TV: Display an Image

For more general TV help, type HELP X11.

TV displays the image in 'buf' on the X11 image display. Only those pixels with data numbers between Z and Z+L are displayed. Data values greater than the maximum are all shown as one color, although the interactive cursor will still be able to display the data associated with the "clipped" pixels.

If no intensity scaling is given on the command line, TV will assume Z=0 and L=3 times the image mean intensity. If the mean has not been calculated, TV will do a quick calculation using every 9th pixel. A peak intensity of 4x the mean is a reasonable values to use for many astronomical images with a large dynamic range. It is not so good, however, for viewing flat fields, skies, or dark images that have a narrow range of data values. In the latter case, you should specify the display levels explicitly on the command line.

BOX=n tells TV to only display that part of the image within box 'n' (see BOX command).

OLD tells TV to use the same display limits as the previous image.

CF=filename loads the color map ("look-up table") used to translate pixel values into colors. 'BW' loads a black and white color map. Once the map is loaded, it remains in effect until changed by either the COLOR or subsequent TV commands.

The basic set of color maps is:

More are available, see the COLOR command for specifics.

The NCOLOR= keyword can be used to limit the number of colors displayed in X windows. By default, 256 colors are displayed. A limited color map is useful when VISTA is being run at the same time as other color-hungry applications (e.g., SAOimage, or web browsers like Netscape). If VISTA asks for too many colors, the X server can refuse to open the image display. An alternative is to shut down the unneeded applications when running VISTA.

The TV display uses a 4-place "ring buffer" to hold the 4 most recently displayed images in memory. You can page through these images using the and $-$ keys when the mouse is on the TV display window. Blinking may be accomplished by displaying two images sequentially, and then using the $+/-$ keys to blink between them.

If the blink buffers are getting filled with a confusing mix of images, the CLEAR IMAGES command may be used to clean them out (except for the currently displayed image) and allow you to start over again.

The default image coordinates are as follows:

        |             |
     |  |             |
     |  |             |
     R  |             |
     |  |             |
     V  |             |
        |             |
            ---C-->   (EC,ER)

The origin is at the upper left-hand corner, with columns increasing to the left and rows increasing from top to bottom. This is the "natural" order in which to directly map images from memory onto the display, and allows us to display images as fast as possible without remapping.

The FLIP keyword is provided to invert the rows axis so as to increase vertically from bottom to top, i.e., in ``Cartesian'' coordinate order. This will give you an orientation that is more ``natural'' from the user's point of view, but which incurs some performance penalty.

The properties of the display (size, whether or not to auto-dezoom, etc.) can be customized using X resources variables set in the user's .Xdefaults file. See HELP X11 for an overview.

Whenever an image is in the display, a set of basic cursor keys are available to provide simple interactive capabilities. See HELP X11 for details.


  1. \begin{example}
\item[{MN 1\hfill}]{}
\item[{TV 1 CF=RAIN\hfill}]{loads image 1 into the display, setting the
span to be 4 times the image mean.}

  2. \begin{example}
\item[{TV 3 100. 30.0\hfill}]{loads image 3 into the display, setting the
span to be 100 and the zero to be 30.}