SETUP | CCD SOFTWARE | READING

Capturing a Comet
Richard A. Bennion
Ewell Observatory

M any people have inquired upon the acquisition and processing workflow used to capture the image of
Comet C/2001 Q4 NEAT which appeared in Sky and Telescope (October 2004). The following is my
technique.

Please note that I make certain references to astrophotography equipment and software that are
specific to my observatory. Some of my technique relies on this equipment and may not be applicable to C/2001 Q4 Neat
everyone. In those cases you will have to use your imagination to arrive to some similar process that Click to V iew Image

achieves the same effect.

STEP #1 - Equipment Setup:

To acquire the data for this image I used a Takahashi FSQ-106N, ST10XM E and CFW8A filter wheel with SBIG LRGB filters all controlled
by a RoboFocus focusing system. The mount used was an Astro-Physics AP900GTO (circa 1998) with upgraded Rev D firmware.

The following software was used to control the above equipment:

TheSky Telescope Control Software connected to the AP900GTO via a serial cable using the built in AP driver.
CCDSoft Camera Control Software connected to the ST10XM E and filter wheel via a USB cable.
FocusM ax connected to the RoboFocus focuser system via a serial cable.
PulseGuide 1.26 by Ray Gralak connected to the AP900GTO via a serial cable connected to the second serial port on the mount.
CCDAutoPilot used to control TheSky, CCDSoft and FocusM ax to create my imaging sequences.

Before attempting to capturing a comet I would suggest the following checklist of tasks as all image acquisition will be taken unguided
and requires special care to equipment setup:

1. 1. Make sure that you a very well polar aligned. Any error in alignment will have a drastic effect on the quality of tracking of
the image due to DEC drift. In order to to achieve good results in unguided imaging you must avoid DEC drift from mis-polar
alignment.
2. 2. Correct your Periodic Error. - In order to reduce (or eliminate) any RA oscillation while tracking, I strongly suggest that you
use the periodic error correction capabilities of your mounts firmware. Some mounts have lower periodic error than others
and therefore will have better success in unguided imaging. I suggest using a product like PEM Pro by Ray Gralak to program
your mount's periodic error correction firmware.
3. 3. Make sure your setup is well balanced: Unguided data acquisition is very sensitive to unbalanced imaging setups. Both the
RA and DEC axis must not allow for any stress on the system. One special note of caution is making sure you cables don't tug on
the system or get caught up on some piece of equipment. I suggest the careful use of cable ties and strategic placement of
cables to avoid any unnecessary drag on the system.

Step #2 - Find the Comet:

The next step is to get your imaging system in position and focus to acquire the data for the comet. This consist of a few easy steps:

1. Start TheSky software, but do not connect to your telescope just yet.
2. Load the positional information about the comet you want to capture into TheSky's database. The following is the procedure
for the TheSky6 Professional Edition:

Select DATA > COM ETS AND M INOR PLANETS from the pull down menu.
From the COM ETS tab, click on the OBSERVABLE button on the lower left of the dialog box.
Select the Comet name you want to image from the list. For instance; C/2004 Q2 (M achholz).
Click Ok.
Now the Comet you selected should be in the scrollable list box on the COM ETS tab.
Scroll down to the bottom of the list box, find the Comet and select it to make it visible in TheSky.

3. Now that the Comet is available for pointing and tracking in TheSky, you can use the Find dialog box to locate the Comet in the
sky.
 4. There are many other great features of TheSky6 that allow you to plot the Comets path over time with date labels and much
more. I encourage you to fiddle with these advanced functions.
5. Once you have the comet selected and the information dialog box available, slew your scope to the Comet.
6. M ake sure the Comet is centered in your camera's field of view.

Now we are ready to move to the next step of programming our mount to track the comet.

Step #3 - Program you Mount:

This is probably the most challenging part of the acquisition process and most time consuming.

