Large Characters

This program displays each printable ASCII character (CHR$ 32 through CHR$ 126) rendered as a large 8×8 pixel grid using a custom UDG character. It reads the ROM character set by locating the font address via system variables at 23606–23607, then uses bit-shifting arithmetic to extract each pixel bit from the font bytes. A single UDG (char 144) is defined with a border/box pattern from the DATA statement at line 100, and its inverse (CHR$ 143, i.e., 144−1) is used to represent set pixels while CHR$ 144 represents clear pixels. The program pauses for a keypress between each character display, stepping through all 96 printable characters before stopping.

---

Program Analysis

Program Structure

The program is organized into three functional phases:

1. Initialization (lines 5–10): Defines UDG ‘A’ (char 144) from inline DATA and locates the ROM character set base address.
2. Display loop (lines 15–65): Iterates over all 96 printable characters (CHR$ 32–126), rendering each as a large pixel grid.
3. Data (line 100): Eight bytes defining a bordered rectangle UDG pattern (255,129,129,129,129,129,129,255).

UDG Setup

Line 5 uses RESTORE followed by a FOR loop to POKE all 8 bytes of the DATA at line 100 into the address returned by USR CHR$ 144 (i.e., USR "\a"), defining UDG ‘A’. The pattern 255,129,129,129,129,129,129,255 creates a hollow box — a full top row, full bottom row, and only the outermost bits set in middle rows.

Font Address Lookup

Line 10 computes the character set base address using system variables:

• PEEK 23606 + 256 * PEEK 23607 reads the two-byte system variable CHARS (address &5C36), which points 256 bytes before the font data.
• Adding 256 corrects for this offset, giving c as the true start of the font bitmap data.

This technique works whether the font is in ROM or has been relocated to RAM.

Pixel Extraction and Rendering

For each character j and each row p (0–7), the program fetches the font byte at c + p + j*8 (line 25). The inner loop (lines 35–50) extracts bits from least-significant to most-significant using integer division:

• LET x = INT(byte/2) — integer right-shift by one.
• LET bit = byte - 2*x — isolates the LSB (0 or 1).
• LET byte = x — advances to the next bit.

The extracted bit is used to select either CHR$ 143 (set pixel, 144−1) or CHR$ 144 (clear pixel, 144−0). Because bits are extracted LSB-first but printed left-to-right starting at column 8 down to column 1 (via AT p, 9-l), the pixel order is naturally reversed back to the correct left-to-right visual orientation.

Display Layout

Screen element	Position
Large pixel grid	Rows 0–7, columns 1–8
Raw byte values	Rows 0–7, column 15
CHR$ label and character	Row 4, column 20
“Press any key” prompt	Row 15, column 0

Notable Techniques

• Using CHR$ (144 - bit) elegantly selects between two UDG states with a single expression rather than an IF branch.
• The PAUSE 0 on line 60 (conditioned on j <> 95) waits for a keypress on every character except the last, avoiding a redundant pause before STOP.
• The font lookup via CHARS system variable is portable and will correctly follow any relocated or custom font.

Potential Anomalies

• Line 60 uses PAUSE 0 alone without reading INKEY$; PAUSE 0 on the Spectrum/TS2068 waits until any key is pressed, so this is valid and sufficient.
• The UDG box pattern is defined but only its code-point relationship (144 - bit) matters; the actual visual appearance of the box/blank pixels is a design choice and not central to correctness.
• Line 10 adds 256 to the CHARS pointer. The CHARS system variable holds a value 256 less than the font start, so this addition is correct and intentional.