| Planetary Photo Techniques (page 3, Updated 11/29/2007) This page continues with the Skynyx Camera & Lucam Recorder |
Getting Good Planetary Colors
7) Proper color balance of an image created by a monochrome camera, like the Skynyx
2-0M, and RGB filters, like the Astronomik type 2c, requires consideration of the differing
relative sensitivity of the camera through the three filters. The graph below shows the
relative response of the 2-0M (using constant gain) through the Astronomik RGB filters.
Notice that the blue & red light
response is slightly lower than
that of the green. This means
that you can either run a slightly
longer exposure for blue & red,
or set the same exposure
times, and turn up the gain for
blue & red. As the gain increases,
so does the noise. As the
exposure time increases, so
does smearing. I have opted to
use constant gain settings, and
increase the blue & red exposure
times to balance the RGB.
Example Exposures Settings
8) Typical Skynyx 2-0M settings, using LR to control capture (HD writes), for a low elevation
Jupiter, when using Astronomik (type 2) LRGB filters on a Mewlon 250 at f/38:
filter weight exposure/gain limit @ fps
R 1.06 (45ms/10.0) 250 @ 15fps
G 1.0 (42ms/10.0) 250 @ 15fps
B 1.52 (64ms/10.0) 250 @ 15fps
clr --- (16.7ms/12.0) 900 @ 60fps (actual 40fps,300 frames dropped)
The exposure times are selected in order to compensate for spectral variations in camera
sensitivity and filter attenuation. One could select constant exposure times and vary gain,
but the blue channel will end up w/ a high gain, and resulting noise. Also note that the
relative red weight/exposure will increase re the green/blue when Jupiter moves higher in
elevation, in a couple years.
Note the high number of dropped frames when I'm asking for 60fps to HD (full frames).
When the HD has been recently defragged, the actual rate goes to 45-50fps, ie, fewer
dropped frames. I've used RAM writes on lum only shots, when 60fps is requested, 60fps is
delivered, but, as noted above, this version of LR requires that the RAM captured frames
be written to HD before a new capture can begin. The overall result (capture to HD) is no
better than just accepting a slower HD write when running a LRGB sequence. The
upcoming LR version 2.0 is suppose to take care of this (as yet TBD).
A few caveats. The exposures/gains given on page 2 result in a relatively high individual
frame and stack SNR, when used under 4/5 transparency w/ the noted scope. The LR
preview meter shows 65-70% FS on each. The resulting stack requires little, if any, levels
adjustment in PS. One could get by with a slightly lower preview meter reading, and
resulting exposure, then use levels in PS to compensate. I'd try a 5-10% reduction in
exposure time on the RGB filters (may slightly reduce object smearing), and if the preview
meter still shows above 1/2 FS, I'd also reduce the gain 5-10% on each (to attempt to
maintain SNR). Under the same conditions, the clear filter gain may be reduced by 5-10% (to
reduce noise).Under better transparency, like 5/5, and with a 1/2 FS meter is accepted,
decrease the clear filter exposure by 5-10% and gain by 10-15%. The RGB exposures could
probably be reduced by 20-25%, making 30fps possible on the R & G filters, but the B will
still be confined to 15fps max. In this case, up the frame limit on the RGB filters to 300 (less
time spent on R & G, more spent on B, remember that Jupiter rotates rapidly, your entire
LRGB image period must be complete in 100-120 seconds). You may even want to omit the
G filter sequence, use PS to synthesize green from red & blue (50%/50%). This is routinely
called LR(sG)B imaging, and works rather well on Jupiter, and even better on Mars (I've not
seen it done on Saturn). Remember that a lower gain will result in lower noise, and less
problems when trying to pull out fine details.
