|

Now that you've captured some images with the HyperStar lens and Starlight
Xpress SXV-H9C camera, you can begin to process the images to get the most
detail possible out of them. While each object is different and may
require some fine tuning in Photoshop, the basic steps used in MaxIm DL will
almost always be the same. You can use the procedure described below for
almost any object. For fine tuning in Photoshop, see the Software
Instructions section on Image Processing
with Photoshop.
Once you've read through the detailed description below, print out the
HyperStar Image Processing Quick Reference to have
handy as you work on your images.
Note: This
tutorial uses a handy color conversion script in MaxIm DL
to convert all your images at once, saving a tremendous amount of time.
Much thanks for Klaus Lux for writing this script and allowing us to share it.
Download the appropriate version for your camera below. Note that some
SXVF cameras may still use the SXV conversion, so if you get purple nebulas, try
using the other script.
Place the script file in a convenient folder, such as Program Files >
Diffraction Limited, or wherever you'll be able to easily find it.
Running the script is described below.
Outline of Procedure
Open & Select Images
Begin by opening all images of a particular object. For this tutorial,
we'll use images of the Horsehead Nebula taken by Dail Terry with a HyperStar
C11. A total of 30 images at 30 seconds each were obtained in this
sequence, but the procedure is the same for any number of exposures.

Above: A single raw frame from the sequence and the final processed
image
Once the images are open, click through and glance at each file. You
want to eliminate any bad images before stacking the exposures. Look for
elongated star images from wind or tracking errors. With an unguided
telescope, expect to toss out a couple images for every 20 to 30 frames. One
of the great things about HyperStar is that if you lose an exposure, you're only
out 30 seconds! Close any bad images, leaving the good ones open.
Color Convert Images
This is where the color conversion script is invaluable. Normally,
converting each image requires selecting the file, clicking on RGB Convert, and
clicking OK. Three steps isn't bad for one or two images, but pretty tedious for thirty
shots. Why do ninety steps when you could do one!
To run the script, go to File > Run Script. Navigate to the folder
where you placed the script file and select it. Click OK. A black
Microsoft Script Host window appears while the images are being converted.
Allow a few seconds for the conversion to run. When finished, the files
have all been converted to RGB images. The script also adjusts the red,
green, and blue color weights to compensate for the spectral sensitivity of the
camera. This eliminates the need to color balance the images, although
there is a section below on fine tuning the color balance.
Align & Combine Images

Above: Close-up of a single exposure and of 30 frames stacked.
Note the significant noise reduction in the stacked frame.
Now for the real magic of MaxIm DL. Select Process > Combine from the main
menu. All the open images will be listed under Available Images.
Click Add All to select all the open images for stacking. (This is why it
is easiest to close the bad images before combining, in addition to speeding up
the computer processing time.)
In the Align window, select Auto - Star Matching under Align Mode and
Median
for Output. These are almost always the settings you will use, and they
will stay selected when you open this window in the future, so you only need to
select them once. Click OK and wait for the images to combine.
Stacking may take 1-2 minutes depending on the number of exposures and speed of
the computer. When finished the final stacked image will appear.
Tip: If your computer chokes and appears ready to light itself on fire
when trying to stack, say, 30 images, there is a workaround (other than buying
more RAM). Simply open only 10 images at a time and combine these.
Save the file, close the first set of raw frames and repeat the procedure with the
next set of ten and so on. Then, open the 3 combined images and stack
those. The end result will be almost identical, plus your computer won't
hate you.
Once the combined image is created, close all the other raw
frames. If MaxIm DL asks, choose No for saving changes. You
want to leave the raw files unchanged for future use.
Save Combined Image
It's a very good idea to save your stacked--but otherwise unprocessed--image
at this point. That way you can start again in the event that you do something you really regret later on in
the processing of the image. Go to File > Save. The File Format
should default to FITS, but if not, select this file type. Give the file a
name you can remember. For example, I would name the current file
Horsehead_med30. That way I know later that this file is a picture of the
Horsehead Nebula and is a median-combined stack of 30 exposures. Use
whatever scheme works for you, but I recommend coming up with a consistent
method so you know what you did at a later date.
Color Balance

Note: The Color
Conversion Script used above automatically color balances the images.
However, if you manually color convert your images, this section describes the
procedure. It also shows you what the script is doing. There is also
an optional procedure described to color compensate objects imaged low in the
sky.
This should be the first step after combining an image, before doing any
other processing. Color balance is necessary for two reasons.
Primarily, it is required because the camera does not have equal sensitivity in
each color. The SXV-H9C is most sensitive in red, slightly less sensitive
in green, and least sensitive in blue. To compensate for this, the Color
Balance window allows you to increase the amount of green and blue to compensate
for their lower sensitivity. For the H9C, the correct balance uses 100%
red, 110% green, and 140% blue. Go to Color > Color Balance on the main
menu. Enter 100, 110, 140 into the red, green, and blue scaling,
respectively. Click on the Auto button under Background Value. This
is an important step!

