Sunday, April 05, 2015
Shooting the Planets Part II: Imaging and Image Processing
I am not Don Parker. Or Chris Go. Or Damian Peach. I can,
however, even given my average equipment, skills, and the varying intensity with
which I pursue planetary imaging, produce pictures that would have caused
anybody’s draw to drop twenty years ago. Even on an average night, my
shots will reveal the basic configurations of Mars, Saturn, the Moon, and,
especially Jupiter. That is, even on less than optimum evenings, I can record
more detail than you will likely see visually, no matter how good your skills.
Now the caveat. As I underlined our last time out on this subject, what will limit you is the seeing. Below, you see the difference.
My average seeing is probably equivalent to a lot of people’s “good,” and it
will deliver the goods, if not as impressively as I’d sometimes like. What does
my average/below average seeing look like? The air is not steady. The image of
the planet onscreen is usually “boiling.”
Then we come to poor.
This is something that I experience in the winter months and even into early
spring some years. What’s poor? One look at a planet on the monitor will tell
the tale. It won’t be boiling, or just
boiling, it will be wavering.
Flapping like a flag. Believe me, you will recognize it when you see it.
Unfortunately, a lot of my northern brothers and sisters can experience this
during much of the year.
The good part? As putrid as the resulting image is, it will
still probably still reveal considerably more than an experienced observer will
easily detect through the eyepiece on such a night. How poor did Jupe look onscreen when I shot
the image on the right? He wasn't always even round
(or oval if you prefer); he was a blob more often than not.
Takeaway? Keep your eye out for the conditions that will
bring good seeing. Temperature inversions. Sticky, humid nights. But don’t
necessarily sit inside on evenings that you and the Clear Sky Clock agree will
be poor. If you are interested in monitoring the planet of your desire, not
merely taking a few pretty pictures, get outside with the scope and camera. Sometimes
seeing can improve suddenly just when you thought that was impossible. Finally,
if nothing else, shooting on nights that don’t look good will at least keep you in practice
with your scope and camera and more able to take advantage of good
nights when they do come.
First things first, you've got to get your camera on your
telescope. If you followed my advice in Part I, you have the camera plugged
into a Barlow and a flip mirror. The tiny chips of planet-cams and webcams make it harder to
center your target than you’d think, even with
the services of a flip mirror. Without one? You ain't got a prayer. Trying to line
up a planet in a high powered eyepiece, remove the eyepiece, and insert a
camera is a losing game. You’ll almost inevitably more the scope the tiny
amount required to put Jupiter out of the frame in the course of doing that. Or
your focuser will flex. Or, if you’re like me, you’ll kick a fraking tripod leg.
Before beginning, you need to have polar aligned your mount carefully (you can use an alt-az mount
for planetary work, but tracking will almost always be better with an EQ).
Especially if your camera supports and you plan to use FireCapture’s ROI (region of interest) feature, which crops the
frame to a size that barely fits the planet. That allows the camera to achieve
high frame rates, but a frame not much more than a minute of across when you
are on Jupiter means tracking needs to be spot on. You need a good polar
alignment and a well balanced scope. Even if your camera has a large frame and
you plan to use full frame, a planet will drift out quickly at high focal ratios if polar alignment
is not good.
 |
| FireCapture |
Once you have the scope properly aligned and the planet on
the cross hairs of the eyepiece in the flip mirror, it’s time to get it onscreen
and focus up. Connect the camera’s USB cable to the PC, light off FireCapture, and have a look. For this
discussion, we’ll assume you are using the new beta of FireCapture, version 2.4.05. There is really no reason not to, and
I find its menus easier to navigate than those of the earlier release. It also
does some things the old one doesn't, like automatically detecting your camera.
I also achieved a somewhat higher frame rate (in ROI mode) with it with my ZWO
ASI120MC than I did with the old version.
After FireCapture finds
your camera you should see an image on the preview screen immediately. If not?
If you don’t even see a bright blob? First off, check the frame size. You don’t
want ROI mode when focusing and framing. Select Max (resolution) at the top of
the control pane. You’ll see a bubble for 16-bit in the same area. Make sure it
is unchecked. You don’t need it and it just slows things down.