9) A couple comments concerning Mars in 2007. I've used the following 2-0M exposures for
capture at a 320 X 240 ROI when the planet was 7" (and covered in a dust storm):
filter weight exposure/gain limit @ fps
R 1.0 (15ms/12.0) 1200 @ 56fps (actual 49fps)
G 1.1 (17ms/12.0) 1200 @ 56fps
B 1.8 (28ms/12.0) 700 @ 28fps (actual 24fps)
lum --- (9ms/10.0) 1200 @ 112fps (actual 95fps)
A note for Mars during November 2007. The planet is now over 14", and I'm still using
similar exposures, gains, and frame rates. The surface brightness of Mars hasn't changed
much. In case you're interested in the IR (742nm Astronomik) filter. I collect 1200 frames
with about the same settings as with the blue filter. In order to account for the IR imaging
time, I collect LRGB frames as 1200, 600, 600, 600. I then combine the lum and IR images, to
act as the lum layer of a photo. I sometimes combine the red and IR to act as the R layer in
the LRGB image.
I know that some amateur astrophotographers think that anything other than pure LRGB
planetary images are simply wrong. Each practitioner is entitled to their own opinion of how
a subject should appear, and how an image is assembled. Those that claim their way is THE
ONLY PROPER WAY, may have arrived at a point in their hobby, were a break is in order.
If you want enhanced contrast of albedo features, blend the IR with the L and/or R layers in
Photoshop (25-35% opacity IR blended as a "normal" layer w/ lum, adjust IR histogram prior
to a 30-40% opacity blend as a "normal" layer w/ red). Be sure that you have properly
exposed, or adjusted the histogram of, the green and blue layers before the final RGB
layering. You want to be sure that the beautiful light blue haze and North Pole clouds are
fully represented along w/ the moderate/high contrast albedo features. And by all means,
do your contrast enhancement/sharpening on the lum layer prior to layering w/ the RGB.
Use wavelets in your stacking program and/or regularized deconvolution (as in PixInsight)
combined w/ a light use of high pass filtering in PS prior to adding the "color" layer (copied
from the prealigned RGB).
10) A note concerning Venus (at less than 50% illuminated, 25 to 35"). I've shot a few LRGB
& UV sequences of this planet at a couple different gain settings. It appears that the higher
gain works well with several hundred, or a thousand frames, but in case you only want to
record a small number of frames, use the lower gain. The LRGB filters can run at high frame
rate, but the UV (due to filter density and lower cam sensitivity) must run at a lower frame
rate (like 15fps). The values I found to work w/ the 2-0M & Astronomik filters are:
filter exposure/gain alt. exposure/gain
R (1.2ms/12) (8ms/5)
G (1ms/12) (7.6ms/5)
B (1.6ms/12) (12ms/5)
lum (0.2ms/10) (2ms/5)
UV (Schuler) (67ms/14) --------
Saturn has a much lower surface brightness than any of the above (~6.4mag/sq arcsec vs
5.3mag/sq arcsec for Jupiter and 4.1mag/sq arcsec for Mars at opposition). I run the lum at
15fps, the RGB at 7.5fps, we're almost back to the old usb1 cameras here. I find the
following exposures to work on it (you'll typically need to over expose the planet to get the
even dimmer moons, at least beyond Titan):
filter exposure/gain
lum 55-60ms/12 (adjust for 1/2+ on exposure meter)
R 133ms/13
G 120ms/12
B 133ms/14
In case you're interested in a photo of Uranus (@ f/38, 10"), which is really bland in amateur
scopes. Try the following (shot at least 500, preferably 1000 frames of each to reduce the
noise):
filter exposure/gain
lum 0.25s/18
R 0.6s/20
G 0.5s/20
B 0.6s/20
Neptune (@ f/38, 10") requires even longer exposures. For the planetary ball try:
filter exposure/gain
R 1.55s/18
G 1.4s/18
B 1.55s/18
If you are interested in catching a little of Triton (very little), without totally saturating the
planetary ball, try (15-20 darks are necessary for exposures of this length):
filter exposure/gain
R 5.8s/18
G 5.2s/18
B 5.8s/18
A decent registration of Triton will take exposures on the order of 10-12 seconds, the
planet will be totally saturated, so you either take a low S/N moon, or a saturated planet. For
a monochrome photo, try using your NIR filter, which will keep the planet (mostly blue) from
saturating while getting some of the (reddish) moon. The exposure gets long for this filter
so plan on darks.
