Category: Astronomy

There’s been some confusion about the meaning of the timestamp in the upper right corner of the Seestar app’s live-stacking screen.

For this discussion let’s assume you’re capturing 10-second exposures.

The Seestar evaluates each exposure, looking for stars it recognizes and making sure there are no star trails, airplanes, satellites, or other abnormalities. If it likes what it sees, it adds that exposure to the finished image. Technically, this is called “live stacking”. Seestar calls it “enhancing”. As each exposure is added to the finished image, 10 seconds are added to the time in the upper right. This is called “integration time” — it’s exposure time multiplied by the number of exposures.

There is a pause of about 1.5 seconds between exposures, so you won’t get 6 10-second exposures per minute of clock time. It’s also the case that the Seestar seems to not even try to accept every image it shoots — more like 1 every 3 at times. I haven’t figured out the pattern for this.

I’ve found if I have ideal conditions (including targeting an object whose path across the sky gives me minimal field rotation in 10 seconds), I get around 4 minutes of images for 5 minutes of clock time. But that’s ideal. Normal for me under Bortle 1 sky (very dark, high-quality seeing) is about 1.5-2 minutes of integration time per 5 minutes of clock time. So in an hour I might get 25 minutes of integration time. Over the course of 6-8 hours, as an object traverses the entire sky — including crossing the meridian at high altitude where there is a lot of field rotation — I’ll get anywhere from 1-2 hours of integration time.

I just processed 2 projects that each had 11 nights of data. The first had about 72 hours of total clock time and about 7 hours and 40 minutes of integration time. The other was the same 11 nights but only about 24 hours of total clock time. It’s path was “flatter” across the sky during those hours. In that case I got over 11 hours of integration time.

These are my experiences and might be dependent on conditions where I am and on my particular Seestar. It might also depend on whether I’m doing a mosaic or not (that is, framing is set to something other than 1.0).

I get asked about this a lot. This is my current workflow. I’ll try to keep it up-to-date but this should at least be close.

Disclaimer: I’m not an expert at this. I’ve just spent a few more hours watching YouTube videos and experimenting than you have.

Why Take the Time to Learn?

Here’s an example of why you should learn to process your Seestar images in something more than a photo editor (Lightroom, Photoshop, GIMP, Acorn, etc.). The first image is right off the Seestar. A lot of people would bring this into an image editor and crank up saturation, brightness, and contrast. That’s what you see in the second image.

The third image is processed using the workflow described below. I just took the stacked .fit file from the Seestar and ran it through GraXpert, Siril, and Topaz. This allowed me to eliminate the noise, remove the stars and process them separately, and bring out details that are present in the original.

Click on an image to see it at full resolution.

This is just a quick example. Processing took about 20 minutes. The point is to show the relative difference between what you get out of the Seestar, what you get from a photo editor, and what you can get with a few minutes of work.

Stacking on the Seestar

For single sessions, the Seestar automatically creates a stacked .fit file. Just use that.

For multiple sessions over several nights, the cheapest and easiest stacking solution is to stack on the Seestar. This works with up to about 2000 frames for sure, but after that it can get iffy. It may do 6 hours of work then crash at the end. But if it works for you, it’s great.

To stack on the Seestar, connect to the device with your PC or Mac and delete (after making a backup copy, of course) everything except the object-name_sub folder containing your subframes. Within that folder, sort by file type/kind and delete all the .jpg files. Just keep the .fit files.

Now use Deep Sky Stacking on the home page of the Seestar to stack your subframes. When done, if your subframes were in object-name_sub, then the result will be in the object-name folder.

Astro Pixel Processor (Stacking)

APP is a paid program (about $60/year or $230 for full license) but does a great job of stacking Seestar mosaics.

1. Select a working folder.
2. Import all your subframes as “Lights” on tab 1.
3. Uncheck the 3 boxes at the top tab 1.
4. Switch to tab 6 and scroll to the bottom and press the integrate button.
5. Go get a cup of coffee, play a round of golf, or run a half marathon.
6. Go through the frame list at the bottom of the screen and de-select all the failed frames.
7. Go to tab 1 and select the Clean button next to Lights to remove the de-selected frames.
8. Back to 6 and integrate again.
9. APP will save the final fit file in the working directory you specified earlier.

GraXpert

1. Start with the stacked .fit file.
2. Crop if needed.
3. Switch to “AI” interpolation method for background extraction. You may have to go to the Advanced tab on the right to download the latest model. Experiment with Smoothing value. I use .8. Then calculate the background.
4. Set Denoise Strength (experiment with best results; I use .8) and denoise the image.

If you are not going on to Siril:

• Adjust Stretch and Saturation at the bottom until it looks good. Saturation won’t get saved with the image.
• Choose “Save Stretched & Processed” and save as a 32-bit Tiff.
• You can do further brightness, contrast, and saturation adjustments in Lightroom, Photoshop, Acorn, Gimp, or whatever. Do this on the .tif file. Save as Jpeg or PNG.