Still no Jupiter? Make sure you haven't, as I almost always
do, forgotten to flip the flip mirror up to send images to the camera. Chances are,
however, that you’ll find that even with a flipper, aiming ain't a piece of
cake. On my last evening out, I had a hard time getting Jupe on the display.
This was one of the first times I used the new FireCapture. Had I set it up wrong?
Nope. The reason was simpler. I needed to adjust my flip
mirror. Over the years, it’s gotten to the point where what’s in the center of
the crosshairs is no longer in the cam. It does NOT take much misalignment of the mirror to do that. I found I had to offset Jupe between
the center and the field edge for it to be onscreen. Like most flippers, my
mirror is adjustable and I need to tune it up (someday). If you’re still having
trouble despite testing the aim of the flipper (the Moon is good for doing that
if it's out), check your exposure. Set the gain slider to about 75%, and
increase the exposure time of the camera and see if anything
appears.
Next up is focus. Remember how I told you you really want
motorized focusing for planetary work? The reason will be obvious now. If you
have moto-focus, focusing will be quick and you won’t mind refocusing as the evening
proceeds. If you ain't got it, focusing will be a pain, even if your scope is
close enough to the monitor that you can focus while watching the image. Touching
the telescope's focus knob will make the planet move unless you've got a big
honking A-P mount or something like that.
How about exposure? You need that right to be able to focus
successfully. Set the gain to 60 – 70% (too low gain can produce odd artifacts in
planetary images), and adjust the exposure until the histogram meter on the
bottom of the preview display peaks at about 60%. Another way of saying that is
to say you should adjust exposure time till the
image looks just right and then shorten
it until the planet looks slightly too dim (never overexpose).
If you have an SCT, you’ll be happy you've set your camera’s
resolution to Max for focusing/framing. Even with an SCT with minimal focus shift, focusing will
still likely cause Jupiter—that’s what I imaged on this night, and I’ll assume
that’s what you’re after right now, too—to move all the way across the frame.
My old Ultima 8, Celeste, has an average amount of focus shift, so while
Jupiter moved from one side of the field to the other, it never quite went off screen, luckily.
When focus is as good as you can get it—under even poor seeing
you should see cloud bands—center the planet up. Use your hand control to put
the planet as close to the middle of the frame as possible. Then, if you are
going to use ROI, switch to that and do final centering. It’s now time to fire
off a sequence.
How long should these movie (.avi) sequences be? Even a
30-second ROI of Jupiter at 50fps plus will produce well over 100 megabytes of data, so be mindful of your hard
drive size. I usually find 30-45 seconds of .avi will result in enough good
frames for me to work with. When you are ready, click the “record” (blue arrow) button on the capture
section of the control pane. Your .avi recording will begin. You can either set
it for a max limit (typically 30-seconds for me) in the Capture section, or
just watch the time displayed and hit the stop button when you've got as much
as you want.
How many .avi captures should you do? As many as it takes. I
generally try to do more rather than less, spacing them out across the duration
of the evening in hopes of catching good seeing. You should be able to tell
when the seeing improves dramatically by looking at the preview image and
hitting “Record” immediately, but it’s not always obvious when it improves slightly,
so take plenty of .avis.
When you've got what you think you need, throw the big
switch, head inside, pour yourself a glass of whatever tickles you, and turn on
the tube or spend some time with your loved one (mucho better). The one thing
you don’t want to do with your
pictures the night you take ‘em is process ‘em. Or even look at them. Trust me,
they won’t look good no matter how good they really are, and no matter how bad
they really are, they will still look much better in the morning.
 |
| Debayer |
OK, it’s morning. Birds are chirping, the Sun is shining,
and you are ready to begin processing your images. The first step, assuming you
use a one-shot color camera like I usually do, is what I nostalgically call, “developing
the pictures.” It is really nothing like putting negatives in soup, but, like
developing film, it is the preliminary step you must take before you do
anything else. The real name for this first step is “debayering.”
Color cameras work by exposing pixels through a matrix of red,
green, and blue filters on the chip. Look at the raw images from your color
camera and you won’t see any color at all. Just a black and white picture that
looks strangely pixilated; it will have what appears to be a grid pattern—and that is exactly what is
going on. To get color, software combines a matrix, a grid, of color filtered
pixels. Three filtered pixels, red, green, and blue, make a final color pixel. The
result is normal looking color images. All one shot color cameras of all kinds
work basically the same way. Naturally, if you are using a monochrome camera, you
don't have to debayer.