Note: The
following section is optional, but it is not done by the script, so if you
choose to do this process, it must be done manually.
A less critical adjustment, but recommended in certain situations, is to
compensate for atmospheric extinction. The lower an object is in the sky,
the less blue and green light will reach the telescope. This is why
sunsets are red. For objects above 45°
elevation, atmospheric extinction is negligible. Ideally you would image
all objects when they are high in the sky, but some targets never get that high
above the horizon. For these objects, color balancing for the atmospheric
effect can help the image considerably.
Compensating for this effect is
easy. Below is a table that shows the recommended color balance values for
different elevations. The table takes into account the necessary balance
for the H9C itself already (since the color conversion script already does that
step for you). If you know an object's elevation, just enter the values
below for the red, green, and blue scaling as described above.
|
Elevation |
Red |
Green |
Blue |
|
Above 45° |
100 |
100 |
100 |
|
40° |
100 |
104 |
108 |
|
35° |
100 |
106 |
111 |
|
30° |
100 |
108 |
115 |
|
25° |
100 |
111 |
121 |
|
20° |
100 |
116 |
131 |
|
15° |
100 |
125 |
150 |
Example: The images below show Omega Centauri
(photographed by Larry Moore) when the object was only 10° above the horizon (as high as
it gets from southern Arizona). The first image is the unbalanced color
image, the middle image is after balancing for the camera's color sensitivity,
and the right image is after balancing for both the camera and the atmospheric
extinction.

Adjusting the Screen Stretch
You may not be thrilled with how the image looks
at this point. In fact, look at the default display settings for our
Horsehead Nebula shot.

Not exactly what you might hope for. But
MaxIm DL is not displaying all the data contained within the image. This
is where the screen stretch comes into play. The screen stretch window
shows the histogram of the image, basically a scale from the darkest to the
brightest parts of the image. There are two triangles, one red and one
green, under the histogram. Learn to love these little triangles; they are
your friends, and they control how much detail your image will show. You
will use them a lot.

The concept is easy. The red triangle is
the black point slider. Everything to the left of this point is displayed
as pure black. The green triangle is the white point slider.
Everything to the right of this point is displayed as pure white. In other
words, only the brightness levels in the histogram between the two sliders will
be shown in the image. This is of critical importance. If any of the
faint detail is to the left of the red slider, it will not appear in the image.
Doing this defeats the purpose of taking and stacking a bunch of exposures to
try to get faint detail in the first place!
The most important thing when setting the screen
stretch is to be sure that no important data is cut out of the picture--what is
known as black point clipping. This is a common beginner's mistake.
In an attempt to make the background appear black, the red slider is moved to
the right, darkening the background but cutting off some of the faint data.
It is much better to leave the faint details intact and to darken the background
later using a more versatile technique such as Curves in Photoshop.
To prevent clipping the black point, run the red
slider almost up to--but not past--the left toe of the histogram curve, as shown
below.

Above: Clipping data, and not clipping
data. Always use the setting on the right.
As for the white point slider, the green
triangle, this is more subjective. Ideally you want this to be as far left
as possible, without washing out too much of the bright areas. This will
bring out the most faint detail without losing the highlights. For some
objects this may prove impossible, and for that there is a solution, discussed
in the following section. For now, we just want the image to look halfway
decent. The screen stretch will become more important when we save the
file after finishing the processing.

Above: The
Horsehead Nebula
after setting the screen stretch
Log Stretch
This step may not be necessary for all images, but most objects benefit from
this procedure. For example, our Horsehead Nebula image does not gain much
from this process, but many bright nebulas as well as most galaxies and globular
star clusters can be greatly enhanced by log stretching.
This function compresses the brightness of an object into a range that can be
displayed more easily by a computer monitor. Objects with both very bright
and very dim components (such as the core and spiral arms of a galaxy,
respectively) often cannot be displayed in all their detail without stretching.
Adjusting the screen stretch to show highlight details renders the faint detail
invisible, while enhancing the faint detail results in the bight areas being
washed out.