A mount tracks objects in the sky by moving at sidereal rate counter acting the rotation of the earth. Deep sky objects don't change
position over time (well not over one night of imaging anyways). Comet's are unlike deep sky targets as they are moving through the sky
at a different rate than sidereal. Imagine trying to image a meteor as it hurls across the night sky but in slow motion.

So the next crucial step is programming (or directing) the mount to move at a speed other than sidereal. Fortunately comets move
very slowly from our earthly perspective, so this is something that can be accomplished with amateur mounts.

PulseGuide by Ray Gralak working with an Astro-Physics mount makes this task possible. The concept is as the mount is moving at its
normal sidereal rate, PulseGuide sends carefully timed 'pulses' to the mount to speed up or slow down the RA and DEC speed of the
motors. This allows you, for example, to slow down RA by .02 arc-seconds per second and speed up DEC by .06 arc-seconds per
second. If you know the RA and DEC speeds of the comet you are attempting to image, you can program PulseGuide to get your mount
to track perfectly on the comet.

I offer this step-by-step tutorial on how to accomplish this feat. Please note, it is advised that you read the PulseGuide manual first to
gain a better understanding of the following steps:

1. Your telescope should currently be pointing at the Comet with the nucleus in the center of your camera's field of view.
2. Take a test shot of the comet to confirm that you can see the comet with your camera control software.
3. Once you have confirmed that the comet is in the center of the field of view, its time to start PulseGuide.
4. Connect PulseGuide to your AP mount. If you do not have a second serial cable connected to your mount, you will need to
disconnect from TheSky and use the same COM port for PulseGuide.
5. M ake sure that your PEM is enabled on your mount to reduce RA errors in guiding.
6. In the TheSky6, open the Comet's Object Information dialog box and select the GENERAL Tab. You should see something like
this:

7. You will notice that TheSky6 provides additional information regarding Comets in the above dialog box. Of particular note is the
RA rate and Dec rate in arc-seconds per seconds. We will use this information as a starting point for programming PulseGuide.
8. Now that we know how the Comet is moving we can enter the RA rate and DEC rate of the Comet on the PULSE GUIDER Tab of
PulseGuide as shown in the following:
 9. Now, start Pulse Guiding by clicking on the ON button. Now your mount, via the serial port, is tracking the comet.
10. We need now to make some adjustments based on our specific location, atmospheric refraction and other factors that might
affect perfect tracking of the Comet. You can perform this operation with most any camera control program, but for this
example, I am going to use CCDSoftV5.

Take a one second unbinned image of the comet in CCDSoft using the clear filter on the Focus Tools tab. M ake sure
you are taking a light frame with Autodark enabled.
Using the historgram tool, modify the background and range so that the comets nucleus looks like a small point of
light similar to a star.
Using your mouse cursor, select a ~100 x 100 box around the nucleus.
Go back to the Setup tab and disable Auto-Contrast.
With the Subframe checkbox enabled and the Continuous check box enabled, start taking 1 second images of the
Comet nucleus.
Hit the '6' key to enable CCDSoft's crosshairs and zoom in on your sub-frame to about 400%.
Adjust the histogram to get a good set of pixels to watch.
You should end up with a screen similar to the diagram below:

12. Now that we have our 'Real-time Comet Tracking Box' enabled we can fine tune PulseGuide to track the comet as best we can.
13. Watch the Comet for a few minutes and note if and what direction it moves. If it does not move after a few minutes, we are all
 done calibrating the speed of the mount. If it does move, then we need to make some adjustments.
14. Lets assume for a moment that your camera is aligned so that the horizontal crosshair represents the RA axis and the vertical
line represents the Declination axis. If the comet drifts to the left of center over time, then we need to slow down the RA
pulses as the mount is moving to fast. If the comet drifts to the right, we need to speed up the pulses. The same goes for the
DEC axis; speed up or slow down the pulses until you get the comet to stay put in the same place for at least two minutes.
15. A few notes regarding fine tuning: M ake very small adjustments. For instance, if your RA rate is -0.0246 and you see that it is
moving too fast, slow down the rate to -0.0200 and see how that affects movement. If you get lost and cant tell which way is
up, then make very large adjustments to get a sense of direction. Also, it is expected that the nucleus will wander second by
second due to seeing and mount imperfections, the better your equipment and PE, the better you will be able to track. The
main goal here is that we can keep the nucleus in the same general position for several minutes. You will also find that the
rates you get from TheSky6 are going to be very close and some cases no adjustments will need to be made. Lastly, don't give
up! I usually spend a good 60 minutes getting to the end of this step. Have patience and take frequent breaks from the screen
to avoid seizures.

