Friday, June 30, 2017

A Beginner's Guide to Solar System Photography with the Celestron 127 SLT (and other Alt/Az Scopes) Part 3

Update: Based on feedback, I have broken the original large post into four smaller posts to make it easier to read.

Focus, Grasshopper....

Believe it or not, the trickiest part of the whole process is getting a sharp focus. You'd think it would be easy, watching the laptop screen, but because the atmosphere is moving, and because the scope mount shakes a bit when you touch it, it's actually kind of hard.

Trust me on this. As early in the process as you can, make or buy a Bahtinov mask. They are a wonderful focusing assist tool. You will save yourself a great deal of frustration. There is nothing quite like the feeling of rolling out of bed at 3:00 AM to shoot Saturn, and later processing the images and determining that you were a touch out of focus. My eyes don't work right at 3:00 AM.

Making one is easy, thanks to the Bahtinov mask generator at AstroJargon. Simply plug in your scope's numbers and print it black. Then have the paper laminated, and cut out the pattern with an exacto blade. I left 4 tabs evenly spaced around the circle that I folded over and added velcro tabs to secure it to the telescope. I also used a Sharpie to black out the back side to minimize reflections.

You can also 3d print one, if you have access to a fairly large printer. The laminated paper one is holding up just fine.

To use it, attach it over the front of the scope. Orientation doesn't matter. If you have a dew cap, you can attach it to the end of the dew cap with no concerns - the distance to the objective is not critical.

Point the telescope at a bright star as close to your target as is reasonable. You need a point light source, not a disk, so pointing it at the moon or a planet is not recommended. Adjust your gain/exposure until you can see three lines making up the diffraction pattern. Two will cross in an X, and the third will be off center when it crosses the other two. As you focus the telescope, that third line will move. When it's centered at the intersection of the other two lines, you're in focus. A gentle touch on the focuser helps, and you'll need to let the scope settle each time you touch it. One of these days I'm going to build an electric focuser, but that's a project for another article.

In the picture below, the middle line, closest to vertical, is the focus indicator. It's not centered between the other two lines, indicating the system is not in focus.

The capture software I'm using has a Bahtinov assist feature, which quantifies how many pixels off center you are. It doesn't always reliably identify all three lines, so it can take a little time for it to settle down, but if that number is oscillating around 0 or 1, you're in good focus. Here's what the sequence looks like.

Way out of focus.  Turn the focuser in the direction that makes this smaller.

Getting better, but still off by 4 pixels

Focus acheived. 

Once you are properly focused, remove the mask and slew the scope over to your actual target. Remember, the mask won't work pointing it at a disk - it needs bright point source. A bright star near your target is best. Slewing the scope can impact your focus, so minimize the movement.

"Lucky Imaging" - making cleaner still images from video

It may seem counter-intuitive at first, but unlike terrestrial photography, you'll get your best results not by shooting a single frame, but by shooting hundreds to thousands of frames. You then use stacking software like AutoStakkert to automatically select the best frames, align them, and combine them into a single image.

The advantage of this approach is that you can overcome some of the variability in atmospheric seeing that is causing the view to swirl and ripple. There are split seconds during which the image is clearer than average, and those frames get selected and combined. By combining many of these images, you can reduce noise, and dramatically improve detail. Essentially, you are improving the signal to noise ratio by integrating data over time into a single image - the details and noise don't happen in the same parts of the image each time, so you get better detail over the whole image as you add frames.

When you are shooting, use a gain in the middle of your camera's range. If you decrease the gain, you'll get cleaner individual frame, because noise increases with gain. However, your exposure time has to increase, and this means that you can't capture as many frames. If the camera exposes for 1/30 of a second for each from, you get 30 frames per second, maximum. Modern astro cameras are much faster than that - connected to a USB3 port you won't be bandwidth limited until 100 fps or more, if you aren't capturing the full view of the camera. More frames increases the odds of getting frames that capture those split seconds of clarity as the atmosphere shifts and ripples.

USB3 and a solid state disk (SSD) are preferred, but I'm currently using a laptop that is limited to USB2 with acceptable results. It does limit my maximum frame rate when shooting bright targets. You can reduce the impact by only capturing a small window of the camera's view - the planet will generally easily fit in the middle 800x600 or 640x480 pixels, so you really only need to capture that.

The resulting files are BIG. It is really easy to shoot 100 GB in a session. Fifty GB is about the minimum. The resulting video files are not compressed.

Start doing video captures of 2-6 minutes. I set the motor speed of my control handset to 3-4 for this, after the planet is centered, so that I can make very small adjustments to keep the planet centered. The tracking is good if you align the scope carefully, but not perfect.

During capture, don't move the scope more than you have to - remember that the stacking software can take care of even significant drifting from center as it aligns the image. Moving the scope more than needed will reduce the available pool of good frames the stacking software has available to it.

Raw video frame 


Best 20% Stack of Video Frames

Wavelet Sharpened

Effects of Planet Rotation and duration of video capture

After a couple tests where you capture a few minutes of video, it will occur to you that you can get more frames simply by recording longer videos. I tried 10 minutes at a time on Jupiter, and couldn't figure out why the results showed very little detail compared to the 4 minute captures.

It's pretty simple - everything's moving. While you are taking video of it, Jupiter is rotating, and it rotates FAST. It completes a full rotation in a bit over 9 hours! If you take a couple of sequential 5 minute captures and produce stacked images from each, you'll be stunned at how much it has moved during this time - you can actually see it rotating if you cycle through the resulting pictures.

Limit your exposures of Jupiter to 4 to 6 minutes, and your exposures of Saturn to perhaps 7 or 8. Repeat this several times if you can - the seeing varies substantially from minute to minute, and one of these will usually turn out better than the others in the same session. The time required to shoot another 6 minute sequence is small compared to the time required to set up and focus.

There is a way to use software to de-rotate the planet in longer exposures, but that is beyond the scope of this article. I will link to it when completed. If you are interested, check out the software package Winjupos and the tutorials online.

No comments:

Post a Comment