Above: The Whirlpool Galaxy screen stretched to show bright detail,
faint detail, and then displayed after log stretching
Log stretch compresses the histogram logarithmically to fit a larger range of
brightness within the range that can be displayed. To run this process, go
to Process > Stretch. Under Permanent Stretch Type, select Log. For
Input Range use Max Pixel. Select 16-bit under Output Range.

After running the log stretch you will have to readjust the screen stretch.

Above: Omega Centauri before and after log
stretching
Digital Development Processing
Digital development processing (DDP) is a more advanced processing step, but
for galaxy images especially, it can make a huge difference.

Above: The Whirlpool Galaxy before and after DDP
DDP works by taking the idea of histogram compression from log stretching
even further, while giving you more control over how the stretch operates.
It also sharpens the image. On galaxies, this filter is like magic, but it
is also useful on nebulas.
To use DDP, go to Filter > Digital Development. Under Filter Type
choose Low Pass. FFT Hardness should usually be set to Mild to begin, but
you may change this later depending on the initial results. Under DDP
Parameters, select Auto under Background. For Mid-Level, check the
Mouse
box, then move the cursor over the image and select a moderately bright, but not
too bright, part of the image. For a galaxy picture, the brightest part of
a spiral arm is perfect. Don't worry too much about it because you may end
up changing this value after the initial DDP run.

Click OK and get ready for an ugly picture. You will definitely have to
adjust the screen stretch after running DDP. Odds are things will still
not be ideal, even after screen stretching. You will most likely end up
clicking Undo and running DDP again with new settings until the picture looks
right. You have to experiment each time as every image is different.
Here are some things to look for after a DDP run and how to correct them:
-
If the screen stretch histogram is clipped at the black (left) end, lower
the Background value in the DDP window
-
If the bright parts of the image are not compressed as
much as you would like, decrease the Mid-Level value. (This value must
be greater than the Background value.)
-
If there are hideous black halos around the stars, select Custom under FFT
Hardness and increase the Cutoff % value.
Saving Images
There is more to this than you think. But it is very easy once you know
what to do. You just have to follow the steps below and everything will
save perfectly.
First, you need to adjust the screen stretch to make the image appear as you
want it saved. Typically after saving a file in MaxIm you will be opening
it in Photoshop for further processing. You will normally want to save
your file as a 16-bit TIFF. Newer versions of Photoshop can operate on
16-bit images. If you have Photoshop Elements, however, you will have to
save the file in 8-bit format. TIFF is the best file format choice because
it preserves all the data--it is not "lossy" like JPEG format. If you want
to save an image as a JPEG for posting online or emailing to a friend, save it
as a TIFF first in MaxIm then use Photoshop to save as a JPEG--you will have
more control.
You must adjust the screen stretch to avoid clipping and
faint you
want to retain. For the black point slider (red triangle), this means
keeping the slider to the left of the toe of the histogram. In fact, it is
best to go far enough left that the background is dark grey rather than black.
You can darken it later in Photoshop with more control, and you want to avoid
losing and faint detail at this stage.

For the white point slider, there is a tendency to bring it left far enough
to make the majority of the detail in the image appear reasonably bright.
But it is important that the bright spots in the image not be washed out.
Adjust the white point slider only as far left as necessary to keep the bright
parts from looking overexposed. If the middle-brightness portions of the
image are still a bit dim, this can be adjusted in Photoshop, but you can't get
back overexposed parts of the picture.
Once the screen stretch is set, you can save the file. Go to
File > Save
As. Be sure not to select Save since this will overwrite your previously
saved FITS file. In the window that appears, write in the desired
filename, and select TIFF Images under File Filter. Under Size Format
select 16-bit Int (or 8-bit Int if using Photoshop Elements).

Finally, click Stretch and in the window that comes up, select
Linear Only,
Screen Stretch, and 16-bit (or 8-bit for Elements).

This will keep the TIFF image looking like (and containing
all the data displayed on) the screen-stretched image in MaxIm DL.
More
Be sure to check out the Software Instructions section on
Image Processing with Photoshop for
tips on further image enhancement.
Print out the
HyperStar Image Processing Quick Reference as a handy guide to the steps on
this page.

Return to HyperStar Imaging
Page
|