Obviously the above procedures are specific to PulseGuide and an Astro-Physics mount. I am sure that the same RA and DEC
adjustments can be made on other select mounts with other procedures.

There is another way to solve the problem of tracking a comet that might seem much easier that I would not discount: An external
guide scope and camera guiding on the comets nucleus. If you have a STV (or other guiding camera) and a guide scope, you should be
able to point the guide scope at the comet and auto-guide satisfactory. Just make sure not to calibrate on the comet as it is moving
and will provide inaccurate mount movement information. Use a star in the same field of view to calibrate your guider.

Step #4 - Planning your Sequence:

Now that we have are mount tracking on the comet, the background stars are moving and the comet is staying put. This will allow us
to get some very crisp shots of the comet's structure. But the very fact that the stars are now moving relative to the comets means
that long exposure photographs are going to have lots of streaks in the background (or foreground) of the image.

This step combined with careful calibration and image processing will allow us to get the best of both worlds, a perfect picture of the
comet without stars and a wonderful star background.

For the sake of understanding, I am going to jump ahead and show the end result of what we are looking for from our acquisition
sequence:
 Figure 1: Non overlapping stars.

The above image represents 10 luminance images taken for Comet C/2004 Q2 (M achholz). The 10 images have been mean combined to
show their respective star locations. As you can see, the comet stays in the same place, but the stars move in a north east direction.
If I were display the same 10 images for the red, green and blue channels, they would like very much like the luminance image above.

The key to the above image is that there are no stars that are overlapping one another. Carefully planned exposures have givin the
mount enough time to move between exposures to allow for no stars to overlap for the entire image set. Compare this to a single long
exposure of the comet. Notice the familiar streaks?
 Figure 2: Long exposure star streaks.

Fortunately, with the first set of images, we will be able to use a special combine technique to remove the stars completely from the
image. The second image however, has no hope of removing the trailing stars. M ore on that later.

Setting up your imaging acquisition sequence takes careful planning based on a several factors: Your imaging systems image scale, the
speed at which the comet is moving and the objects that are going to enter and exit the field of view. For the following example, I will
plan for an acquisition sequence using my Takahashi FSQ-106N refractor and SBIG ST10XM E camera.

The main goal is to space out our exposures so that we do not end up with any overlapping stars in any of our master luminance, red,
green or blue channels. The best way to spread out the amount of time between exposures and channels is by using a combination of
sequencing and delays. Take this sequence as an example:

1 X Clear @ 40 seconds @ 2x2 bin.

60 second delay.
1 X Red @ 40 seconds @ 2x2 bin.
60 second delay.
1 X Green @ 40 seconds @ 2x2 bin.
60 second delay.
1 X Blue @ 40 seconds @ 2x2 bin.
60 second delay.

Lets assume it takes 5 seconds to download the image and 3 seconds to change filters adding 8 seconds to each exposure. This would
yield a total running time of the above sequence of 432 seconds (108 seconds per exposure + delay * 4).

By repeating this sequence 10 or more times, we ensure that each integration for each channel is 438 seconds apart from one
another. The question is, is this enough time to avoid star overlap?

In my case, I shot the above comet on Dec 23, 2004 when it's RA and DEC rate was -0.0240, 0.0649 respectively. M y imaging system
yielded an image scale of 5.3 arc-seconds per pixel at 2x2 bin mode. The brightest object in the field was HIP 19508 at a magnitude of
7.36 (seen on the far right in the above images). The reason 40 seconds was chosen as the exposure time was because of blooming
limits on my ST10XM E. So lets assume that this is a constant. The factor we need to calculate is the delay between exposures.

