This chapter deals with the VDU software - anything to do with how things are put on to the screen (i.e. the television or monitor). What 'modes' are and why they are there is covered first, followed by a section on writing text and then details on the graphics routines. Lastly the palette is covered. All the individual VDU commands are listed for reference in the next chapter.

When a letter is written to the screen it has foreground and background colours - the colours of the ink and paper. When the machine is turned on, it is always white foreground on black background. Colours (or more strictly logical colours - see the section on the palette) are labelled from 0 upwards. To take a definite mode for simplicity, mode 1 has four colours labelled from 0 to 3. Try the following:

This puts you in mode 1 with white text on a black background.

This sets the foreground colours to number 1 (red). Text after this command is red on black.

This sets the background colour to number 2 (yellow). Text after this is now red on yellow, and clearing the screen with CLS makes the entire screen yellow. Why 130? Because to change the background instead of the foreground colour you must add 128 to the colour number. Thus, to get background colour 2 (above), add 128 to give 130.

Changing mode resets the colours to white on black. As said before, any VDU commands (including COLOUR, GCOL, MOVE, DRAW, etc) can be either typed straight (as a 'direct command') or used as part of a BASIC program.

Each letter position has its own address in the usual columns and rows format. The column numbering is from left to right starting from column 0 and the tows, as for all VDUs, are labelled from the top (row 0) downwards. How many rows and columns there are depends on the mode - the drawing below shows the labelling for mode 6.

IMAGE MISSING

The cursor may be positioned to any part of the screen with the TAB(X,Y) command, thus the following program prints out a diagonal line of As.

Normally, the text may appear anywhere on the screen. However, a text window may be set, which allows the text to appear only inside the window. To do this, the VDU 28 command is used as follows:

VDU 28,a,b,c,d

where a,b is the bottom left and c,d the top right position inside the window (see the drawing below)..

IMAGE MISSING

Nothing outside the text window is affected by text commands, such as screen clearing, scrolling, cursor positioning, etc. Note that the TAB(X,Y) measures from the position of the top left of the current window. Try the following program:

Both text and graphics windows are removed by VDU 26.

Each character is an 8by 8 matrix of dots (pixels). All the normal letters, numbers and punctuation marks are defined, but it is possible to define your own. 256 bytes of RAM are set aside for the definitions of characters whose codes are from 224 to 255.

Character definitions are entered thus:

where CODE is the code of the character to be defined (it is then printed using either VDU CODE or PRINT CHR$(CODE);) and

L1 is the bit pattern of the top row
L2 is the bit pattern of the second row from top, and so on until...
L8 is the bit pattern of the bottom row..

What is the bit pattern? Each dot in any one row is given a number, and the bit pattern is the sum of the numbers corresponding to those bits in foreground. These numbers, labelling the bits from left to right, are 128 (for the leftmost pixel), 64, 32, 16, 8, 4, 1 (for the rightmost pixel). Specific examples are easiest to understand.

The space character obviously has no foreground, thus all the bit patterns are zero, so to assign the space character to the code of 224, the command

would be used. To define a large X, the top line has the left and rightmost pixels set only, thus L1=128+1=129. The next line has the second from left and the second from right pixels set, thus L2=64+2=66. Similarly, L3=32+4=36 and L4=16+8=24. The fifth through to eighth rows are the mirror image of the first four, so to define the character 225 as an X, type the following line:

All the characters from 32 to 255 may be defined, but to define those outside the codes 224-255, it is necessary to allocate more memory for the fount. This is called 'exploding the fount' and is done via FX call number 20.

Here is another example of defining a character. The alien in the BUGZAP program on the Introductory Cassette was made up on the matrix in the drawing below:

If you use the code 224 for the new character definition, here is the VDU statement which defines the complete character:

By changing L7 and L8, the 'upright' alien shown in the drawing below can be defined. The code for this character must be a different one from the one above (eg 225), otherwise you will lose the original alien.

Have a go at defining this new character, then check the result by displaying it on the screen with

The program below shows how you can produce an animated alien by using both these characters.