Actually, you could skip this step even with a color camera.
FireCapture and some other image
acquisition programs will debayer on the fly. They will convert raw images to
color before they are saved on the hard drive. That's not usually a good
idea, however. What we want more than anything else when we are recording planetary .avi movies is lots of frames, a high frame rate. Debayering in real
time will inevitably slow the camera down.
So, exactly how do you "develop" pictures? With FireCapture, you use a little utility
that comes with the program, Debayer.
You’ll use it frequently, so find this application in the directory where FireCapture resides and put a shortcut
to it on your desktop. The rest is simple: start it, click “Open AVIs,” and
select your image files. Normally, you’ll select multiple image files (shift-click
in the file window) and let the utility debayer all your sequences at once. You
can choose the type of debayering process, but I find the default, “bilinear,”
works fine. Click “Start” and the program will begin its work.
When all your files have been processed, you’ll find debayered copies of them in the same
directory as the undebayered ones. The new .avis will have the original file
names, but “bilinear” will have been appended to them (assuming that was the process
you used). If hard drive space is an issue, you can now delete the non-debayered files.
 |
| AutoStakkert |
Time to stack. What exactly does that mean? What you
have on your hard drive are movies,
.avi movies, videos, but what you want are high resolution stills. The way to
get those is with a program that will select the best frames of your video and
stack them. It’s easy to understand why it is good to select the best frames,
those taken in the best seeing, but why stack them?
Stacking does one thing for you: it reduces noise. Not only does that mean the
finished still image looks smoother, it means it is easier and more practical
to apply sharpening tools. Try to sharpen a noisy frame and you sharpen the
noise too, making it more prominent and making the image look worse rather than
better.
If you, as most planetary imagers are, running
Windows on a PC or on a Mac by means of emulator software like Bootcamp, you have two major choices of
stacking program today, the venerable Registax
and a newcomer, AutoStakkert. As I
said in Part 1, Registax still does a
good job, but I must admit AutoStakkert
is at least somewhat better. It is, like Registax,
freeware and is no more difficult to use, so there is no reason not to
use AutoStakkert.
Also like Registax,
AutoStakkert
has many settings and adjustments that can be used to improve your results once you've gained some experience with the program. The nice thing, though, is that
both programs can produce impressive results with a few simple settings and/or just
leaving (most) controls at their defaults. Which is what we are going to do
this time out—keep it simple.
Execute AutoStakkert
(download the latest Beta, 2.3.0.21; it is more than ready for primetime). When
it comes up, you’ll find two windows on your desktop as shown above. Step one
is, natch, to press the Open button in the left window and open your .avi file.
Make sure you choose the debayered version, of course.
When your file is loaded, you’ll find one of its frames
displayed in the preview window (right). It won’t look like much; it will be
both fuzzy and noisy. Don’t worry; we are going to cure that now. Immediately
below the Open button is the Stabilization control area. Here, select either
Planet or Surface. The latter is used if you are processing an .avi of the
lunar or Solar surface. Since I presume you've, like me, been shooting Jupe,
select Planet.
 |
| Registax |
Now, click the Analyse button. AutoStakkert will think for a while, but only for about 15-seconds
unless you have a very large video file. When it is done, direct your attention
to the right pane, the image pane. You need to do two things here. The first is
to specify how many alignment points (to keep frames properly aligned during
stacking) you want on your image. I find that for a high resolution planetary
image, “50” is good. So, tick “50” in the “Auto AP” area of the frame.
Next, push the “Place APs in Grid” button just below. Alignment
points (red points inside boxes) will be automatically positioned. You can
place APs manually or delete or add points after placing them automatically.
Some folks tell you to delete alignment points near the edge of the disk, but I
haven't found this to make a difference.
Almost done. Go back to the left window, to the Stack
Options area on the upper right, and tick “TIF.” That will cause your stacked
image to be saved as a .tif file, which is normally the best option. And that
is it. Just push the “Stack” button at the bottom of the window and let the
program do its thing. In about one minute, depending on the horsepower of your
PC and the size of the .avi file, AutoStakkert
will finish and will have written a .tif file of Jupiter to a new subfolder
labeled “AS” in the folder in which your original .avis were located.
