EP0132454A1 - Method and device for the high definition typographical display of text - Google Patents

Abstract

The invention relates to a method for the high-definition typographical display of text, in which the picture elements which are to be displayed for the letters are stored in the form of half-tone matrices and recorded as half tones.

Description

Process and device for high-quality typographic representation of fonts

Technical field

The invention relates to the field of computer-controlled rendering of fonts.

Underlying state of the art

The computer-controlled rendering of fonts is used today for the dialogue between humans and computers and is used to record, check, change and reproduce data. For this purpose, screen terminals are used, which essentially consist of the following components: alphanumeric keyboard, computer interface, letter memory, display and control electronics and picture tube. The letter memory consists, for. B. from PROM memories (Programmable Read Only Memory), the image information for the representation of each character, for. B. letters, numbers or punctuation marks etc. on the screen.

The display on the picture tube is conventionally, typically with around 460 pixels per line and 240 picture lines. Therefore, up to 80 letters with a pixel space between successive characters can be displayed in one line - characters - 80 x (7 + 1) = 640 - and up to 20 lines - 20 x (9 + 3) = 240 - with 3 image lines each Line penetration can be recorded. The height of the capital letters is therefore usually shown with 3 - 4 mm.

Usually, matrices of 7 x 9 bits are stored for each letter and displayed on the screen. Each bit corresponds to a pixel, which can be either black or white. In this way it is not possible to reproduce different fonts with typographic subtleties and the legibility of some letters is poor. So z. B. the letter "w", if it is to be reproduced in a font in which it is provided with serifs, is not readily recognizable as such, while in the representation without serifs the white portion of some letters, especially the letter "I" is too big.

Disclosure of the invention

The object of the invention is therefore to improve the legibility of the fonts and to better express the typographical features. The invention achieves this in that the pixels to be displayed for the letters are stored in the form of gray value matrices by means of multi-bit bytes and recorded as gray values.

An advantageous development of the invention consists in that the gray value for each pixel corresponds to the gray value of a rectangle of m points lying one behind the other in a line and n successive lines, preferably m = _. 4 and n = 4 can be selected.

Another embodiment of the invention is that a multi-bit key of half Byte is used.

Advantageously, the gray value matrices all have the same height, preferably 16 gray value points and different widths, preferably 3-16 gray values, adapted to the width of the characters to be displayed.

According to the invention, a picture tube with more than 512 x 512 pixels, preferably 1024 lines of 1024 pixels or more without field separation can be provided.

The inventive device for implementing the method may be characterized preferably in that a bit _- Map of the presentation of each screen character storage for storing character with at least n times as many rows and at least m times as much points per line as indicated on the Screen can be displayed, is provided and is read out n-line by line and summarized in m-pixel width by means of a computer, the corresponding gray value being determined for each mxn pixel and this information as multi-bit byte in the form of gray value matrices for the characters is storable. Here, preferably m = 4 and n = 4 and a multi-bit key of half a byte can be provided.

Brief description of the drawings

• Fig. la bis lf Darstellungen von Zeichen ohne Grauwerte (Stand der Technik),
• Fig. 2a und 2b Darstellungen von Zeichen mit Grauwerten,
• Fig. 3 ein Beispiel für die Darstellung eines Zeichens in einer Schriftart (Optima),
• Fig. 4 die Darstellung einer Schrift mit einer Bit-Matrix von 64 x 64 Bildpunkten (Stand der Technik),
• Fig. 5 die Darstellung dünner Striche mit Grauwerten
• Fig. 6a bis 6c ein Beispiel für die bessere Lesbarkeit der Schrift bei der gegenüberstellenden Wiedergabe mit und ohne Grauwerten,
• Fig. 7 eine Einrichtung zur Durchführung der Erfindung
• Fig. 8 ein Diagramm für die Übergabe der Halbbytes an den Video-Controller
• Fig. 9 ein Beispiel für eine Schnittstelle zwischen Bus und Halbbyte-Speicher und Video-Controller und
• Fig. 10a u. 10b ein Schaltungsbeispiel für die Übergabe vom Halbbyte-Speicher zum Monitor

The invention is explained below with reference to FIGS. 1-10.
Show it:

• La to lf representations of characters without gray values (prior art),
• 2a and 2b representations of characters with gray values,
• 3 shows an example of the representation of a character in a font (Optima),
• 4 the representation of a font with a bit matrix of 64 x 64 pixels (prior art),
• 5 shows the representation of thin lines with gray values
• 6a to 6c an example of the better legibility of the font in the contrasting display with and without gray values,
• Fig. 7 shows a device for carrying out the invention
• Fig. 8 is a diagram for the transfer of nibbles to the video controller
• Fig. 9 shows an example of an interface between bus and nibble memory and video controller and
• 10a u. 10b shows a circuit example for the transfer from nibble memory to the monitor

Best We ^g for practicing the invention

In the figures la to lf the letters I, M and W are shown in the form of pure black and white values once with and once without serifs. The letter I in FIGS. 1 a and 1d provides legibility in both cases, but not in the case of the two more complicated letters M and W in FIGS. 1b and le and 1c and 1f.