Line 40 gets rid of the flashing cursor, which would otherwise be a distraction. You can retrieve it by typing

The graphics instructions are pretty extensive in the Electron, and they all have certain things in common. The easiest commands to understand are the MOVE and DRAW commands, and these will be used for illustration in the following section. The ideas presented here are true for all graphics commands (including CLG).

Remember: when you press BREAK, the computer is in mode 6. This is not a graphics mode and nothing will happen when you try to plot things. Always remember to go into a graphics mode to try these things out. Mode 1 is a good one to start with. Similarly, programs should always have a MODE command in them, as described at the beginning of this chapter.

Firstly, we must describe the coordinate system, that is to say how positions of points are labelled. This is similar to the text coordinate system but there are three differences.

Firstly the system have the point (0,0) in the bottom left hand corner, and row numbers are labelled upwards.

Secondly, the top right hand point on the screen is (1279,1023), the same in all modes (see the drawing below). This is so that drawing a line from, say (100,100) to (400,400) always draws a line in the same place, even though the pixel size varies with the mode.

Lastly, points off the screen are well defined, that is to say, drawing a line from, say (-300,-400) to (300,400) is perfectly legal, and what appears is what you would expect - that portion of the line that is in the area viewed by the screen.

The graphics screen

IMAGE MISSING

The graphics cursor is an invisible point on the screen, and is where you are about to draw from. Move it about the screen with the MOVE command, and drawing is easiest with the DRAW command. Thus:

moves the cursor to (100,100) and draws a white line to (400,400). Try lots of lines in different modes to get the feel of the coordinate system.

Just as the foreground and background colours of text were defined using the COLOUR command, so the corresponding colours in graphics are defined using the GCOL command. Try the following:

This draws a line in colour 1 (red) from (0,0) - where the graphics cursor is when the mode is changed - to (300,300). However, you will notice that the GCOL command has two numbers after it. The second is just like the COLOUR command's number, that is the foreground colour number, or, if 128 is added to it, the background colour number. CLG is the graphics equivalent of CLS and clears the graphics area to the current graphics background colour. Thus

sets the graphics background colour to 1 and clears the graphics screen to this colour (red). Note that the CLG command is much slower than the CLS command.

The first number in the GCOL command is unusual. It tells the computer what to do with the graphics point. The following values are defined:

Other values do stripey things which may change with different releases of the software.

What is meant by OR, AND, EOR and INVERT? Each pixel has a colour - in mode 1 with four colours, this is from 0 to 3, or 00, 01, 10 and 11 in binary. What appears on the screen is the result of a logical operation between the pixel you want to plot and what is already on the screen. The OR and the AND are the same as for the BASIC commands. EOR means 'exclusive OR', which is the same as OR unless both bits are one, in which case the result is zero. Again, all this is most easily explained by specific examples. The following assumes that you are in mode 1.

Assume there is a red screen (from GCOL0,129:CLG above). Setting GCOL 1,2 sets the foreground colour to 2 (10 in binary), and the colour 'mode' to OR. Drawing a line then takes the red pixel on the screen (red - colour 1 = 01 in binary), and ORs it with the yellow pixels you are plotting. The pixel colour that appears is the 01 ORed with 10, which is 11, colour 3, which is white. (Try it.)

Given this white line on a red background, set the colour with GCOL 2,2, which has foreground colour 2 and colour mode AND. Plotting a line then takes what is on the screen and ANDs it with the yellow pixel, colour 2 or 10 in binary. Therefore when the line is plotting on the red background, 01 (red) is ANDed with 10 (yellow), then result being 00 (black). If it crosses the white lone - the white pixels (11) are ANDed with the yellow pixels (10) to give 10 (yellow).

Setting GCOL 3,131 sets the background colour to 3 (white) and the colour mode to EOR. Doing a CLG then EORs its pixel with 11 - that is to say 00 goes to 11, 01 to 10, 10 to 01 and 11 to 00. The screen is thus inverted in colour, and repeating the command restores it to its original state.

If this does not seem too clear, playing around with it for a little should help. It has two main purposes - setting the EOR mode allows erasure of a line by plotting it again. In four colour modes, two independent two-colour pictures may be drawn and selectively displayed using the palette.