The following steps will assist you in coming up with the proper delay between exposures:

1. Take one 40 second, 2x2 binned image through the clear filter. and measure the diameter of the brightest object in the field in
pixels. M ake sure to stretch the object enough to replicate the stretching that will take place later in processing. In my case,
the 7.36 magnitude star measured 5 pixels in diameter.

2. Knowing that our brightest object in the field is 5 pixels in diameter means we need to make sure that we have at least 7 pixels
of movement of this object for each sequence to avoid overlap.

3. To calculate the amount of time needed for our entire sequence to run, we calculate the following:
( ( reqPixelMovement * imageScale ) / rateOfComet ) where rateOfComet equals the greater of both the Comet's RA or DEC
movement rates. In my case, (( 7 * 5.3 ) / 0.0649 ) or ( 37.1 / 0.0649 ) = 571 seconds.

4. Since we know that 160 seconds will be taken up by exposures (4 x 40) and 24 seconds by other functions (4 x download times
and filter wheel changes), we have already used up 184 seconds with our exposure time. By subtracting the constant 184
seconds needed for exposure from the 571 seconds required to avoid overlap, we end up with an extra 387 seconds needed for
delay.

5. By dividing 387 by four and rounding up a few seconds we end up with a required delay of 100 seconds between exposures.

You can see that our original estimate of 60 seconds was not long enough to completely avoid star overlap. (BTW, this was the case
with the recent image I took of C/2004 Q2 and I ended up with less than desirable results. Now that I know the math, I can avoid this
next time).

Its is important to understand that a comet's RA and DEC rate changes over time. For instance, on Jan 25, 2005, C/2004 Q2 will have a
RA rate of -0.0144 and a DEC rate of 0.0546, M uch slower than on Dec 23, 2004. M ake sure to check the rate of your comet every night
you attempt to image it and adjust your delay times accordingly.

There is nothing like a real world test to check our assumptions, so the next step is to take the proposed sequence three times to
confirm we are avoiding overlap. In my case, I use CCDAutoPilot for image acquisition sequences. The following is an example of how I
would set up the test sequence:

After running your test sequence, load the three clear exposures into your image processing software. Do an average combine with all
three integrations and confirm that the stars are spaced appropriately as in Figure 1 pictured above.

One last note, you may need to align the three images on the comet's nucleus before you do your average combine to see exactly
how the stars line up. We will be aligning on the comet in our calibration stage, so its a good idea to check it now. M ore on that
later...

Step #5 - Acquiring the Data:

Congratulations! You have passed all the difficult steps. Its on to acquisition and image processing.
First off, lets discuss some reasons for my exposure and binning reasoning. I have chosen to image comets with my ST10XM E and FSQ-
106N because of its wide field of view (comets are big), its fast focal ratio and camera sensitivity. I will soon attempt to image C/2004
Q2 with my STL11000. I suspect that I will bin the camera 3x3 for added sensitivity and shorter exposure times. But why the binning
and short exposures?

I choose to keep my exposures short to avoid blooming and to create a large stack of exposures in which to combine. I will be using
one of several different combine types that will all benefit from lots of sub-exposures. The reason for binning is to increase the
sensitivity of the camera, once again for several reasons; overcome the read-noise of the camera, bring out more color from the dim
comet and increase my image scale to make unguided imaging more reliable. You can experiment with your imaging set-up to find the
best combination of exposure length and binning modes.

Now, on to acquiring our data. We are going to do three separate imaging runs to capture the required data for processing. I call this
the 'comet-stars-comet' sequence. Its fairly simple at this point assuming we are all set-up with our tracking and timing. The sequence
goes like this:

Run this sequence five times with Pulse Guiding enabled:

1 X Clear @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Red @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Green @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Blue @ 40 seconds @ 2x2 bin.
100 second delay.