In Figures 2a and 2b, the letter M and the small dot are shown with gray values. The height of the matrices is constant, e.g. B. with 16 gray dots, and the width variable, z. B. with 16 gray dots, depending on how the width of the respective character requires, about 3 for the small dot and 16 for the small M.

In order to optimally enable the display, a different picture tube than was previously used for screen terminals should be used, namely a picture tube with more than 512 x 512 pixels without field separation. Particularly preferred is a picture tube with 1024 lines of 1024 pixels or more without field separation, e.g. B. to represent the area of a DIN A4 page of about 200 mm wide and 300 mm high. In this case, the line width is about 0.3 mm and the height of the capital letters is 3.3 mm.

• _Man wählt Rechtecke von m in der Zeile hintereinanderfolgenden Punkten Länge und n aufeinanderfolgenden Zeilenhöhe aus, errechnet - zweckmäßig mit Hilfe eines Computers - aus den m x n Bits einen durchschnittlichen Grauwert, entsprechend dem Verhältnis der Anzahl schwarzer Punkte zur Gesamtpunktezahl, und speichert ihn in einem Mehr-Bit-Schlüssel. Jedes der ausgewählten Rechtecke erscheint demnach bei der Bildwiedergabe auf dem Bildschirm als einziger Bildpunkt mit einem Grauton, der dem Durchschnitt der Schwarz-Weiß-Werte des Gesamtrechtecks entspricht im Schwerpunkt dieses Rechtecks.

The optimum choice of the most favorable gray value for each image point for the purpose according to the invention can be determined according to the invention by firstly creating a square for the letter finer black and white bit map created than the pixel grid shown on the screen corresponds, and preferably with 64 x 64 pixels. The gray scale matrices that are to be loaded into the PROM memory of the screen terminal are determined as follows - expediently with the aid of a computer:

• _M an selects rectangles of m in the row of successive points length and n successive row height, calculates - expediently with the help of a computer - an average gray value from the mxn bits, corresponding to the ratio of the number of black points to the total number of points, and saves it in a more Bit key. Each of the selected rectangles therefore appears when the image is displayed on the screen as the only pixel with a shade of gray that corresponds to the average of the black-and-white values of the overall rectangle in the focus of this rectangle.

Preferably, m = 4 and n = 4 and the multi-bit key should be 1/2 byte (4 bits). However, a differently determined gray value can also be used to achieve a more favorable gradation. An advantage of the method according to the invention is that the optical control of text entries for typesetting computers is made possible, which results from the possibility of variable letter width and variable font.

An example of this is given in FIG. 3 and FIG. 4 shows the representation of this character according to the prior art with a 64 x 64 bit matrix per letter. As can be seen from this, the quality is worse than in the representation with gray values according to FIG. 3.

As can be seen from FIG. 5, with the help of the gray values used in accordance with the invention, the letter lines can get a desired thinning with lighter gray points in the middle.

The mode of operation of the method according to the invention can be seen particularly well from FIGS. 6a to 6c by means of the letter "e". Fig. 6a shows the bit map for this letter with the z. B. in newspaper printing, the usual number of bits of 34 x 34 bits per 9p letters, with the "over pixels" combined according to the invention to form rectangles (in the example shown, 4 x 4 pixels). The number of black bits contained in each of these rectangles is indicated. FIG. 6c shows how this letter would look like a quarter with 8.5 x 8.5 pixels per letter in pure black and white reproduction (black the rectangles with at least 7 and white, those with 6 or less black bits like rectangle) - namely completely illegible.

In comparison to this, it can be seen from FIG. 6b that the same letter with the same number of pixels per letter is easily readable when the gray values shown are suitably graded according to the invention with pixels or "over-pixels". This is particularly noticeable when the letter is the correct size reduced by about 3 mm or, which amounts to the same thing, from a greater distance, e.g. B. 4 m is considered.

As can be seen from the foregoing, the method and device according to the invention can achieve a number of advantages over known terminals, e.g. B. a high-quality rendering of fonts, thus a significant improvement in legibility - that is, ergonomically cheaper -, the reproduction of typography and the choice of various fonts as well as the possibility to save characters in a much nicer form than before and display them on the screen.

In Figure 7 (1) is a bus line via which the modules (2 to 7) can communicate with each other, for. B. a commercially available VME bus. A font memory (2) is provided in which the bit matrices for the individual characters are stored permanently, preferably on magnetic disks (floppy disks). The sum of all characters of a certain font forms a so-called font. Usually, a bit matrix of 34 x 34 bits is provided for a font that is to be displayed in 9p font size, cf. Fig. 1, in which the number of black pixels per 4 x 4 bit square are also entered. The text to be set is from a sentence calculator (3), z. B. a computer of the type MC 68000 from Motorola, built in that it the bit maps of the individual letters in the correct geometric location in the bit map memory copied. It expediently does not fetch these single-bit maps directly from the font memory (2) each time, but first loads the entire font into an intermediate memory (4), which is advantageously a RAM memory (random access memory). As memory modules for this buffer z. B. FORCE boards can be used. According to the described preferred embodiment of the invention, the text commands to be processed by the typesetting computer (3) contain, in addition to the text, information about the thickness of each letter, ie the exact distance in bits, which it has before with a small white area (width) and a small white area afterwards. In this way, the sentence computer (3) can determine the most advantageous starting position for the bit map of the respective subsequent letter from the information of the character just set.

Independent of the sentence computer (3), the partial byte computer (6) used according to the invention and also connected to the BUS (1) works. In the preferred embodiment described, it is a nibble byte computer, eg. B. a FORCE CPU 68000 Sys 68 K / CPU-1.

It reads sequentially from left to right lines, each preferably 4 bits high, and forms a nibble from 4 x 4 bits, which is loaded into the corresponding position of the nibble memory (7) (a in FIG. 2). The computer (6) proceeds line by line and forms the first 1 x 1000 nibbles of the memory (7) from the first 4 x 4000 bits of the bit map (5) and likewise from the second 4 x 4000 bits the second 1 x 1000 nibbles, etc. Independent of both the sentence processor (3) and the nibble processor (6), the video controller only connects to the nibble memory (7), but not to the BUS line (1) (8) from the nibble memory (7) line by line 1000 nibbles each. From these, the television signals for a television line are generated via the digital-to-analog converter (9) required for the screen input and are offered to the electronics of the monitor (10).

In order to ensure a perfect continuous image display on the monitor, it is expedient in the example described to use a video controller input with a bandwidth of at least 64 x 10 ⁶ nibbles / sec.

The transfer of the nibble information from the nibble memory (7) to the monitor (10) is shown in more detail in FIGS. 7 to 10.

The video controller (8) requests successively from the nibble memory (7) packets of 8 nibbles = 32 bits of certain addresses, z. B. the 15th package of the 27th line.

The timely completion of these orders can be seen in FIG. 8, in which the top line represents the transfer cycle, which can be, for example, 125 ns. The successive order numbers are given in the second line.

Of the lines a _l , a ₂ , a ₃ ..... the left starting point represents the beginning of a request and the right end point represents the completion of the transfer of the requested nibble packet from the memory (7) to the video controller 8. Each request is only accepted when the memory (7) has a positive edge of the transfer clock (point C ₁ , C ₂ , C ₃ ..... see also top row).

After completion of the processing (points D ₁ , D ₂ , D ₃ ), the requested nibble packet is transferred to the video controller (8) (transfer points F ₁ , F ₂ , F ₃ ... ₎ , ₍ end points E ₁ , E2, E3 ^... ) ^.

As can be seen from FIG. 8, it is provided according to the preferred embodiment of the invention shown that several partially overlapping jobs - in the example shown 4 - are taken over and executed at the same time, as a result of which the transmission speed is considerably accelerated.

Figure 9 shows a circuit suitable for this purpose. It is a memory into and from which data can be transmitted asynchronously in both ways. The block circuit enables multiple accesses, four times in the example shown. ,

It is connected to the bus line (1) via the interface a and to the video controller (8) via the interface b.

In the event of a memory access by the video controller, an address sent by an address generator, which will be shown later in FIGS. 10a and b, under which a 32-bit word is to be read, passes via the transfer interface b, which is denoted by (34) is in one of the 4 address / data registers (23), (26), (29), (32). Which one exactly depends on the address. With consecutive addresses, the registers (23), (26), (29), (32) are addressed in succession.