7) Proper color balance of an image created by a monochrome camera, like the Skynyx
2-0M, and RGB filters, like the Astronomik type 2c, requires consideration of the differing
relative sensitivity of the camera through the three filters. The graph below shows the
relative response of the 2-0M (using constant gain) through the Astronomik RGB filters.
Notice that the blue & red light
response is slightly lower than
that of the green. This means
that you can either run a slightly
longer exposure for blue & red,
or set the same exposure
times, and turn up the gain for
blue & red. As the gain increases,
so does the noise. As the
exposure time increases, so
does smearing. I have opted to
use constant gain settings, and
increase the blue & red exposure
times to balance the RGB.
Example Exposures Settings
8) Typical Skynyx 2-0M settings, using LR to control capture (HD writes), for a low elevation
Jupiter, when using Astronomik (type 2) LRGB filters on a Mewlon 250 at f/38:
filter weight exposure/gain limit @ fps
R 1.06 (45ms/10.0) 250 @ 15fps
G 1.0 (42ms/10.0) 250 @ 15fps
B 1.52 (64ms/10.0) 250 @ 15fps
clr --- (16.7ms/12.0) 900 @ 60fps (actual 40fps,300 frames dropped)
The exposure times are selected in order to compensate for spectral variations in camera
sensitivity and filter attenuation. One could select constant exposure times and vary gain,
but the blue channel will end up w/ a high gain, and resulting noise. Also note that the
relative red weight/exposure will increase re the green/blue when Jupiter moves higher in
elevation, in a couple years.
Note the high number of dropped frames when I'm asking for 60fps to HD (full frames).
When the HD has been recently defragged, the actual rate goes to 45-50fps, ie, fewer
dropped frames. I've used RAM writes on lum only shots, when 60fps is requested, 60fps is
delivered, but, as noted above, this version of LR requires that the RAM captured frames
be written to HD before a new capture can begin. The overall result (capture to HD) is no
better than just accepting a slower HD write when running a LRGB sequence. The
upcoming LR version 2.0 is suppose to take care of this (as yet TBD).
A few caveats. The exposures/gains given on page 2 result in a relatively high individual
frame and stack SNR, when used under 4/5 transparency w/ the noted scope. The LR
preview meter shows 65-70% FS on each. The resulting stack requires little, if any, levels
adjustment in PS. One could get by with a slightly lower preview meter reading, and
resulting exposure, then use levels in PS to compensate. I'd try a 5-10% reduction in
exposure time on the RGB filters (may slightly reduce object smearing), and if the preview
meter still shows above 1/2 FS, I'd also reduce the gain 5-10% on each (to attempt to
maintain SNR). Under the same conditions, the clear filter gain may be reduced by 5-10% (to
reduce noise).Under better transparency, like 5/5, and with a 1/2 FS meter is accepted,
decrease the clear filter exposure by 5-10% and gain by 10-15%. The RGB exposures could
probably be reduced by 20-25%, making 30fps possible on the R & G filters, but the B will
still be confined to 15fps max. In this case, up the frame limit on the RGB filters to 300 (less
time spent on R & G, more spent on B, remember that Jupiter rotates rapidly, your entire
LRGB image period must be complete in 100-120 seconds). You may even want to omit the
G filter sequence, use PS to synthesize green from red & blue (50%/50%). This is routinely
called LR(sG)B imaging, and works rather well on Jupiter, and even better on Mars (I've not
seen it done on Saturn). Remember that a lower gain will result in lower noise, and less
problems when trying to pull out fine details.
