Friday 27 January 2023

Astrophotography - attempt 2

There were a few hours of clear skies on the evening of 25/26 Jan, so I did another session. My initial attempt to block out the street lamp flare by using the lens hood didn't work - test exposures showed a nicely illuminated 'circle' around the outside of the image -- with the lens hood encroaching into the FOV and being well lit by the street lamp. I think the lens hood must be meant for use with my 50-200 mm zoom lens rather than the 18-55 mm zoom lens.

So I removed the lens hood and relocated from the front porch the our back yard, where our house and boundary fencing block out all direct illumination from street lights and the neighbouring houses (can't do much about the Bortle 5 sky darkness in my back garden. Driving 15 minutes to a nearby national park provides a slight reduction in light pollution (Bortle 4) which will be helpful to try and image the E3 comet on 11 Feb. If I drive 3.5 hours to my hobby farm/lake house I have access to a 'dark sky' which is Bortle 2. The best dark sky site I could access is at Coonabarabran, close to the Anglo-Australian Telescope (AAT) at Siding Spring Observatory -- unfortunately that is a 6 hour drive from Sydney!). Orion was getting low by the time the skies had cleared, so I simply aimed at a patch of sky above Orion. Unfortunately having the camera pointed more towards the zenith made it a bit awkward to check the focus after each test shot, and my homemade Bahdinov mask (cut from some paper and cardboard) didn't seem to work very well, so after ten minutes I gave up and simply turned the focus ring all the way over to 'infinitiy'. Big mistake. Especially as I didn't bother taking another test exposure to check the focus, but simply cleared the SD card and commenced taking a series of lights, darks, biases and flats.

I took 11 sets of 9 x 3 second exposures at F/5,5, 55mm FL and ISO 12,800. The sets of 9 exposures were due to the D3300 self-timer mode only supporting up to 9 shots in a sequence. You then have to press the shutter button again for another set of 9 shots (after the initial delay of 5 seconds I had set to minimize camera shake from pressing the shutter button). I also had to reset the self-timer mode after each set as the camera display shut down after a few seconds and reverted to single exposure mode. I could probably fiddle with the shut-down delay setting to avoid having to reset the self-timer setting after each series, but that might cause an issue with battery life (the fully charge battery lasted for two nights - about 3 hours of camera use - before needing to be recharged).

I had also worked out how to get properly exposed 'flats' using the 'A' mode (aperture priority) that kept all the manually setting unchanged, but worked out the correct exposure time. I used a piece of fine weave white cloth (a lot of people use a 'white T-shirt') over the lens and the 'light box' app on my tablet to provide a uniform illumination. The flats seemed OK (the automatic exposure time was 1/8th second) with the histogram being fairly central for the flat exposures.

I processed the lights, darks, biases, and flats using SiriL, and all went well until it got to the 'find stars' stage required to register (align) and stack the lights. The software reported the process failed, with '0 stars found' - which is due to my lights being too out of focus. The software finds stars by finding the centroid of a star (a couple of illuminated pixels), and won't work if the image is too out of focus. Looking at a couple of my 'lights' the stars were totally out of focus and spread over dozens of pixels! I definitely should have taken a single test exposure and checked the focus was 'reasonable' before wating an hour taking exposures.

I've made two small purchases to improve future astrophotography sessions with my DSLR and tripod - I bought a commercially produced 'Bahdinov mask' for about $6 (it is being shipped from China, so probably won't arrive until the end of Feb -- well after I want to image Comet E3 around 11 Feb). I also ordered a cheap ($24) intervalometer from Amazon.com that will allow me to 'program' a series of exposures and then execute them with a single button press on the hand controller. This will allow me to execute an entire series of 100 or more 'lights' with a single button press, rather than having to use a series of 9 shots each individual started. Also, because I won't have to touch the camera at all during the series of exposures, I won't have to have a delay between each exposire -- so rather than a 3 second exposure being followed by a 2 second delay, I can just fire off a series of 3 second exposures one after the other -- reducing the overall time by around 40%. Once the 64 GB SD card on order has arrived (to replace the current 8GB SD card) I will also be able to increase the number of lights, darks, biases and flats from 99/54/45/45 to 300/100/50/50 - which should improve the signal to noise ratio and bias compensation. There isn't another clear night predicted for a week or more, so hopefully my SD card (and possibly the intervalometer) will have arrived before my next imaging session.

How to waste an hour by not checking focus before taking a series of 99 3-second exposures. D'Oh!

Subscribe to Enough Wealth. Copyright 2006-2023

No comments: