Sunday, March 3, 2024

Automated Solar Eclipse Sequence in NINA

Everybody always says to not image a solar eclipse but just witness it instead -- but why have just one when you can have both!

In the past, I've successfully automated my eclipse imaging using BackyardNikon with my DSLR, and it worked great. I started the main eclipse sequence exactly 1 minute before 2nd contact, and I had carefully timed everything and practiced a bunch, and I got some great images in 2017 and 2019 without having to look at my camera or computer once during totality.

But now for 2024, I'm ready to use an astro cam instead, which opens up some great possibilities with automation. NINA is my tool of choice since there is so much flexibility in the exact ordering and timing of events.

I've been talking about making this sequence for a while, but with PhD things, I'm finally just now having some time -- and I finally got it done and tested! A lot of people have asked if they can have a copy of the sequence, and I'm happy to share.

UPDATE (April 2, 2024): I've created a "simplified" version that has just the partial phases, Bailey's Beads/Diamond Ring, and then a giant loop over all bracketed exposure times in Fred Espenak's table. This way you only have to modify a couple of timestamps around the contact times. How many seconds before or after C2/C3 to end Bailey's Beads/Diamond Ring is annotated. This makes it a lot easier to adjust if you have to change locations (and it's just easier to use to begin with).

Files:

Whole eclipse, with time spans instead of clock times for testing
Excel spreadsheet for planning times
README file with the same information in this post
Link to the folder with all of these



IMPORTANT NOTES

You will need to modify this sequence! It is set specifically for timing at the location I will be at, and the exposure times are set for a specific calibration point in Fred Espenak's table (more on that in a second). I'm also including the spreadsheet I used to work out all the timing of each imaging phase. 

Exposure Times

First, you need to calibrate your astro cam to Fred Espenak's exposure table. Scroll down on this page to the "Solar Eclipse Exposure Guide." It's designed for DSLRs, so you'll need to choose a gain on your camera to match an ISO/f-stop value so you know which column to use. 

To do this, set up the telescope and camera in the configuration you're going to use for the eclipse. Put on the same solar filter you're going to use, and find out whether it is ND 4 or ND 5 (how dark it is). My Seymour Solar filter is ND 5. Get the sun in your camera's field of view. Then, in the live-view app of your choice (SharpCap is perfect for this), choose an exposure time from the table in the ND 4 or ND 5 row, depending on which filter you have, and set that as the exposure time on your camera. For instance, I used the column where the partial phase with an ND 5 filter has an exposure time of 1/250s, or 4 ms. Then I adjusted the gain until the sun looked good. I used the Histogram tool and looked for when the combined histogram peak (white) had a hump around 50% (you'll also have a peak at darker values -- this is the black background). You don't want too high of a gain so you don't sacrifice dynamic range, nor too low that you end up in the last column and the outer corona images end up being very long. On my ZWO ASI2600MC Pro, I chose 210. So these settings put me in the column with the partial phase ND 5 exposure time of 1/250s, and I can base the exposure times for all the rest of the phenomena from there. 

For the eclipse, I'm using the whole FOV, not the ROI I'm showing here.


Eclipse Timing

To get all the contact times, I used Xavier Jubier's website, where you can put in the coordinates of where you plan on imaging from, and it will tell you the "local circumstances," including each of the precise contact times, the altitude & azimuth of where the sun will be at that time, and some other numbers. If you have a backup location, I would make a separate sequence for that location if it is different enough in its timing. 

Organization

There are three Sequential Instruction Sets -- pre-eclipse partial phase, totality, and post-eclipse partial phase. The totality instruction set has several sub-sets for the different eclipse features -- the sliver while you remove your filter before 2nd contact, Bailey's Beads right around 2nd contact, chromosphere shortly after 2nd contact, the corona out to multiple solar radii (shorter exposures for the bright inner corona, and longer exposures for the dimmer outer corona), prominences when the moon is centered at mid-eclipse and a "long" exposure for outermost corona and hopefully Earthshine, then another round of corona shots because there's time (if you're uncertain about your gain setting, you could do a different set of gain/exposure times for this second runthrough), and then finally chromosphere just before 3rd contact, Bailey's Beads & diamond ring through 3rd contact, and then some short exposures while you put your filter back on. Then the post-eclipse partial phase picks back up.

Each segment loops until the start time of the next feature; no need to time how fast your capture rate is and estimate a number of exposures to fit the right time frame. The segment will stop when it hits the "loop until" time and move onto the next segment automatically. Most of the segments are 15s long. During the partial phases, the sets of bracketed exposures are 5 minutes apart; feel free to change this.