If you are going to bring the image into Siril:

• Choose “Save Processed” and save as 32-bit Fits.

Siril

Siril has more options than you can learn in one YouTube video or by reading these steps. Here’s my general workflow as of today.

• Press the Home button and choose the folder containing your .fit file.
• Press Open and open the .fit file.
• At the bottom of the screen, choose “Autostretch” instead of “Linear”.
• Open the menu in the upper right, choose “Image Information” then “Image Plate Solver”. Choose “OK” to identify the object and flip your image to the “correct” orientation. Close the dialog. (If this doesn’t work, enter the name of the object in your image at the top of the dialog and press the find button to look it up, then press “OK”.)
• Open the “Image Processing” menu” and choose “Color Calibration”, then “Photometric Color Calibration”. Press “OK” to calibrate colors. (Again, if this doesn’t work, enter the name of the object at the top of the dialog and press the find button to look it up. Then press “OK”.)
• Open the “Image Processing” menu (top left) and choose “Star Processing” then “StarNet Star Removal…”. Choose the option to pre-stretch and the option to generate a star mask, then “Execute”.
• After a few seconds you’ll have a starless image.
• Change from “Autostretch” to “Linear”. Your image will be mostly black.
• Choose “Image Processing” then “Histogram Transformation”.
• Repeat the following: Note the position of the histogram. It will be mostly vertical on the left side of the window. Find the triangle that is closest to it on the right and move it left until it almost touches the histogram. It doesn’t have to be super close. If there’s room, move the leftmost triangle to the right until it almost touches the histogram. Press “Apply” then repeat this step as many times as you can.
• You’ll get to where the left and right triangles are just to the left and right of the histogram curve. You can move them slightly into the histogram as long as the image looks better. Press “Apply” each time you move them. When you’re done, close this dialog and choose “Save”.
• Apply any of the following that make the image look better:
• Asinh Transformation
• Color Saturation
• Noise Reduction (I find this doesn’t work well)
• Background Extraction (I find this doesn’t work well)
• If something you do makes it look worse, or if you want to compare what it was before, use the “Undo” and “Redo” buttons to the left of “Image Processing”.
• When you’re happy with the starless image, press “Save”.

If you have a tool to do noise removal or further editing, like Topaz Labs’ Photo AI:

• Save as a 32-bit TIF
• Load the .tif file into Photo AI or your image editor for further work. Don’t change the size or do any cropping. If you change the size at this point, your stars won’t recompose correctly.
• When you’re satisfied, save as TIF.
• Open the TIF in Siril.
• Save as 32-bit FITS.

Continue in Siril:

• Choose “Open” and open the file that starts with “starmask”.
• In “Image Processing”, “Star Processing”, choose the gear icon in the lower right corner to show the “Dynamic PSF” dialog. Select the button with stars that have circles around them to detect the stars in your image. You can leave this dialog open while you do the next couple of steps.
• In “Image Processing”, “Star Processing”, choose “Desaturate”.
• In “Image Processing”, “Star Processing”, choose “Full Resynthesis” and apply that.
• If you dismissed the “Dynamic PSF” dialog (gear icon in “Star Processing”), bring that up again. Select the delete button in the lower right corner to remove the circles around your stars.
• No need to stretch the stars. Just save the starmask by pressing “Save”.
• Open “Image Processing” menu, choose “Star Processing”, then “Star Recomposition”. Open your starless image on the left and starmask image on the right. If you edited the starless image outside of Siril, make sure you select the FIT file you created from the TIF you got from your image editor.
• Stretch the starmask image to about .5 for now — just enough to see the stars — and “Apply”.
• Select an “interesting” portion of your image by clicking and dragging your mouse to produce a small rectangle. You’ll have to experiment with what to select here. I try to choose a feature I want to emphasize or better separate from the background. After making a selection, press the eyedropper next to “Symmetry point” on the left side of the recomposition window.
• Adjust “Stretch factor”, “Local stretch intensity”, and “Black point” until you like the image better. If you don’t like it, choose “Reset”. Once you get an image you like, choose “Apply”. Repeat as often as you want until it looks good.
• Now go to the “Star Stretch” side and adjust “Stretch factor” until your stars look like you want them. You can adjust “Black point” to get rid of smaller stars. Press “Apply” when you get it how you like it. Repeat as often as you want.
• Close the “Star Recomposition” dialog.
• Choose the down-arrow button to the right of “Save”. Name the file something you will recognize. I use object name, number of subframes, and exposure time. So for M 31 the name might be “M 31 999x10s star recomposition result”. Choose the file type in the bottom right, and save.

Photo Editor

• Open the file in your favorite photo editor for final adjustments. Save as JPG 80% or higher (the higher the number, the more accurate the pixel values are to your original).