Each of the 4 registers is connected to a memory matrix (24), (27), (30), (33) with a capacity of 64 K 32 bit words and a memory controller (124, 127, 130, 133). The memory controller, which ensures the correct timing for the memory modules, is known per se and is made up of standard TTL modules. The memory matrix is made up of 64 K-bit dynamic RAM memories (e.g. _M ostek _M K 4564). After the access time of the memory matrix (24), (27), (30), (33) has expired, the data are available at the output of the matrix and are written into the address / data registers (23), (26), (29), (32 ) loaded. Via the transfer interface (34), they arrive in a nibble byte catch register of the video controller (8). The access of the nibble computer (6) is analogous to the memory access of the video controller.

The connection to the bus (1) via the interface a takes place via the registers (22), (25), (28) and (31).

The transfer from the nibble memory (7) to the monitor (10) is shown in greater detail in FIGS. 10a and 10b. An address generator (35) is provided (e.g. AMD AM 2932), which is controlled via a basic clock d. This clock is supplied by the synchronization generator (20) and is identical to the basic clock shown in FIG. 3, ie the transfer clock. The address generator (35) via gives the current address of the nibble packet to be called to an address register (36) (e.g. TI SN 74LS374), which takes over the address with the system clock and makes it available at the output. The address reaches the nibble memory (7) via a bus driver (21).

The data supplied by the nibble memory arrives in the nibble byte register (12) via a selector (11), where it is accepted with the system clock c.

A nibble packet consisting of 8 nibbles is available at the output of the synchronous register (12). The individual nibbles are selected in sequence via a 1 out of 8 selector (15) and passed on to a nibble representation register (18), into which they are adopted with the system clock c. The 1 out of 8 selector (15) has 8 inputs and a control unit (17). The control unit (17) consists of a register (171) which is controlled by the system clock c. The register (171) is connected to a selector (172) which is controlled by the basic clock d, which, like the clock c, is supplied by the synchronization generator (20). The clock c is 8 times faster than the clock d since the 1 out of 8 selector (15) is queried 8 times in a clock period of d. The selector (172) is also connected to a step-up circuit (173) and to a zero generator (174), which is also driven by the clock d. When a new nibble packet is loaded into the synchronous register (14), the control unit (17) is reset via the control signal d, so that the first nibble via the 1 out of 8 selector (15) to a nibble representation register (18). The system clock c increases the value of the control unit by 1, and the next nibble reaches the nibble representation register (18) via the 1 from the B selector (15). After increasing the value of the control unit (17) seven times, the last nibble of the nibble packet is switched through to the nibble representation register (18). A new output cycle of the nibbles then begins with the resetting of the control unit (17) via the control line d.

The nibble in the nibble representation register (18) is converted into an analog voltage signal via the digital / analog converter (19) (e.g. Analogic MP 8318) and sent to the monitor (10) as a video signal in the video input.

A known and commercially available synchronization generator (10), which can be constructed discretely from standard TTL modules, supplies the necessary synchronization signals (horizontal and vertical synchronization signals) for operating the monitor (10) as well as the basic clock d and the 8 times higher clock c.

Claims (12)

1. A method for typographically high-quality representation of fonts, characterized in that the pixels to be displayed for the letters are stored in the form of gray-scale matrices and recorded as gray-scale values. 2. The method according to claim 1, characterized in that the pixels for the letters are stored in the form of gray-scale matrices by multi-bit bytes in PROMs. 3. The method according to claim 1 or 2, characterized in that the gray scale value for each pixel corresponds to the gray scale value of a rectangle of m in each row points in a row and of n consecutive lines. 4. The method according to any one of claims 1-3, characterized in that m = 4. 5. The method according to any one of claims 1-4, characterized in that n = 4. 6. The method according to any one of claims l _- 5, characterized in that the multi-bit key is 1/2 wide. 7. The method according to any one of claims 1-6, characterized in that the gray scale matrices have the same height, preferably 16 gray scale points and have different widths of preferably 3-16 gray value points adapted to the width of the characters to be displayed. 8. The method according to any one of claims 1-7, characterized in that a picture tube with more than 512 x 512 pixels is used, preferably 1024 lines of 1024 pixels or more without field separation. 9. Device for performing the method according to one of claims 1-8, characterized in that a bit map of the template each character to be stored in the character memory at least n times as many lines and at least m times as many points per line as it can be displayed on the screen, it is provided that the bit map can be read out n-line by line by a computer and summarized in m-pixel width, the corresponding gray value being determined for each mxn pixel and this information as a multi-bit byte in the form of Grayscale matrices for the characters are saved. 10. Device according to claim 9, characterized in that m = 4. 11. The device according to claim 9 or 10, characterized in that n = 4. 12. Device according to one of claims 9 - 11, characterized in that the multi-bit key is 1/2 wide.

Images