Michael Ravine, Advanced Projects Manager at Malin Space Science Systems, Inc., has kindly agreed to share this fascinating post about how he and his team translate black and white images from JunoCam on the Juno Jupiter Orbiter into color images.
This is an attempt to summarize the main points of making images from Junocam.
Junocam acquires images as small, rectangular framelets through three different color filters. The spacecraft is rotating, which we use to scan Junocam’s field of view to make of a picture. The raw frames look like this:
I should point out that all of the images we take with Junocam get released as soon as we at Malin Space Science Systems (http://www.msss.com) finish the initial processing. The data are then available on the Mission Juno website:
https://www.missionjuno.swri.edu
The images available there include the raw “Venetian blind” images like the one above, and various processed versions, include some done by a number of interested amateurs who’ve made a real contribution to getting the most out of the camera.
It turns out that the three color filters in Junocam, which we refer to as blue, green and red, are not a particularly good approximation of the human color visual response. So, if you just mosaic together the framelets for each color channel and composite them together, you get this ghastly mustard color:
If you want more technical background on this, see the spectral response plot (Fig. 13) in the Junocam instrument paper, available here:
https://www.missionjuno.swri.edu/pub/e/downloads/JunoCam_Junos_Outreach_Camera.pdf
Anyway, this is most assuredly not the color of Jupiter you would see if you were there. If we just fix the color balance to something we think is closer to reality, we get this:
Now, the problem is that we’re looking at the whole planet from the sunward limb to the terminator. So, the brightest place (in this picture, high noon) to sunset, where the brightness goes to zero. Meaning that most of the brightness variation you is just telling you the direction to the Sun, which isn’t what you are interested in. If Jupiter had a phase function with an nice, analytic form (say, Lambertian, which just goes like the cosine of the solar incidence angle), we could divide that out to “flatten” the image. In reality, we flatten the image using an empirical function that does a better visual job. Also, as is the case with every spacecraft image you’ve ever seen, we stretch the contrast up some to make the cloud features more visible, ending up with this:
You might think this last step is a bit dishonest, but I would argue that’s a matter of interpretation. If you were actually there, you wouldn’t be taking a single picture of all of Jupiter with your eyes, you’d be building up an image in your mind by scanning your eye all over the disk. When you looked at the South Pole, you pupils would dilate and let in more light so you’d be able to see the details of the circumpolar cyclones. And when you looked and the limb, which is much, much brighter, your pupil would stop down, letting in less light so you would get a clearer impression of the lower latitude belt-and-zone banding.
And you might wonder, why such a goofy way to take pictures? It made sense at the time, but that’s another story (there is some of this background in the paper linked above).
Juno is in a highly elliptical orbit around Jupiter, and once every 53 days it does screaming past the planet really, low, really fast. Since we take a series of Junocam images on each one of these passes, it’s possible to use them to generate a movie of the this. Such a movie (done by Gerald Eichstädt, who also did the image immediately above) can be found here:
https://www.youtube.com/watch?v=y3qnWRMRqQI
And if you make it out that way, you’ll have to let me know if I was right about the colors.
Michael Ravine
Advanced Projects Manager
Malin Space Science Systems, Inc.