9) A couple comments concerning Mars in 2007. I've used the following 2-0M exposures for
capture at a 320 X 240 ROI when the planet was 7" (and covered in a dust storm):
filter weight exposure/gain limit @ fps
R 1.0 (15ms/12.0) 1200 @ 56fps (actual 49fps)
G 1.1 (17ms/12.0) 1200 @ 56fps
B 1.8 (28ms/12.0) 700 @ 28fps (actual 24fps)
lum --- (9ms/10.0) 1200 @ 112fps (actual 95fps)
A note for Mars during November 2007. The planet is now over 14", and I'm still using
similar exposures, gains, and frame rates. The surface brightness of Mars hasn't changed
much. In case you're interested in the IR (742nm Astronomik) filter. I collect 1200 frames
with about the same settings as with the blue filter. In order to account for the IR imaging
time, I collect LRGB frames as 1200, 600, 600, 600. I then combine the lum and IR images, to
act as the lum layer of a photo. I sometimes combine the red and IR to act as the R layer in
the LRGB image.
I know that some amateur astrophotographers think that anything other than pure LRGB
planetary images are simply wrong. Each practitioner is entitled to their own opinion of how
a subject should appear, and how an image is assembled. Those that claim their way is THE
ONLY PROPER WAY, may have arrived at a point in their hobby, were a break is in order.
If you want enhanced contrast of albedo features, blend the IR with the L and/or R layers in
Photoshop (25-35% opacity IR blended as a "normal" layer w/ lum, adjust IR histogram prior
to a 30-40% opacity blend as a "normal" layer w/ red). Be sure that you have properly
exposed, or adjusted the histogram of, the green and blue layers before the final RGB
layering. You want to be sure that the beautiful light blue haze and North Pole clouds are
fully represented along w/ the moderate/high contrast albedo features. And by all means,
do your contrast enhancement/sharpening on the lum layer prior to layering w/ the RGB.
Use wavelets in your stacking program and/or regularized deconvolution (as in PixInsight)
combined w/ a light use of high pass filtering in PS prior to adding the "color" layer (copied
from the prealigned RGB).
10) A note concerning Venus (at less than 50% illuminated, 25 to 35"). I've shot a few LRGB
& UV sequences of this planet at a couple different gain settings. It appears that the higher
gain works well with several hundred, or a thousand frames, but in case you only want to
record a small number of frames, use the lower gain. The LRGB filters can run at high frame
rate, but the UV (due to filter density and lower cam sensitivity) must run at a lower frame
rate (like 15fps). The values I found to work w/ the 2-0M & Astronomik filters are:
filter exposure/gain alt. exposure/gain
R (1.2ms/12) (8ms/5)
G (1ms/12) (7.6ms/5)
B (1.6ms/12) (12ms/5)
lum (0.2ms/10) (2ms/5)
UV (Schuler) (67ms/14) --------
Saturn has a much lower surface brightness than any of the above (~6.4mag/sq arcsec vs
5.3mag/sq arcsec for Jupiter and 4.1mag/sq arcsec for Mars at opposition). I run the lum at
15fps, the RGB at 7.5fps, we're almost back to the old usb1 cameras here. I find the
following exposures to work on it (you'll typically need to over expose the planet to get the
even dimmer moons, at least beyond Titan):
filter exposure/gain
lum 55-60ms/12 (adjust for 1/2+ on exposure meter)
R 133ms/13
G 120ms/12
B 133ms/14
In case you're interested in a photo of Uranus (@ f/38, 10"), which is really bland in amateur
scopes. Try the following (shot at least 500, preferably 1000 frames of each to reduce the
noise):
filter exposure/gain
lum 0.25s/18
R 0.6s/20
G 0.5s/20
B 0.6s/20
Neptune (@ f/38, 10") requires even longer exposures. For the planetary ball try:
filter exposure/gain
R 1.55s/18
G 1.4s/18
B 1.55s/18
If you are interested in catching a little of Triton (very little), without totally saturating the
planetary ball, try (15-20 darks are necessary for exposures of this length):
filter exposure/gain
R 5.8s/18
G 5.2s/18
B 5.8s/18
A decent registration of Triton will take exposures on the order of 10-12 seconds, the
planet will be totally saturated, so you either take a low S/N moon, or a saturated planet. For
a monochrome photo, try using your NIR filter, which will keep the planet (mostly blue) from
saturating while getting some of the (reddish) moon. The exposure gets long for this filter
so plan on darks.