Run this sequence two times with Pulse Guiding disabled and the mount moving at sidereal rate:
1 X Clear @ 40 seconds @ 2x2 bin.
0 second delay.
1 X Red @ 40 seconds @ 2x2 bin.
0 second delay.
1 X Green @ 40 seconds @ 2x2 bin.
0 second delay.
1 X Blue @ 40 seconds @ 2x2 bin.
0 second delay.

Run this sequence five times with Pulse Guiding enabled:

1 X Clear @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Red @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Green @ 40 seconds @ 2x2 bin.
100 second delay.
1 X Blue @ 40 seconds @ 2x2 bin.
100 second delay.

Total imaging time for the above imaging session will take approximately 100 minutes to
complete.

Lets look at each segment of the above sequence to understand what we are trying to accomplish.

The first segment captures 5 integrations for each LRGB channel while tracking on the comet. This segment will bring
out the detail of the comet while spacing the stars out so to not overlap.
The second segment is acquired with the mount moving at sidereal rate so that we can capture the background star
field. The reason for taking two exposures is strictly for redundancy. We don't want a rouge satellite ruining our star
field.
The third segment is identical to the first and will be combined together with the first segment to create the image
of the comet.

The concept is to capture the comet for the first half of our imaging session, stop tracking the comet, take exposures of the starfield,
and then capture the comet for the second half of our imaging session.

Why not take the star field first, then 10 comet exposures? This is personal integrity at play here. I feel that for the shot to be
considered accurate, one should take the star-field in the middle of the total imaging session to preserve some sense of reality. The
fact is, with the following image processing techniques you could place the comet on top of any star field you wish. I leave the choice
up to you, but urge you to follow my technique for the sake of scientific integrity. BTW - The official time of the image would be
clocked at when you took the star field exposures.

M ake sure to check ahead with TheSky6 to ensure you don't have a M agnitude 3 star entering the field of view during your imaging
session as this will surely bloom your camera and exceed our calculated star diameter threshold.

I also encourage you to take more than 10 total comet exposures if possible as the more exposures you have, the better your signal to
noise ratio and the more subtle detail you will bring out in the comet's tail.

Step #6 - Calibrating, Aligning and Combining the Data:

This step assumes that you have a basic understanding of calibration and alignment techniques for CCD imaging. So I present a simple
checklist of items we will need to calibrate our image sets.

20 x 40 second dark frames to create a master dark frame to reduce all images. Remember these are 2x2 binned.
10 flat frames for each of the LRGB filters. Some just create a flat-field from the clear filter to calibrate all filters, I
recommend flats for all filters.
20 bias frames to reduce our flat frames.

We have four channels we need to process, Clear, Red, Green and Blue. I will use the data we acquired for the clear channel for my
walkthrough.

1. Load the 10 clear light frames we acquired while tracking on the comet into your image processing software. This would
include 5 light frames from our first imaging sequence and 5 frames from our third imaging sequence from Step #5.
2. Calibrate the 10 light frames with your master dark and flat frames.
3. Align the 10 light frames on the centroid of the comet's nucleus. You may need to do a min/max histogram stretch to get the
nucleus to a point size that you can see clearly.
4. TIP: These are very under-sampled images. Using a Bilinear re-sampling alignment algorithm is going to provide much cleaner
results than a Bi-cubic algorithm. You can experiment with both and see the difference yourself.
5. At this point you should have 10 calibrated and aligned images. Save them. As a test, average (mean) combine all 10 of your light
frames and confirm that they look something like this:
 Figure 3: C/2001 Q4 Neat - 10 Light Frames Average (M ean Combined)

6. Now that we have confirmed that are images are in alignment relative to the comet, and our stars don't overlap. A job well
done!
7. At this point, make sure you have all 10 calibrated and aligned light frames saved as FITS files.
8. Repeat the calibration and alignment step for the remaining red, green and blue channels.

