T 0158/01 (Video display adjustment/BELISHA) 20-05-2003

1. Amendments

The Board is satisfied that the amendments made to the claims do not contravene Article 123(2),(3) EPC.

2. Clarity

Claims 1 and 7 are regarded as fulfilling the requirements of Article 84 EPC. It may be noted that the use of the word "fairly" ("keeping the absolute vertical size of said visual representations fairly constant across different frequency modes") is allowable since the invention only permits approximate size adjustment and thus cannot be defined more precisely.

3. The prior art

3.1. D6 describes an apparatus for adjusting video display controls in a multi-frequency video display. The display is tuned to the frequency of a horizontal sync signal of a wide variety of video adaptor cards of computer systems (column 1). It has a screen (CRT 68, figure 1A) for displaying information received from these computer systems. The apparatus comprises an input control block (4) for providing user input; a microcontroller (1) capable of receiving said user input from said input control block and controlling the adjustment of the video display controls (column 1, lines 27 to 30); a memory block being capable of storing parameters of said adjusted video display controls, this memory block being electrically connected to said microcontroller (column 1, lines 31 and 32); and a display adjustment block capable of providing the parameters of the adjusted video display controls to the multi- frequency video display in order to set the video display controls, this display adjustment block being coupled to and controlled by said microcontroller (column 1, lines 44 to 47). Thus D6 discloses the features of the preamble of claim 1. There is no on-screen display (OSD).

3.2. D2 describes OSD for adjusting parameters (such as brightness) of a common TV screen (see in particular page 1, lines 48 to 65).

3.3. D9 describes an OSD system for TV which is compatible with different scanning standards (see the abstract). It is explicitly mentioned that the width (ie horizontal size) of the characters displayed can be kept constant over non- standard horizontal frequencies by making the character pixel read-out rate proportional to the horizontal sync frequency (see page 804, section 2.2: "The PLL Time Base"). If the same is achieved for the character height (vertical size), this is not said.

4. Inventive step

4.1. Novelty not being in dispute, only the question of inventive step need be considered. D6 is taken as starting point. The described apparatus, having the features contained in the preamble of claim 1, has no OSD. OSD was however a well known technique at the priority date of the opposed patent, as exemplified by D2, and offered a convenient way of setting parameters. As to combining D6 and D2 the Appellant has pointed out that if an OSD technique would be used for the D6 monitor the displayed characters would in general be distorted because of the varying number of scan lines used in different computer systems. Therefore, in the Appellant's view, it was not obvious to combine the two teachings.

Respondent 02, on the other hand, is of the opinion that the skilled person would have made an effort to overcome any such distortion problem.

4.2. In the Board's view, merely desiring to add OSD to the system known from D6 was an evident aim which cannot be become inventive because of problems appearing when the combination is tried. This is all the more the case since the need to adjust the size of OSD characters to various TV standards (rather than computer standards, but the technique is the same) was not a new problem but had previously been recognised, eg in D9. Starting from D6, the actual technical problem with which the skilled person was faced was therefore to find a way of keeping the size of the OSD characters (more or less) constant.

4.3. The size of a character involves its horizontal and vertical dimensions, but the invention according to claim 1 is only concerned with the vertical size: each pixel line of a character is read out repetitively depending on the received horizontal sync signal. The underlying assumption is that "higher horizontal frequencies indicate an increased vertical resolution" (page 4, lines 22, 23 of the patent specification) . If the resolution is high the number of displayed scan lines is larger than normal and a character stored as a certain number of pixel lines would appear compressed if each pixel line were read out only once.

4.4. The solution as claimed to this problem can, in the view of Respondent 02, be found in D9. According to the Respondent, although this document does not explicitly mention keeping the vertical size of OSD characters constant the skilled person would deduce from figures 5 to 7 and certain passages in the text that in the high definition modes (IDTV, EDTV, and "Japan HDTV") each line of a character matrix is read out twice, thus doubling its height.

The Appellant denies that D9 discloses this.

4.5. The Board first notes that the Respondent's argument based on D9 was presented for the first time at the oral proceedings before the Board. Since the text in D9 does not mention the feature in question explicitly, the circuit diagrams in the drawings are the only possible basis for it. Respondent 02 has referred to three diagrams, each showing one part of the circuit. Any relationship between the signals involved must be deduced from these drawings but since no signal waveforms are given it is not possible to verify the result. Thus the basis for the argument is rather weak. Moreover, the Respondent admitted during the discussion between the parties that his interpretation of D9 relied on there being an error in the text: it is stated at page 805, left column, that "/t/he feedback signal, M15, to the phase comparator is 15 kHz when mode flag is set and changes to 31 kHz when mode flag is reset", but the argument in fact requires the opposite relation. At least at first sight it is not apparent from the associated drawings that the text actually contains this error. The Board therefore cannot accept the conclusion drawn by Respondent 02, namely that the skilled person would necessarily conclude that individual lines of the character matrix are read out repetitively in the high definition TV modes. Nor is it self-evident that every high definition mode requires this measure. As the Appellant has pointed out, at least in the Japanese HDTV (known as MUSE) a frame is split up in four fields, each of which contains roughly the same number of lines as a standard field (eg NTSC). Therefore a displayed OSD character would have the same height in a standard field and a MUSE field, a conclusion which is incidentally supported by figure 8 of D4. Thus, it cannot be determined with a sufficient degree of certainty that the skilled person would regard D9 as involving the problem to be solved or leading him to its solution.

4.6. Respondent 02 has relied solely on D9 to demonstrate that the last feature of claim 1 is obvious, and indeed no other cited document appears to disclose more relevant art. Thus, the subject-matter of claim 1 involves an inventive step. For analogous reasons also the method defined in claim 7 involves an inventive step.

5. It should be stressed that the fact that the Board does not accept the Respondent's argument in this case does not mean that drawings of electrical circuits in the prior art would be of minor relevance. On the contrary, the skilled person in the field of electronics would always be likely to study circuit diagrams closely. But clearly caution is necessary whenever information which can allegedly be drawn from diagrams goes beyond - or even contradicts - the accompanying text. $

It may also be added that although there is usually no objection against a new line of argument based on previously cited prior art being developed at oral proceedings (cf G 4/92, headnote II, EPO OJ 1994,149 in this respect) it is advisable to communicate such reasoning beforehand if it is complex or relies on assumptions. This would have been possible in the present case since the essential features of the claim under consideration were known for almost two months before the oral proceedings. To do so may also be in the interest of the party presenting the new reasoning since unexpected arguments which are difficult to assess at short notice may be met by equally unexpected counter-arguments.

6. Although the independent claims are regarded as fulfilling the requirements of the EPC some of the dependent claims - which are still those of the patent as granted - may need revision (eg claims 3 and 9). Furthermore, the technical problem which is solved by the now considerably limited claim 1 does not become clear from the introduction to the description, which is mostly about adjusting video display controls (cf Rule 27(1)(c) EPC). The case is therefore remitted to the first instance for examination of any necessary modifications of the dependent claims and the description.