If you were to open your stacked .tif file with Photoshop (or your image processing
program of choice) now, you’d be somewhat impressed, but not blown away. The
picture would look a lot smoother, and noticeably less noisy, but it would
still be awfully soft. The way to fix that is with Registax 6. While AutoStakkert
does a fine job of stacking, it lacks Registax's
famous “Wavelet” sharpening tools.
Run Registax 6,
and, when it comes up, click Select (upper left) to select and load the .tif
file that AutoStakkert produced. When
your picture loads, Registax may ask
if it should stretch intensity levels. Say “yes.” Now, choose the Wavelet tab.
What will appear on the left is a series of sliders and some selection bubbles.
First choice is Wavelets Scheme, Dyadic or Linear. The former seems to give
each Wavelet slider more power and range, and is what I choose most of the
time.
Below “Scheme” is Wavelet Filter. You can choose Default or
Gaussian. In most instances, I find “Default” works best with the planets while
Gaussian is a little better on the Moon. If you choose the latter, you’ll find
you can specify Denoise and Sharpen percentages for each slider. I generally
set Denoise to about 10 and sharpen to 100 for each slider, but you may want to
experiment with that.
 |
| Final tweaks in Photoshop... |
Now you can began tweaking the sliders till your image looks
good. In Dyadic mode, you will usually only have to adjust the top three.
Remember, exercise a light touch; you
want detail and sharpness, but you also want Jupiter to look normal. Think of
the pictures you've seen by masters Damian
Peach and Chris Go. You want
your pix to look as much like that as possible. Oh, the top slider works on the
smallest detail and each succeeding one affects larger and larger features.
When you are satisfied, click the Do All button and then
Save Image. While Registax has plenty
of other good processing tools to adjust the histogram of your image and do
other things, I find I am more comfortable using an outboard program like Adobe Photoshop or LightRoom. Just about any image processing program will work, since
all you’ll likely need to do with the Registax
output is tweak brightness and contrast and do some cropping.
So, what do you think? Yes, I know that while you have
striven to make your shot look like something done by Mr. Peach, it’s unlikely
it will—mine sure don't. Still, I think you’ll have to admit that what you
wound up with ain't bad. And your images will just get better, muchachos. Better seeing,
better focus, most of all, more experience doing this will all help. The only secret to planetary imaging these days, other than good seeing, is PRACTICE, and I am betting your first pix have you excited enough to ensure you do plenty of that.
Comments:
<< Home
This blog actually is a nice tutorial on on solar system imaging, concise and interesting. thanks Rod. Howard
# posted by 
: 2:36 AM
: 2:36 AM
Hi Rod,
Especially as a newbie to astroimaging, I REALLY appreciate these posts! Just one quick question. I've been shopping around for an illuminated reticle eyepiece, and most of the options I've found are at the extreme ends of the price/quality spectrum. Any advice on where to find a good one?
Especially as a newbie to astroimaging, I REALLY appreciate these posts! Just one quick question. I've been shopping around for an illuminated reticle eyepiece, and most of the options I've found are at the extreme ends of the price/quality spectrum. Any advice on where to find a good one?
The safe bet is just get the cheapest one you can find. You probably want it to be illuminated, though, and that drives up the price. Good compromise: http://www.telescope.com/Accessories/Telescope-Eyepieces/Orion-125mm-Illuminated-Reticle-Plossl-Telescope-Eyepiece/pc/-1/c/3/sc/47/p/8450.uts
Great no nonsense primer for planetary imaging. Thanks! I Read this before going out for first try with ZWO 120MM-S and same software and my C8. Spot on...tough to hit that little chip w/o flip mirror, didn't have one. I added a RACI Finder recently with illuminated reticle and tightly aligned it with main scope. Helped immensely by putting the planet in the cross hair square. Landed Venus, Saturn, and Jupiter with very little nudging. I ordered a flip mirror afterwards. Turns out it probably gives me the spacing I need for back focus as I had to slide the Barlow out to get focus. Spot on about the JMI focuser. Mine worked fine w/o adding crayford. Turns out, mirror flip was minimal and the image stayed within the window.even at ROI. Setting recommendations saved hours.
pgandy
Post a Comment
pgandy
<< Home
