Minnesota Mining and Manufacturing Company v Fuji Photo Film Co Ltd
[1994] APO 56
•23 September 1994
official notice
decision of a delegate of the commissioner of patents
Application : No. 574479 in the name of Minnesota Mining and Manufacturing Company
Title: Continuous Tone Recording System
Action: Opposition Under Section 59(1) of the Patents Act 1952 by Fuji Photo Film Co,. Ltd.
Decision: Issued .
Abstract: Consideration of "definition by result", novelty, inventive step, sufficiency of description, clarity and fair basis of claims. Claims amended between opposition and hearing. Opposition to amended claims dismissed. Part costs awarded to opponent.
PATENTS ACT 1990
DECISION OF A DELEGATE OF THE COMMISSIONER OF PATENTS
Patent Application Number 574479 by MINNESOTA MINING AND MANUFACTURING COMPANY and Opposition Under Section 59(1) of the Patents Act 1952 by FUJI PHOTO FILM CO LTD
BACKGROUND
Patent application 574479 in the name of Minnesota Mining and Manufacturing Company (3M) was filed under application number 44789/85 on 11 July 1985. It claims convention priority of 27 August 1984 from an application made in the United States of America. It was advertised accepted on 1 July 1988. Fuji Photo Film Co Ltd (Fuji) filed a notice of opposition on 5 October 1988 under section 59 of the Patents Act 1952.
I heard the opposition in Sydney on 5 April 1993. The opponent (Fuji) was represented by Mr David Yates and Mr Stephen Burley, instructed by Mr Chris O'Sullivan, patent attorney, of Griffith Hack & Co, Sydney. The applicant (3M) was represented by Mr David Catterns QC, instructed by Mr John Hinde and Mr Lee Pippard, patent attorneys, of Spruson & Ferguson, Sydney. The attorneys for 3M were assisted by Ms Lorraine Sherman, a patent attorney registered in the United States, Mr Gary Ashton, one of the inventors, Mr Paul Schubert and Mr Sid Schrtzfield.
Although the notice of opposition recites all the grounds set out in sub-section 59(1) of the 1952 act, the submissions at the hearing concentrated on issues of lack of novelty, lack of inventive step, and ambiguity and lack of fair basis in the claims.
SUMMARY OF ISSUES
This opposition is about a high speed laser printer, which is also called a laser "imager" in the evidence. As the specification states,
"The present invention relates to a system and method for providing photographic quality continuous tone images from a digital data base using a laser diode light source. The images are useful, for example, in medical electronic imaging systems, high quality facsimile, and in the graphic arts."
The only independent claim in the specification recites a combination of five integers, referred to in the evidence and at the hearing as integers (a) to (e). The last six lines of the claim then characterise or define the capabilities of that combination. The opponent argues that these last six lines merely state a problem in the form of desiderata to be met without defining how those desiderata could be met. The applicant argues that these last six lines of the claim are a legitimate form of "defining by result".
The opponent also argues that integers (a) to (e) were known, both separately and in combination, in the prior art. In particular, the opponent argues that the combination of all of integers (a) to (e) was disclosed in Japanese patent specification number 58(1983)-106953 by Ricoh K K ("the Ricoh specification"). This was published in Japanese at the Patent Office library, Canberra, on 2 August 1983. This is before the priority date of the claims of the opposed specification.
THE SPECIFICATION
The specification was amended after acceptance. It currently has one independent claim, 16 dependent claims, and an "omnibus" claim. It is entitled "Continuous Tone Recording System."
The specification says that:
"The present invention relates to a system and method for providing photographic quality continuous tone images from a digital data base using a laser diode light source. The images are useful, for example, in medical electronic imaging systems, high quality facsimile, and in the graphic arts."
Under the heading of "Background of the Invention" the specification says:
"Optical scanning and recording systems in which lasers are used for exposing images from digital data stored in computer memory onto plates or film are known. An externally modulated beam is used to write information on a photosensitive medium. These systems include office printers for text, rotating drum scanners for color separation for printing, and flying spot scanners using gas lasers for newspaper plate making. The capability of these systems to reproduce photographic images is limited to half-tone methods of information recording."
The specification then refers to some prior art documents, followed by a "Summary of the Invention" which is in the same terms as claim 1. That claim is as follows:
"1. An image recording system comprising in combination:
a. light means including a solid state laser diode for generating an analog signal-controlled, intensity-modulated beam of light;
b. a feedback circuit responsive to the instantaneous intensity of said beam of light for continuously providing a feedback control signal coupled to said light means for adjusting said laser diode to stabilize its light output, allowing virtually instantaneous changes in output intensity in response to an input signal while maintaining the output constant regardless of changes in the characteristics of said laser diode;
c. means for at least one of collimating, focusing, and controlling the path of said beam of light;
d. means for scanning said beam of light across the width of an image plane at which a photosensitive recording medium having a sensitivity and exposure range matched to the intensity-modulated beam of light may be positioned, said means providing scan lines on said recording medium,
e. means for positioning said scan lines consecutively and accurately along the length of said image plane;
said combination being adapted to enable the formation at the image plane of a continuous tone image having at least 64 gray levels when imaged onto a transparent medium or at least 32 gray levels when imaged onto an opaque medium, said scan lines being placed with an accuracy having a standard deviation of less than 2 percent and said scan lines being placed at a rate not less than 80 lines per second."