MOVE and DRAW are two special cases of the more general PLOT command, which is as follows:

where K% is the plot mode (i.e. what you are actually going to do); X% and Y% are the coordinates of the point to which you are plotting.

K% takes the following values:

This plots a solid triangle using as vertices the point specified, the graphics cursor and the previous graphics cursor. This can be used to fill many different shapes.

This plots the line sideways from the specified point, left and right, until either the edge of the window is reached or the line meets a pixel of non-background colour. The graphics cursor is set to one of the end points, and the previous graphics cursor (used in triangle plotting) to the other.

The values of these may be found out via FX call number 13 (decimal). If the point specified is outside the graphics window, or is not on background colour, then the Y coordinates of the returned points are different.

Filling right until background colour plots right from the specified point as far as either the edge of the graphics window or a pixel of background colour is found. The endpoints (note that it does not go left) are retrievable via FX 13 in the same way. The endpoint is actually the first pixel of background colour found, thus if the specified point is background colour, the endpoint returned is the same as the specified one.

These two routines together enable a fast routine to be constructed to fill any enclosed shape.

The VDU command writes a series of bytes to the screen in a similar way to the PRINT command. Thus the following two commands are exactly the same:

(Note the semicolon at the end of the PRINT statement - the VDU command does not send a carriage return unless you explicitly tell it to). Most numbers that you need to write to the VDU are single bytes (characters, for example). However, the graphics coordinates are all double byte quantities and are sent lower byte first, higher byte second. The VDU command enables this to be done easily. If a number in the VDU command is followed by a semicolon, that number is interpreted as a double byte quantity. If you are unsure of bytes and double bytes, the quick rule is that if you are doing a graphics operation using the VDU command, you must always follow a graphics coordinate with a semi-colon. This only applies to the VDU command.

There are two more graphics commands, both of which are done via the VDU command.

Just as text may have a text window defined, outside of which no text command has effect, so a graphics window may be similarly defined. This is done with

where L%, B% and R%,T% are the coordinates of the lower left and upper right pixels inside the window. Setting a window thus prevents any plotting outside it. Also, because CLG is just another plot command, defining a graphics window and going a CLG is a quick way of plotting rectangles.

So far it has been said that the point (0,0) is a the bottom left hand corner of the screen. This point (called the origin) may be specified to live elsewhere with the origin command.

sets the position of the origin on the screen for future graphics commands. Thus to set the origin in the middle of the screen, use VDU 29,640;512;. It does not move the physical position of what is on the screen, the graphics windows or the graphics cursor.

IMAGE MISSING

If VDU 5 is entered, the text and graphics cursors are said to be joined, that is text appears at the graphics cursor which then moves as the text is written. This is mainly used for labelling graphs. The graphics cursor points to the top right pixel of the 8 by 8 character cell to be written, and is moved 8 pixels along by writing letters. This is seen in the following program:

VDU 4 restores the text cursor.

Colours defined through the COLOUR and GCOL commands are more properly referred to as logical colours. When a mode is changed, these logical colours appear as certain physical colours, thus in mode 1, colour 1 is red and colour 2 is yellow. The palette allows this to be changed, thus colour 1 may be made to be blue and colour 2 is yellow. The palette allows this to be changed, thus colour 1 may be made to be blue and colour 2 flashing black on white. To be exact, we must distinguish between two types of colour:

The logical colour is what is output by the COLOUR commands. The maximum logical colour is limited by the number of colours available in the mode.

The physical colour is what appears on the screen. The physical colours and their numbers are listed below:

Each logical colour has a physical colour assigned to it, which may be changed by reprogramming the palette. This is done as follows:

where L% is the logical colour and P% is the physical colour. So in mode 1, to change all the pixels with logical colour 3 (usually white) to blue (logical colour 4), the command VDU 19,3,4,0,0,0 is used. Thus while the very detailed mode 0 is a two colour mode, the colours themselves may be anything available from the palette, such as green on red. Note also that the palette reprogramming is very fast as it does not involve a lot of the screen memory being reprogrammed.