Now that we have all of our light frames calibrated and aligned, its time to combine them to make the master luminance frame.
Depending on what software you use, there are different methods that will produce the best results. I offer the following examples
below:

Mira Pro 7

1. Load all 10 of your calibrated and aligned clear light frames into an image set.
2. Combine the images using the 'M inimum value' method making sure to normalize the images using the 'Scale' option.
Maxim DL 4.0

1. Load all 10 of your calibrated and aligned clear light frames into M axim.
2. Combine the images using the 'SD M ask' Output.
3. Use the following SD M ask settings:

Sigma by Ray Gralak:

1. Load all 10 of your calibrated and aligned clear light frames into Sigma.

2. Use the following settings to combine the images:

If everything goes well, you should end up with an image of the comet completely devoid of any stars while retaining a high signal to
noise ratio. Your image should look something like this:

Figure 4: Combined Light Frames with M ira Pro M inimum Value - Comet only sans stars

Other image processing programs will have similar combine methods which I urge you to experiment with.

Now repeat the entire process for the remaining red, green and blue channels to create your four master light frames.

Step #7 - Creating our Luminance, RGB and Starfield images:

We will now create three images which we will import into Photoshop for final image processing. A luminance channel, a RGB color
channel, and a star field channel.

At this point we should have four master light frames. I will refer to them as neat.clear.fit, neat.red.fit, neat.green.fit and
neat.blue.fit. We should also have ready one set of our clear, red, green and blue star field images which we aquired at sidereal rate
in the middle of our acquisition run. If you recall, we called this Segment 2.

In the following examples, I will use M axim DL 4.0 to process our raw data into our three master channels for import into Photoshop.

Master Luminance Channel:

I find comets difficult to stretch manually in programs such as Photoshop as there is much dynamic range to bring out. The bright
nucleus, the faint tail, or in most cases multiple tails of different intensities. If you are feel comfortable stretching the images manually
than I urge you to give it a try as you will always have greater control.

I find the DDP stretching works quite well with comets and therefore, this is my preferred method of making histogram adjustments to
the raw, combine data. I present the following method for DDP stretching the clear data to create our Luminance channel.

1. Load neat.clear.fit into M axim DL 4.0.

2. Open the Histogram (Screen Stretch) window if it not already opened.

3. From the pull-down on the right select M ax Val.

4. Select FILTER > DIGITAL DEVELOPM ENT to bring up the Digital Development dialog box.

5. For Filter Type, select Kernal - User Filter.

6. Select the 'Set User Filter...' button and use the following settings:

Selecting the User filter to the above settings will avoid any sharpening to the image during the DDP process.

7. Now we are ready to stretch the image. Start by selecting the Auto check boxes and show in the above Digital Development
dialog box.

8. You will notice when looking at the Screen Stretch window that M axim's automatic settings will usually clip the low-end of the
histogram and compress the white point a little more than we want to at this point. Fine tuning to the image can be done later
in Photoshop. Here is an example of what you might see with the Auto setting:
 Black point clipping and excessive white point compression.

9. Uncheck the Auto check boxes (and the M ouse checkboxes) in the Digital Development dialog and manually adjust the
Background and M id-level values.

10. You will want to lower the Background Value and raise the M id-Level to achieve a histogram that looks more like the following:

The resulting image should look something like this:

 Figure 5: DDP Stretched Comet in M axim DL 4.0

Save this image as a 16 bit TIFF file called neat.lum.tif. We will use this later in Photoshop. Congratulations, we have just created our
master luminance channel!

RGB Color Channel

The next step is to create our RGB Color Channel.

