TV Device Tasks

At the ATNF, the only TV device of interest is the X-windows based server, xmtv. Refer to Section 3.1 for details on how to start them and refer also to the on-line documentation ( help server). You can compute on one machine and display on another if you wish.

The tasks that use TV servers are tvdisp , tvinit , tvcycle , tvflag , tvall , and tvset .

Of these, the functionality of tvall and tvset are now included in the local capabilities of the servers. tvflag has been discussed in Chapter 9. We will not discuss these tasks further. You know what to do for more information.

WARNING: Note the following very unpleasant features of the xmtv server. There is a status line on the display which lists the location of the cursor and the value of the pixel under the cursor. The location is in server coordinates and has nothing to do with image pixels. The value is only just useful. First, it only gives correct numbers for the last image displayed on the server. If you have two images loaded, one in each channel, it will only give correct numbers for one channel, and that will be the last one into which you put an image. Second, if you have displayed the image with an intensity range set (see keyword range in tvdisp

below), any pixel intensity from the original image outside the range will be truncated to the nearest range extremum. This is because the server does not know anything about the image, just the scaled numbers given to it.

• The task tvinit is used to initialize the TV display. Its only input is the server keyword.

• The task tvdisp is the basic task to display images on a TV device. It can display single images, or many planes (channels) from a 3-D image and then optionally run a movie. You can fiddle with the image (zoom, pan, look up table) with the buttons on the local server. When displaying an image with a large dynamic range, it is often necessary to map only a certain range of image intensities to the TV grey scales. For example, if you want to see low-level artifacts in an image with a strong point source somewhere in the field, mapping the image minimum and maximum to the available TV grey levels will result in your being able to see the point source only. You would need to set the mapped image minimum and maximum to a subset of the available full range. Image pixels whose values are outside the specified intensity range (keyword range) are set to either black or white (or the appropriate colour). The number of channels in the TV is a function of the server. The window-based servers at the ATNF have only two channels (so you can blink images).

  Here is an example of how to display four channels from an image.

  

  Once the image is displayed, you can fiddle about with it (zoom, pan, modify transfer function, activate another TV channel, etc.) using the server's local buttons. They are pretty self-explanatory.

  Note that tvdisp tries to put the entire region you select from the image on the TV. Thus, if you ask it to display a large spatial area and many planes from a cube, it may not be able to fit them into the two available channels (the TV has a finite number of pixels). In this case, tvdisp will display only as many planes from the cube as it can and still fit all the selected spatial pixels on the TV. You can use the incr keyword to tell tvdisp

  what spatial and spectral increments to use if you wish.

  The next example shows how to run a movie of all the channels in a 3-D image (cube).

  

  When the movie option is given, and if the `control panel' task, xpanel, is running (see Chapter 3), then a control panel will appear to give you control over the movie. Most of the buttons are self-explanatory. We describe the more obscure ones:

  • Stop stops the movie sequence running, but allows you to re-start or single step afterwards. The Exit button causes tvdisp to exit completely.
  • The Loop/Osc button toggles the behaviour of tvdisp

    when it reaches the end of a sequence. It can either loop through all channels (e.g. wrapping around from the last frame to the first frame), or oscillating back and forward (e.g. the movie sequence reverses after it reaches the last frame).

  • The ReZoom button causes tvdisp to adjust the zooming of each frame. When tvdisp starts the movie sequence, it adjusts the zoom so that (as close as possible) one frame fits within the TV window. If you adjust the TV window size after the movie starts, then use ReZoom to re-adjust the zoom factor.

    Adjusting the window size can be quite useful, so that the window encomposes a single frame (flickering of neighbouring at the edges of the current frame can be annoying). Also making the window small allows the movie sequence to be much fast.

• The task tvcycle can be used to cycle through channels on the TV. The local server button enables you to switch between channels manually, but often a faster blink is wanted. tvcycle is the task for you. The x-location of the cursor controls the cycle speed from a base interval set with the delay keyword (it defaults to 0.5 s).