ISSUES TO BE DECIDED
The opposition raises the following issues of fact or of law.
Do the claims adequately define the invention by result?
Which, if any, of integers (a) to (e) were known before the priority date?
What, if any, combinations of the integers (a) to (e) were known before the priority date?
Does the specification fully describe the invention?
Is claim 1 novel and inventive?
These issues are inter-related. The determination of whether or not there is adequate "definition by result", requires a consideration of whether a person skilled in the art could interpret the claim. This involves a consideration of the state of the prior art base and of common general knowledge. I also have to consider the state of the prior art base and of common general knowledge in deciding whether the invention is novel, non-obvious, and fully described in the specification.
These issues raise some quite complex questions of fact. To deal with these questions adequately, I have found it necessary to set out various submissions of counsel, and passages from the evidence, in more detail than is usual in a Patent Office decision. This has the unfortunate effect of making the decision less "readable" than I would like. However, readers who are not interested in the details of the questions of fact should be able to comprehend my findings without considering my quotations from the submissions and from the evidence in detail.
DO THE CLAIMS ADEQUATELY DEFINE THE INVENTION BY RESULT?
Law on Defining an Invention by Result
The most recent Australian authority on "defining by result" is the decision of Mr Justice Gummow in the Federal Court of Australia in Rescare Ltd v Anaesthetic Supplies Pty Ltd, (1992) 25 IPR 119. That decision includes some observations which are appropriate to the facts of the present case.
At page 126 lines 1 to 9, his honour approved the following quotation from the fifth edition of Patents for Inventions by Mr Blanco White QC:
"To amount to a limitation by result, what is in the claim must at least be a limitation; something that draws a line between two classes of things that would otherwise fall within the claim; with the implication that conditions of the manufacture can be adjusted, by the reader of the specification, to secure the specified result. It is, of course, a matter of construction to determine whether the words in the claim effect a limitation or merely assert that complying with the claim will secure a certain result, and this like other questions of construction affecting validity is likely in present-day conditions to be decided in favour of an otherwise meritorious patentee."
His honour also made the following observations, which should be kept in mind when considering a case of "defining by result":
[at page 127, lines 24 to 28] ".. at various stages of its submissions upon infringement the respondent tended to confound the function of the claims (to define the area claimed for the monopoly) with that of the body (the teaching of the performance and putting into effect of the invention)"; and
[at page 143, lines 27 to 30] "However, it must be borne in mind that it is not the task of the claims to teach the application of the invention: the task of the claims is to define the invention and mark out the area of the monopoly."
My understanding of these principles, restated in terms more directly relevant to the facts of the present case is as follows.
·It may be necessary to distinguish between (a) a claim which "defines by result"; and (b) a claim which, for example, recites a list of integers and asserts that the combination of integers invariably has some inherent result.
·A claim which "defines by result" recites some limitation (by result) which draws a line between two classes of things.
·It is not the function of the claim, in a situation of "defining by result", to instruct the reader how to go about achieving that result. This is the function of the rest of the specification.
[Although this decision was appealed to the full federal court, the decision on that appeal does not affect these conclusions.]
As a matter of interpretation, the present claim 1 is not a claim which "recites a list of integers" and "asserts that the combination of integers invariably has some inherent result". The concluding lines of the claim specifically recite "said combination [of integers (a) to (e)] being adapted to enable the formation at the image plane of a continuous tone image ... [of a certain quality]." [My emphasis.] The claim does not assert that all possible combinations of integers (a) to (e) will necessarily have the property of enabling the formation of an image of the recited quality. Rather, it asserts that to infringe the claim, it is necessary to adapt those integers so that they can form an image of the recited quality. (I note in passing that some of paragraphs (a) to (e) of the claim may need further interpretation to decide whether or not they contribute to a limitation of the claim by result. For example, paragraph (e) recites "a feedback circuit ... for adjusting said laser diode to stabilize its light output, allowing virtually instantaneous changes in output intensity ... maintaining the output constant ..." [my emphasis].)
This conclusion does not answer the other issues set out in my summary above. It is still necessary to decide whether the concluding lines of the claim set out a "result" which is sufficiently precise to "draw a line between two classes of things". It is also necessary to decide whether the description sufficiently instructs the reader in how to go about achieving that result.
