Venus cloud tracking -- July 31, 2006

<- Previous page

7. Making movies

The best way to see the cloud motion in a series of aligned images is to make a QuickTime movie. This is also a good way to encapsulate a sequence of images for distribution over the web, since it merges several files into one.

Making TIFFs

The first step in making a movie is to convert the sequence of FITS files to compressed TIFFs. This can be done with the 'fits2tiff' command line program:

2: ~/Desktop/New071304/code > fits2tiff ../processed/c1230.fits 1e3 0

This utility takes 3 arguments: the name of the file to convert, and the display scale maximum and minimum values. Values outside this range will be clipped.

(Click for full-size)

Compressed TIFF files are a fraction of the size of FITS files, in this case 128 Kb from a 2 Mb image.

Using QuickTime Pro

To make movies from TIFFs, you need to have QuickTime Pro, which is available for $30 from:

Arrange the series of TIFFs in a folder with consecutive names, and then choose 'File->Open Image Sequence...' and select the first image. Although you can change the animation rate, the default of 2 fps is reasonable.

(Click for full-size)

The sequence of images will be imported and assembled into a movie. Choose 'Movie->Loop' or 'View->Loop' (depending on your player version) to continuously loop the movie.

(Click for full-size)

Choose 'File->Save' or 'File->Save As...' to save the assembled movie, and make sure to click the 'Make movie self-contained' radio button before saving. Note that this movie of 7 images is only 1/2 the size of a single individual FITS file.

(Click for full-size)

Raw and composite movies

QuickTime movies can be uploaded to a website for display. Although a movie can be embedded directly into a web page, I do not recommend doing this. Instead, create a link from the page to the .mov file (you can use a picture as the link), so that the stand-alone movie will be downloaded and opened in a separate page or QuickTime Player window.

For example, here are 2 movies made from the raw first images of each interval chosen previously, and from the processed 10 image composites of each interval:

(Click for movies)

8. Enhancing composites

Before making vector fields, each composite image was enlarged to 2x size, masked, and enhanced by an unsharp mask using a Gaussian filter. The mask used was a combination of a circular cutout plus half an ellipse of minor/major axis ratio 0.45 oriented at an angle of 10 degress clockwise. All pixels not within the mask were set to zero so as not to affect the cloud tracking program.

A Gaussian filter of sigma = 10 was convolved with the enlarged and masked composite and then subtracted from the original at a mix ratio of between 0.6 and 0.56 to create an unsharp masked (i.e. 'sharpened') version of the composite so that features in the clouds could be more easily detected by the tracking program.

Here is an example of the enlarged and masked image, and the convolution filter used:

(Click for full-size)

Here is an example of the blurred intermediate and the resulting sharpened image:

(Click for full-size)

This technique was more effective on composites from later intervals in the image sequence, as they were generally of poorer quality than earlier images. Although this procedure did bring out substantial detail in the clouds, it also has the adverse effect of increasing the noise and non-flat field abberations.

Here is a movie made from the set of enhanced composites:

(Click for movie)

Next page ->

İSky Coyote 2006