The field is being imaged in the y direction and dispersed in the x direction. The system is treating the point sources just as a conventional spectrograph would. The y direction holds the wavelength information.

The instrument takes images while rotating 180 degrees yields a full data set. The images are then compressed along the y direction so that each is one pixel wide while keeping the original hight of the image. Below is a description of how the image is compressed.

We create one image-line for a given span of time, for this we will use one second. The spin rate of our payload was 360 degrees in 100 seconds, 3.6 deg/sec. So, for 180 degrees of data, the payload has 50 image-lines. A typical stack of image-lines would look like the below graphic.

Looking at the picture, you can see that the point sources in the image are represented as sine curves. The amplitude of the sine is the radial distance out from the center of the rotation, and the phase offset is the angular distance around the field. In 2 dimensional spherical/cylindrical coordinated, the amplitude represents 'r' and the phase offset represents 'theta'.

The stacked image-lines are referred to as a sineogram. This is the form that the data is fed into the reconstruction algorithm.

This method allows us to gather much more information than either a conventional spectrograph or a conventional camera. The spectrograph yields spectral information of an area of limited size, and the camera yields spatial information of an area for a given wavelength range.

This method allows us to gather full spectral information for a large region at one viewing position as well as imaging that same area.

We can reconstruct the image for any given wavelength range. Also, the individual spectra of the point sources can be extracted or removed. As was previously stated we are trying to gather absorbtion information as well as dust scattering across the region. By extracting individual stellar spectra, we can look for absorbtion features. By removing all the point sources from the data set, we are left with the dust information. Below are some examples of flight data.

This is a flight image integrated for about 5 seconds. You can see the point sources being dispersed across the image. The bright vertical line in the image is hydrogen ly-a (Lyman-Alpha= 1216 Angstroms). This is a local feature that is very bright in the region of the atmosphere where we observed. It appears that ly-a is cenetered in a vertical band of lower intensity. This area was designed to have lower sensitivity because of the strength of the ly-a line.

This is a sinogram representing 180 degrees of data (50 seconds of observing) Each image-line is about 0.3 seconds of data, the reconstruction works best when given a roughly square input matrix.

This is a reconstruction of the above data. The bright stars shown here are (from left to right) beta, delta, pi, sigma, all in the constellation Scorpius.