Drawing "a line between two classes of things"
Most of the "limitation by result" is in the concluding lines of the claim, which read as follows:
"said combination being adapted to enable the formation at the image plane of a continuous tone image having at least 64 gray levels when imaged onto a transparent medium or at least 32 gray levels when imaged onto an opaque medium, said scan lines being placed with an accuracy having a standard deviation of less than 2 percent and said scan lines being placed at a rate not less than 80 lines per second."
At the hearing, Mr Yates made several submissions concerning the clarity of this passage from claim 1. When he was specifically addressing the clarity of the expression "continuous tone image" in claim 1, Mr Yates said that:
"It is then said that the apparatus has to form at the image plane a continuous tone image. Now that limitation we submit is completely subjective. It is recognized to be so by Professor Orr on page 14 of his declaration where he seeks to make certain comments on paragraph 10 of the Sceats declaration. We submit the problem is not overcome by gray levels as being defined. As amended that definition now reads at page 5 lines 19 to 23 `gray levels means controlled variations in density of an exposed recording medium which can be distinguished by measurement with a Tobias Model TBX diffuse densitometer of sensitivity capable of measuring steps of 0.01 optical density units, that being the threshold of average human eye perception.' Now we submit that definition does not remove the subjectivity which is inherent in the claims. It is one thing to state that gray levels are to be distinguished by stated sensitivity in the definition by reference to the densitometer but that tells you nothing about whether a continuous tone as defined results. A continuous tone is defined on page 5 lines 10 and following to mean gradations in intensity in levels of gray so that the transition appears relatively smooth. Now one asks relative to what? Clearly that introduces a significant subjective element which no doubt will be open to debate and the claim accordingly lacks clarity in that connection."
Mr Yates drew an analogy with the expression "white or gray or some shade similar thereto" which was in issue before the Federal Court of Australia in Australian Solar Mesh Sales v Tomlin Industries, (1991) 21 IPR 447 at 455, and said:
"Now His Honour held that the claim was bad for lack of clarity because of the subjectivity introduced by that feature of the claim and here there is similar subjectivity by saying that the image has to be a continuous tone image, that is one which appears to be relatively smooth. One can no doubt well anticipate the kind of debate that that would provoke. If one had an image recording system and it could be determined [that it had the defined features] of the accuracy and speed it would still be open to debate as to the relative smoothness of the image. And we say that the claim therefore lacks clarity in that respect."
In these submissions, Mr Yates referred to the declarations of Dr Sceats and Professor Orr. Dr Sceats is a declarant for the opponent. Paragraph 10 of his declaration, referred to by Mr Yates, is as follows:
"10. It is clear to me that the Ricoh specification is concerned with the accurate reproduction of images, and the apparatus it describes would be capable of producing a high quality image in which the gray level appears to the eye to change smoothly, and which has a large number of measurable gray levels. There is no reason to believe that the apparatus described in the Ricoh specification would not give a high quality image with as many gray levels as the apparatus described by the 3M specification; and every reason to believe that the 3M apparatus would be less accurate through neglect of the non-linearity addressed by the Ricoh specification."
Professor Orr is a declarant for the applicant. Page 14 of his declaration, referred to by Mr Yates, reads as follows:
"Para 10 - The Ricoh specification, of itself, does not give sufficient information to enable the production of high quality images in which the gray levels appear to the eye to change smoothly. This would require, for example, the 2% accuracy feature now specified in amended claim 1 and in no way suggested by Ricoh. The Ricoh gamma-correction factor was obviously of importance in conjunction with a photoconductor medium although of less importance in respect of silver halide film. There are many reasons to believe the Ricoh specification should not give a high quality image. It does not require 2% accuracy of line placement; it is also open to doubt as to whether the Ricoh circuitry would provide sufficiently rapid correction to the intensity of light generated by the laser diode to overcome deficiencies of the laser diodes of that time (including use of the back facet of the diode for emission purposes)."
This passage from Professor Orr's declaration does not seem to me to support Mr Yates' submission. That is, that Professor Orr recognized that the limitation of "continuous tone image" is completely subjective. It may be that Mr Yates intended a reference to page 10 of Professor Orr's declaration (rather than a cross-reference to paragraph 10 of the Sceats declaration), where Professor Orr refers to paragraph 6 of Dr Paul's declaration. Here, Professor Orr states:
"The definition of "continuous tone" in the 3M specification would be perfectly clear to the skilled addressee, such as a radiologist. The comparison, in the art, is with standard photographic x-ray films viewed on a standard lightbox. Those are "continuous tone" images such as are required to be produced by the apparatus of the current application."
In addition to the declarations which Mr Yates cited, there are other declarations which give evidence on this point.