Other Notes


Bracketing

Each phenomenon segment has three exposure times -- the one from Fred's table, and then half that, and then double that. It's a long eclipse, so there's plenty of time for this.

Computing Power

While NINA isn't especially resource-heavy, you'll definitely want to test the capture speed if you're using an older machine or a light tablet. It's a lot of image download very rapidly, and it might get bogged down.

If you have a computer with a solid-state hard drive, highly recommend using that to maximize frame rate. If you're using an astro cam with USB 3.0, definitely use a computer with a 3.0 port. 

I am using my Microsoft Surface Pro 7 -- less power-hungry than a laptop, but pretty capable, and it as a USB 3.0 A port.

Bailey's Beads

On my camera and computer, it takes about 2-3 seconds per frame to capture and download, with star analysis and stretching turned off. Make sure you turn these off! (Toggle the buttons for these at the top of the Image tab in the Imaging view). This isn't ideal for very quick events like Bailey's Beads, so I decided to try something, but only on the 3rd contact Bailey's Beads, just in case -- using video mode. There's a plugin for NINA called LuckyImaging, which adds a container type called Lucky Target Container, and within that, I use the Take Video Roi Exposures. I turned off using an ROI -- I could probably get a faster frame rate using one, but just in case the sun ends up not quite centered, I don't want to miss it. You can try using an ROI if you have good enough tracking. This lets me take 3 fps at a single exposure time. It still saves out FITS files. In the # box, I've got 100 frames set (it will stop when it hits the "loop until" time), and for the exposure time, the Bailey's Beads exposure time from Fred's table. I should also get the Diamond Ring here. The other reason I'm not doing this for the 2nd contact Bailey's Beads is so that I can bracket, just in case these exposure times end up a little off.

Safety Notes

As always, never point your scope at the sun without a solar filter firmly in place. Don't take it off until at most a minute before totality; you're safe within 30 seconds of totality for sure with a refractor, maybe shorter if you're using a large-aperture reflector. Also make sure you cover any guide scopes or spotting scopes, or remove them. Never put solar eclipse glasses on the eyepiece side of a telescope or pair of binoculars, you can burn out your eyes and your optics!

Make sure your computer is in the shade! In 2017, my tablet shut off about 40 minutes before totality because it overheated; fortunately I was able to get it cooled and back on before totality. I put it in the shade of my telescope case.

My Equipment

Wondering what I'm planning on using? 
  • Telescope: Astro-Tech 72ED
    •  It has a focal length of 430mm, so a good wide FOV for corona. A bit longer would probably be better, but this scope is a lot lighter and smaller than my Takahashi FSQ-106N.
  • Camera: ZWO ASI2600MC Pro
    • Color camera, APS-C format sensor. You could certainly use an uncooled camera, or one with a smaller sensor if your focal length is short enough. I don't recommend doing mono; it would be a lot of filter changes and focus shifts and basically a lot of dead time to get each color. Note: you don't need an H-alpha filter to see the prominences and chromosphere -- they become visible during totality!
  • Mount: Celestron NexStar SE
    • I chose this mount because it is alt-az instead of equatorial, which means I don't need to polar-align the night before (or with a compass/tilt meter during the day, which is approximate at best). I can just plop it down, put in my coordinates and the time, say "here's the sun," and it will track it pretty well. I may have to nudge it once or twice, but it will definitely keep the sun centered for the duration of totality. As far as tracking stability goes, when I used to use this mount for deep sky back in the early dates, I could take images as long as 20 or 30 seconds before too many of them started to streak.
  • Computer: Microsoft Surface Pro 7, a tablet of pretty good computing power but in a compact and not-power-hungry package.
At this position, I might not be able to reach the 68 degrees the sun will be at for me -- I'm going to add a longer dovetail so I can sit it further forward.

Troubleshooting

I was having trouble with my 2600 having an error and disconnecting for exposure times longer than 1s; I figured out that even though it can run off USB power alone for shorter exposures, it needs the 12V power, even though I'm not cooling, for longer exposures. Plugging it in solved the problem. I'm not cooling so that I can run on battery power if I have to (using my Jackery); if I have AC power, I'll probably run the cooler at -5C or -10C, if it's not too hot, to reduce noise for the longer exposures. The max exposure time I'm using is 15s, so not much noise will accumulate, but still. 

The sequence is quite large -- it might take several seconds to open.

If you notice any errors or have questions, you can reach out to me either on Facbeook or via email at astronomolly.images at gmail dot com. You can also post a comment below.

Want to see my previous eclipse notes?