyx 2-0, just scale to get the required focal ratio. My earlier Atik had
5.6 um pix, so it would require a focal ratio of 29, from this, the Mewlon would be using a FL
of 7250mm, a multiplication of X2.4 was required. There is something to be said about a
system without ANY refractive optics, like barlows. If you have an f/25 scope, your best
options are to go with a camera using a smaller (lower quality) pixel, or find a good X1.5 to
X1.7 multiplier. If you have an f/4 scope, save it for deep space imaging, where the speed is
a real advantage. Buy something in the f/10 to f/20 range for planetary work.
In the (high frequency) electronics arena, we typically sample 15-30% higher than nyquist
due to imperfect sampling apertures, filters, and a number of other reasons. I know what
you're saying, ...my audio system is based on samples taken at 192Ksps, much higher than
the nyquist rate for audio frequencies. There is a BIG difference between highly over
sampled sigma-delta conversion, as used in todays audio systems, and simultaneous, or
parallel sampling of a optical (spatial) scene. I'm confident that I'll loose NOTHING by
running f/40 or f/45. I'm also more confident that I'll have the lum layer properly focused.
The exposures will be shorter, and the resulting frame rates will be higher at f/40 vs f/50 or
more, hence I'll have a better chance of reducing seeing induced blur. If I want a larger
presentation, I use a good resampling routine, such as Genuine Fractals, or even the
bicubic (sharpener) in Photoshop to upscale the image by 40% (presenting an image at an
effective f/56). If your seeing is 9, or better, on the Pickering scale (ie. perfect), go ahead
and crank up the capture focal length (f/50 or more). Otherwise, forget the big BLURRY
exposures, and set your focal length (and resulting focal ratio) at a reasonable value, you'll
be much happier with the results.
How long should you record frames?
There is another item to consider when deciding on the focal length you will use. The
planets rotate, a feature on the surface moves. If you just start running an image
sequence, and go long enough, you'll find that the object will be located in a different
position on later frames in your sequence. You need to allocate time for LRGB sequences
such that the total time for all is less than the time for the object on the planet surface to
move a noticeable amount. How long do we have before an object on the planet's surface
moves 1" as seen from your telescope? Mars allows at least 1160sec, Saturn allows 590sec,
and Jupiter allows 230sec (assumes closest planetary approach). This is for 1", not the
angular extent that a pixel sees when you are using f/40 on a 10 inch scope. For this case,
we are using 10000mm focal length, the image scale (which is what really counts) for the
Skynyx 2-0 is 0.15"/pixel (slightly oversampled in blue light, even greater oversampling in
red). If we want to be critical, one would want to keep smearing at, or below one pixel
(0.15"). This means that we have only 180sec for the entire LRGB sequence for Mars, 90sec
for Saturn, and 34sec for Jupiter. No problem for Mars, but for Saturn, and especially
Jupiter, this is a problem. The only solution is to let the smearing go to 1 1/2 pixel, which
allows us to take the LRGB sequence over a period of 135sec for Saturn and 51sec for
Jupiter. No problem on Saturn, but this is not much time for many frames in a complete
LRGB set on Jupiter, therefore, you may want to consider using the LR(sG)B color layering.
Instead of four filters to run through, you need only three. The color is not as good, but the
increased number of frames for those colors collected, decreases the noise, smoothing
out the image.
If you want examples of both good and bad focal length selections, take a look at the yahoo
groups that focus on planetary images and/or ALPO websites. You'll see a few nice sharp
images, some upsampled post capture, taken at focal lengths that the seeing was able to
support. You'll also see a lot of oversized, blurry photos. Some were captured at excessive
focal lengths, and then made worse by upsampling. I might also add that it may help you to
take a look at several of the planetary imaging sites to see what results are being
delivered with various systems, operated in various ways. You may start to see some
patterns, some methods just yield better results.
If you are willing to accept a lower resolution, but want to simplify your life, go with the
color version Skynyx cams. The problem is that the bayer filter matrix, over the CCD, means
that the RGB capture sampling (alternating GRGR/BGBG) is not representative of the post
interpolation effective sampling, and trying to select the appropriate FL for critical
sampling will be questionable. The Bayer filtering /interpolation results in a scene
dependent resolution, or effective pixel size. I guess that you will choose a FL based on
the pixel size, then multiple that by 20-30%% (looks like you may need f/50 for this case), to
be sure, or just accept the potential undersampling. You may also want to consider the
(potentially) lower effective RGB resolution on these cameras. Their resolution,
interpolated from the raw RGB camera data, is scene dependent. Discussions with a couple
camera manufacturers indicate that the RGB resolution on these one-shot cameras CAN be
as much as 20-25% lower than what would be achieved with a monochrome cam through
filters. If you're looking for "ease of use" go with the color cam. If you want "best possible
results" go with the monochrome camera and filter set.
An alternative to the use of filters is to do as I did with the usb1 cams. Use a monochrome
for lum, and a RGB for chroma capture (both equal size pixels). You'll need a turret so that
the cams can be brought into the optical path quickly. Take a look at the previous page for
a more complete description of two cam use. The two cam approach works well.
How to improve your frame rate
2) The Skynyx 2-0M monochrome camera w/ LR is fast. Don't run a usb2 cam through an
external hub (unless you are willing to take a lower frame rate to your PC in the event of an
underperforming hub), hook it straight into one of the usb2 ports on the PC. If you must
use a hub, look at the reviews, and get a known good one (some aren't). If you use the
direct RAM capture on 8b SER files (AVI too), you get whatever frame rate you ask for, ie
60fps full frame (640X480), 112fps using a 340X240 ROI. The 60fps is nice for Jupiter lums
and full FOV Mars lums, the 112fps is useable for the clear filter with Mars on a 320 X 240
ROI, (it would probably be good on lunar shots too). The disadvantage in using RAM writes
is the additional time required to move the image sequence from RAM to HD (version 2.0 of
LR is suppose to remove this delay by allowing the data move to occur while the next
sequence is being written to RAM, we'll see if it works properly when reviews of that
version become available, that's the only new feature in LR 2.0 that I'm interested in, and
for me, it's not worth the upgrade price).
Hard drive capture is dependent on your usb2, bus, and HD speed. If you periodically
defrag your HD, don't try to run several programs at the time you are using LR to capture
images, you'll see 40-45fps (full FOV) written to my laptop 5400rpm HD. I expect one of the
7500rpm HDs, with a large cache, and the latest internal bus would allow 50-60fps HD
writes. I typically see an actual of 85-95fps when running the 320X240 ROI on Mars lum
shots at 112fps. I don't have any complaints, that's still a lot of frames real fast (can your
DMK do that, they couldn't in 2006 when I was looking for a new cam).
3) You won't be able to take advantage of the blazing camera speed on Saturn, RGB
exposures on Jupiter, IR exposures for Mars, or when using a UV filter on Venus (the
response of the cam at deep blue is poor, adding to the already low response of the UV
filter). You can count on 15-30fps (full FOV) under these filter conditions though, much
faster than you can run the usb1 cams (w/o compression artifacts). Any PC should be able
to handle these frame rates. If you go with a sub frame, or ROI, say 320 X 240 on small
subjects like Mars (7" as of August 07), you can get a good clear exposure at 112fps, and
write 85-90% of that (uncompressed) frame rate to HD.
4)The Skynyx is a bit less noisy than the Atiks, you don't need darks, even at large fractions
of a second exposures. You do need to take darks on a multi-second exposure, as may be
the case on Neptune, when trying to also catch its moon, Triton. You do need to take flats
(to get rid of dust donuts), and biases to properly use the flats on the lights. Registax
allows convenient use of biases (considered to be darks) and flats. Use proper lighting for
the flat. I use the same light box with this scope/cam that I use on my long exposure setup.
5) LR's filter wheel interface is nice, but there isn't a focuser interface yet (at least not as of
version 2.0). This means that you need to focus on a star, or planetary moon through the
lum filter, before you start a sequence, return to the planet, start your sequence with lum,
then when the LR sequencer changes filters and pauses, refocus. This is repeated for
each color filter. This is a pain. With a focuser interface, a fixed offset (re the clear filter)
would be programmed for each color filter. As LR calls up each new filter, the appropriate
focuser offset would be commanded. Once the focuser interface is implemented in LR, I'll
add an encoder equipped external focuser to the Mewlon and upgrade from LR 1.x to the
new version (unfortunately, the Tak scope designers didn't think ahead, and add an
encoder to the fine internal motorized secondary focuser).