1. Close all open images in M axim DL 4.0.

2. Open up our four combined raw files (neat.clear.fit, neat.red.fit, neat.green.fit and neat.blue.fit).

3. Using M axim's align feature, align all four images using the centroid of the comet nucleus as your reference point. Also make
sure to use the clear channel as your reference image and to uncheck Bicubic Resample. Use the following Align Images
settings:

Align Images Dialog Box

4. Now that you have all four channels aligned, its time to combine the channels to create our RGB image.

5. Select COLOR > COM BINE COLOR to bring up the Combine Color Dialog box.

6. Set the conversion type to LRGB.

7. Uncheck 'Allow Resize'.

8. Check 'Bgd Auto Equalize'. (I have found that manual normalization of each channel for this type of data is not necessary).

9. Set your Luminance Weight to 100%.

10. Set your Red, Green and Blue color channel weights accordingly.
 Combine Color Dialog Box Settings

11. Once you have confirmed that your settings are correct, click the OK button. You should get a result similar to the following:

Figure 6: Un-stretched LRGB combined comet image.

Now that we have our raw LRGB color channel, we will repeat the process of DDP stretching the image to produce the following
results:

Figure 7: DDP Stretched LRGB color combined comet image.

You can now start to see some of the wonderful detail of the comet's nucleus and tails consisting of wonderful green, blue and subtle
red shades.

Save this image as a 16 bit TIFF file called neat.rgb.tif. We will use this later in Photoshop. Now on to the star field.

Star Field Channel

Creating the star field channel is straight forward.

1. Close all open images in M axim DL.

2. Load the clear, red, green and blue FITS images from our star field exposures.

3. Align the four channels on the stars NOT the comet Nucleus.

4. LRGB Color combine the four channels.

5. DDP Stretch the LRGB result.

You should end up with the following result. The comet is very faint compared to the above images, but the stars are well pronounced
and will be slightly sharpened later in Photoshop.
 Figure 8: Star field channel DDP stretched.

Save this image as a 16 bit TIFF file called neat.stars.tif. We have now created all of our channels for processing in Photoshop. On to
the next step.

Step #8 - Image Processing in Photoshop:

Now we come to the fun part! Importing our images into Photoshop, creating our layers, and making our final adjustments. The finish
line is well in sight.

1. Open neat.lum.tif, neat.rgb.tif, and neat.star.tif into Photoshop.

2. Combine all three images into one file and name the three layers Lum, RGB, and Stars.
 Photoshop Layers Dialog

3. Set the Blend mode for the lum Layer to Luminosity and the Opacity to 20% as a starting point.

4. Set the Blend mode for the rgb Layer to Lighten and the Opacity to 100%.

5. Set the Blend mode for the stars Layer to normal and the Opacity to 100%

6. Adjust the levels and curves of the lum Layer to bring out details to taste.

7. Adjust the levels, curves and saturation of the rgb Layer to taste.

8. Adjust the saturation and color balance of the stars layer to taste.

9. Apply a 1.0 pixel radius Un-sharp M ask to the stars Layer to tighten up the stars a bit. Dont over do it our you will lose the
color of your stars.

10. We also need to mask out the bright comet nucleus in the stars Layer as it creates a strange artifact with the rgb Layer. The
stars Layer should look something like this:
 Figure 9: M asking out the Bright Nucleus of the Stars Layer in Photoshop.

At this point, its all about making fine tune adjustments to your image until you are pleased with the results. I personally do not like to
alter the original data that I imported into Photoshop, so I use Adjustment Layers to make adjustments. They are very handy because
the changes you make are not permanent allowing you to go back and change settings without having to use the undo or history
pallete.

Here is what the my final Layers dialog box ends up looking like when I have completed my processing:
 Using Adjustment Layers in Photoshop.

It took quite a bit of planning and patience to capture this comet, but the results are quite rewarding. Here is the final image after
image processing in Photoshop:
 Figure 10: Final Composite of C/2001 Q4 Neat.

Conclusion:

I hope you have found my tutorial useful and it inspires you to capture a comet! I really enjoy the challenge of imaging comets and
always look forward to the arrival of a new ice ball.

If you have any questions regarding any of these techniques, or have techniques of your own, I would love to hear from you. I can be
reached at rbennion@ewellobservatory.com.

Happy Comet Hunting!