Jupiter, Europa and Io

I filmed the giant Jupiter a whole night as its Great Red Spot entered and left view. I also filmed Io exiting from behind Jupiter then move away and Europa moving towards the planet. And put all data together to show it as a time lapse movie.

Setup and metadata

I used my known setup: the 150/750 newton scope on the motorized EQ3 mount, a 2x barlow + extension tube (another barlow tube with the lens removed) + Scopium webcam (640×480, 8 bits RGB). I disassambled and adjusted the barlow+extensino tube+webcam assembly two times during the night to get the most out of the setup in terms of a great disk photo, knowing when putting all the frames together compensating for the rotation, xy offset and size will be only a small fraction of my problems.

To be noted that the observation started with Jupiter relatively low, near the horizon (about 30 degrees, in a city) and that the seeing varied substantially during the session, peaking at an almost perfect seeing, one I’ve never seen before from within this city.

I took the first video on November 3rd, 2014 at 0:53 UTC and the last one at 4:36 UTC. Location: Kolozsvár city, Romania, EU.

The setup (this was taken on October 10, 2014)

Laptop with at least 100 GiB free space (this was taken on October 10, 2014)

The screen (this was taken on October 10, 2014)

Tech data and fun facts

I recorded a total of 68 videos, varying lengths (1-2 minutes), totalling 100 GiB of raw data. To go on with some napkin math, this translates to about 3838 seconds (about 64 minutes, ie. 1 hour and 4 minutes) of video data, which in turn means 115,050 frames recorded during the night. I wore five pants and stilled wished I had a space suit. During the observation Io moved at least 186,000 kilometers (1,33 Jupiter diameters on the images) – but knowing that Io has an orbit 2,649,619 km to complete in 42,5 hours, its speed should be 62,343 kmph – at this speed in 3,76 hours Io moved 231,916 km. That’s 60% the distance at which the Moon orbits us.

I processed each video in Registax 6 and/or Autostakkert (AS!2) to get the individual frames of the rotation video. Some results were just bad, some not so bad and some left me wondering how this relatively humble setup shot a picture this great.

Europa and Jupiter. Io is about to come out

Europa and Jupiter. Io is coming

Io coming out… a better frame of the video

Io is coming into view

Io well out

GRS facing us

GRS mocing away

GRS is almost out

Io moved a lot

One stacked video, becoming a frame for the rotation-video

Moments…

Moments…

Second setup: note the larger disk and the different hour angle

Second session, fourth video stacked. Before wavelets

Second session, fourth video stacked. After processing

frame

Third session…

GRS makes the rotation easy to see

GRS moving away

moments…

GRS almost out

The three sessions: each have their own hour angle and disk size

FujiFilm HS20EXR image, tripod, 1 s. Io is lost in the glare.

FujiFilm HS20EXR image, tripod, 1 s. Io is lost in the glare. (crop)

Jupiter’s family, composite. Ganymede and Callisto added using their position recorded with the Fuji.

Creating the rotation movie

So I had data from basically three sessions, although consecutive ones. Since the mount was not guided and no polar alignment holds for hours at this magnification, Jupiter slowly but constantly drifted. These combined meant that the three data sets needed to be (1) rotated to fit, (2) resized/resampled (I also adjusted the extension tube resulting in slightly different magnifications), and (3) all the frames aligned (Jupiter moved a lot around).

From a previous attempt of mine at imaging Jupiter’s rotation I had a home brewed, rather basic php/gd program to align the frames. Back then I did the aligning manually using a blinking interface. Now I added some code for the software to detect Jupiter’s location on the frame, measure its size and estimate how much resizing/resampling is needed. And now take a step back. Detecting a light blob on a dark background is easy. However, athmospheric conditions changed, exposure and contrast needed to be adjusted as well, so the apparent diameter of the resulting planet varied even when the scope was left untouched. After my program estimated, I entered manually the values that seemed correct. There were some rounding errors too making the planet wobble one pixel – corrected manually. Although made my best, the video revealed some further, minor errors which made Europa jump around a bit. I used the position of the slow moving Europa on neighboring frames of each session to match the hour angle of the sessions.

And the outcome is nothing more than a half minute video. Watch it in HD

I also put together these animgifs:

A different crop and colors

And below is a longer video about what and how I recorded on October 10 – the same equipment and principles explained in the video below were used on both ocasions.