Is refocusing between filters necessary?
Is refocusing between filters important? It depends on how critical you want the focus.
You're running an expensive scope, camera, etc., using an image scale of a fraction of an
arcsec per pixel, so that you get the most out of your system, then you let the focusing go
were it will??? Point at a star under conditions of good seeing, carefully focus through your
lum filter, note the focuser position, now repeat for your other filters. Was there a (small)
change in the focuser position, color vs lum? If so, plan on refocusing between filters, if
not, you're fortunate. Don't just take some other hobbyist's word that you can "let it slide",
check it out yourself.
6) A comment concerning low elevation subjects (like Jupiter is in 2007-08 for those of us
in the US), use a dispersion corrector when your subject is at or below 45 degree elevation.
Differential refraction of wavelengths across the visible spectrum, for a subject at 45
degree elevation, results in a 0.6" shift between red and blue components. This means that
a broadband filtered image (like a UV/IR blocked lum), will have components from the
red/blue spectrum smeared in excess of the diffraction limit of an 8 inch scope. Even
images taken with individual color filters will experience smearing, that approaches the
diffraction limit of a 14 inch scope. A Risley prism can be used to correct this effect, and
allow one to capture sharp, unsmeared broadband images. The prism makes a difference
on broadband (UV/IR blocked lum) shots at 45 degrees, and a BIG difference at 30 degrees.
It is even noticeable on RGB filtered exposures for subjects at 30 degrees. Contact
Adirondack for their 1 1/4 inch dispersion corrector, then Precise Parts for a 2 inch holder if
you use a 2 inch diameter imaging train. I use the Baader Quick Change couplings to allow
independent rotation of the corrector within the imaging train. Differential refractive
effects for subjects at elevations of 30 degrees, and above, is NOT an excuse for smeared,
blurry photos (probably exacerbated by running excessive focal lengths, and improper
focusing, and....what about collimation).
2) The Skynyx 2-0M monochrome camera w/ LR is fast. Don't run a usb2 cam through an
external hub (unless you are willing to take a lower frame rate to your PC in the event of an
underperforming hub), hook it straight into one of the usb2 ports on the PC. If you must
use a hub, look at the reviews, and get a known good one (some aren't). If you use the
direct RAM capture on 8b SER files (AVI too), you get whatever frame rate you ask for, ie
60fps full frame (640X480), 112fps using a 340X240 ROI. The 60fps is nice for Jupiter lums
and full FOV Mars lums, the 112fps is useable for the clear filter with Mars on a 320 X 240
ROI, (it would probably be good on lunar shots too). The disadvantage in using RAM writes
is the additional time required to move the image sequence from RAM to HD (version 2.0 of
LR is suppose to remove this delay by allowing the data move to occur while the next
sequence is being written to RAM, we'll see if it works properly when reviews of that
version become available, that's the only new feature in LR 2.0 that I'm interested in, and
for me, it's not worth the upgrade price).
Hard drive capture is dependent on your usb2, bus, and HD speed. If you periodically
defrag your HD, don't try to run several programs at the time you are using LR to capture
images, you'll see 40-45fps (full FOV) written to my laptop 5400rpm HD. I expect one of the
7500rpm HDs, with a large cache, and the latest internal bus would allow 50-60fps HD
writes. I typically see an actual of 85-95fps when running the 320X240 ROI on Mars lum
shots at 112fps. I don't have any complaints, that's still a lot of frames real fast (can your
DMK do that, they couldn't in 2006 when I was looking for a new cam).
3) You won't be able to take advantage of the blazing camera speed on Saturn, RGB
exposures on Jupiter, IR exposures for Mars, or when using a UV filter on Venus (the
response of the cam at deep blue is poor, adding to the already low response of the UV
filter). You can count on 15-30fps (full FOV) under these filter conditions though, much
faster than you can run the usb1 cams (w/o compression artifacts). Any PC should be able
to handle these frame rates. If you go with a sub frame, or ROI, say 320 X 240 on small
subjects like Mars (7" as of August 07), you can get a good clear exposure at 112fps, and
write 85-90% of that (uncompressed) frame rate to HD.
4)The Skynyx is a bit less noisy than the Atiks, you don't need darks, even at large fractions
of a second exposures. You do need to take darks on a multi-second exposure, as may be
the case on Neptune, when trying to also catch its moon, Triton. You do need to take flats
(to get rid of dust donuts), and biases to properly use the flats on the lights. Registax
allows convenient use of biases (considered to be darks) and flats. Use proper lighting for
the flat. I use the same light box with this scope/cam that I use on my long exposure setup.
5) LR's filter wheel interface is nice, but there isn't a focuser interface yet (at least not as of
version 2.0). This means that you need to focus on a star, or planetary moon through the
lum filter, before you start a sequence, return to the planet, start your sequence with lum,
then when the LR sequencer changes filters and pauses, refocus. This is repeated for
each color filter. This is a pain. With a focuser interface, a fixed offset (re the clear filter)
would be programmed for each color filter. As LR calls up each new filter, the appropriate
focuser offset would be commanded. Once the focuser interface is implemented in LR, I'll
add an encoder equipped external focuser to the Mewlon and upgrade from LR 1.x to the
new version (unfortunately, the Tak scope designers didn't think ahead, and add an
encoder to the fine internal motorized secondary focuser).
Is refocusing between filters necessary?
Is refocusing between filters important? It depends on how critical you want the focus.
You're running an expensive scope, camera, etc., using an image scale of a fraction of an
arcsec per pixel, so that you get the most out of your system, then you let the focusing go
were it will??? Point at a star under conditions of good seeing, carefully focus through your
lum filter, note the focuser position, now repeat for your other filters. Was there a (small)
change in the focuser position, color vs lum? If so, plan on refocusing between filters, if
not, you're fortunate. Don't just take some other hobbyist's word that you can "let it slide",
check it out yourself.
6) A comment concerning low elevation subjects (like Jupiter is in 2007-08 for those of us
in the US), use a dispersion corrector when your subject is at or below 45 degree elevation.
Differential refraction of wavelengths across the visible spectrum, for a subject at 45
degree elevation, results in a 0.6" shift between red and blue components. This means that
a broadband filtered image (like a UV/IR blocked lum), will have components from the
red/blue spectrum smeared in excess of the diffraction limit of an 8 inch scope. Even
images taken with individual color filters will experience smearing, that approaches the
diffraction limit of a 14 inch scope. A Risley prism can be used to correct this effect, and
allow one to capture sharp, unsmeared broadband images. The prism makes a difference
on broadband (UV/IR blocked lum) shots at 45 degrees, and a BIG difference at 30 degrees.
It is even noticeable on RGB filtered exposures for subjects at 30 degrees. Contact
Adirondack for their 1 1/4 inch dispersion corrector, then Precise Parts for a 2 inch holder if
you use a 2 inch diameter imaging train. I use the Baader Quick Change couplings to allow
independent rotation of the corrector within the imaging train. Differential refractive
effects for subjects at elevations of 30 degrees, and above, is NOT an excuse for smeared,
blurry photos (probably exacerbated by running excessive focal lengths, and improper
focusing, and....what about collimation).
Talking about collimation, use a star in the part of the sky
that you will be working, especially is you have a large/long
imaging train, 2" tubes bend, thanks to gravity. You want
the bend as it will be when you are imaging. Once you have
the collimation close, as measured in/out of focus, go to
focus (under good seeing) to finish up. The photo to the
right shows the Mewlon at focus (10000mm eFL), it's not
quite there, but close.
that you will be working, especially is you have a large/long
imaging train, 2" tubes bend, thanks to gravity. You want
the bend as it will be when you are imaging. Once you have
the collimation close, as measured in/out of focus, go to
focus (under good seeing) to finish up. The photo to the
right shows the Mewlon at focus (10000mm eFL), it's not
quite there, but close.
