This question touches on the basic point of the program. SIP is an astronomical image processing program. It is not simply an image display program. There are already plenty of good image viewers that will display gif or jpg images (or whatever is your favorite, common format). Some of those viewers even allow you to play with the display parameters (grayscale, color, etc.). SIP is different, in that (1) it let's you display FITS images (the standard file format of astronomical images), and (2) it lets you carry out standard astronomical processing procedures on the image data (some of these processes, such as adding/subtracting images, or dividing one image into another image, are not possible in typical digital image processing programs, which are not written specifically for astronomical image processing).
Finally, a gif image (to take a particular format at random) of some astronomical object is usually a particular representation of the data in the original image (which was probably a FITS image) --- the person that produced that gif image enhanced it in whatever way they wished, and you cannot see/recover the original data values. For many astronomical purposes, the gif image is "damaged goods." It may be pretty, but you're limited in what you can do with it. On the other hand if you start with an original FITS image you can select how you display it, and you can select what gif image you end up using to represent that data. You can also do astronomy with the FITS image (e.g., photometry), using SIP.
My image was taken with a square CCD, but it doesn't look square in SIP's display. What gives?
If your image was obtained with a CCD that has a square chip but uses non-square pixels, the image will not look square. SIP assumes the pixels in your CCD are square. SBIG CCDs such as the ST5, ST7, and ST8 have square pixels (so the appearance of the displayed SIP image will accurately reflect the appearance of the object). However, the ST4 CCD has pixels that are 13.75 by 16 microns in size (i.e., rectangular), yet the chip is 2.64 by 2.64 mm (i.e., square). So an ST4 image will look rectangular in SIP with the long axis being 16/13.75 = 1.164 times longer than the short axis. There are only two minor problems this will present for a SIP user: (1) the image will appear slightly stretched in one direction, and (2) calculating the distance between two locations requires a little extra thinking. The accurate distance (in microns) between a pixel located at coordinates (x1, y1) and another pixel at (x2, y2), is the square root of (a^2(x1-x2)^2 + b^2*(y1-y2)^2), where a=13.75 and b=16.00, assuming the y direction is along the short length of the displayed SIP image; a=b for an image taken with a CCD having square pixels. Similarly, an accurate area measurement for the region bounded at one corner by the pixel (x1, y1) and at the opposite corner by the pixel at (x2, y2) is a*(x1-x2)*b*(y1-y2). Any camera using the TI-211 chip (as does the ST4) will present these difficulties. Another such camera is the Lynxx CCD camera from SpectraSource.
The ST6 camera has pixels of size 23 by 27 microns in a 375 by 242 array; the chip is not square and the pixels are not square. The resulting ST6 image will look rectangular in SIP; the real extra length of the image along one axis (due to the rectangular nature of the chip) is exaggerated in the displayed SIP image (due to the rectangular pixel size). In computing accurate distances between pixels use a=23, b=27, where, once again, b corresponds to the factor appropriate for pixel coordinate differences along the short length of the display SIP image.
Why not allow SIP to perform [your-favorite-processing-task] on a image?
I am certainly open to suggestions for processing tasks to incorporate into SIP. However, keep in mind that the philosophy behind SIP is to provide an astronomical image processing program that is simple. The tasks provided are meant to cover the basics of image processing, yet provide the basis for performing complicated tasks in a step-by-step fashion using the tools provided. That way students (the main target of this program) see and learn how the complicated tasks actually work.