Paragraph 6 of the declaration by Dr George Lang Paul states:
"Page 4 of the 3M specification defines "continuous tone" (line 9) and I note this definition requires a subjective relative evaluation of the appearance of the transition between levels of gray; this is impossible to determine with any degree of accuracy since the discriminating power of the eye to gray level transitions varies from person to person and depends on the ambient light conditions."
Paragraph 12 of the declaration by Mr Philip Emanuel Ciddor gives similar evidence, namely:
"Page 4 of the 3M specification defines "continuous tone" to mean "gradations in intensity of levels of gray so that the transition appears relatively smooth" (lines 9 to 11), and I note this definition requires a subjective evaluation of the appearance of the transition between levels of gray; this is impossible to determine with any degree of accuracy since the discriminating power of the eye to gray level transitions varies from person to person. The definition is also inadequate because it requires a relative assessment of the smoothness of the transition while no indication is given of what the smoothness should be compared to."
Paragraph 4 on page 2 of the declaration by Dr Philip Anthony Wilksch is also material. It states:
"The 3M specification defines "continuous tone" to mean "gradations in intensity of levels of grey so that the transition appears relatively smooth." I understand this term, as would any worker skilled in this field, to refer to images in which levels of light intensity in the original image are represented by corresponding fine gradations in density within the recording medium, rather than by small-scale arrangements of elements with a very limited range of densities, as in half-tone or binary images. In the latter cases, gradations in density between different tones are immediately apparent, at least on a microscopic scale."
A patent specification, including the claims, must be interpreted from the point of view of a person skilled in the art. See, for example, the decision of Lloyd-Jacob J in Barking Brassware Co Ltd v Allied Ironfounders Ltd, [1962] RPC 210 at page 217:
"Argument at the hearing was directed to three expressions appearing in the specification and claims, namely, `fluid mixing apparatus,' `co-axial members' and `well-like recess.' In approaching the resolution of the conflict between the contentions of the parties, it must be stated at the outset that the conception of the court as a depository of dictionaries and encyclopaedias for the elucidation of technical terms is wholly erroneous. Reference to such works may be and frequently is of assistance in assessing the comprehensiveness and reliability of the evidence of witnesses of opinion, but the primary duty of the court is to identify itself with the knowledgeable persons in the particular art to whom the specification is deemed to be addressed, to seek to understand the terms used in the specification in the sense that he and his kind would attribute to them, and thus to arrive at an ascertainment of the alleged invention as it would present itself to a competent potential user of it. This is not to decry the merit of intellectual criticism of language, but merely to mark the bounds within which it will be found of advantage." [My emphasis added.]
The issue of whether or not an image amounts to an infringing "continuous tone image" is an issue which would have to be determined by a court from the point of view of "knowledgeable persons in the particular art to whom the specification is deemed to be addressed".
The issues in this opposition differ in both fact and degree from the issues which were before the court in Australian Solar Mesh Sales v Tomlin Industries, cited by Mr Yates. The relevant facts of that decision are set out at 21 IPR 455, where Wilcox J said:
"Counsel for Tomlin argue that the reference to colours is ambiguous: `white or grey or a shade very similar thereto.' They ask whether the concluding reference to a similar shade qualifies both white and grey, or merely grey. And what is `very similar' to grey? Or does it refer to something which is not grey but which has some tonal similarity, such as blue? I think that there is substance in counsel's point about the ambiguity of the colour specification."
The relevant art in that case, mesh screening, may well have been so straightforward that the court did not need expert evidence in order to determine that "white or grey or a shade very similar thereto" was irresolvably ambiguous. If the same expression was being interpreted in the context of a more complex art, such as colorimetry, colour photography, or colour television, I can speculate that there would be some expert evidence to the effect that, in that art, "white or grey or a shade very similar thereto" meant "an unsaturated colour".
The state of the evidence in the present case is such that I cannot conclude that persons skilled in the art could not say that a particular image was or was not a "continuous tone image". To quote Mr Catterns' submissions:
"Now the next criticism on the ground of sufficiency or perhaps ambiguity is the idea of the continuous tone image being subjective. Now it is my submission virtually every patent would have subjective terms like "substantially inextensible" 3M, or "continuous tone image". Just because it requires a human being to perceive it, doesn't mean it's not sufficiently certain to provide a workable test to those skilled in the art. And I submit that it is clear on the evidence that on both sides the witnesses had no trouble at all in knowing what "continuous tone image" meant. It is something that is perceived by humans but it is sufficiently certain to be part of our claim. Whereas to say "white or gray or similar thereto" as in the Solar Mesh case just really leaves you every colour known to man. It's a completely different thing from a scientific term relating to human perception."
Professor Orr's evidence may well establish that a skilled radiologist, viewing an image under standardised conditions, could judge whether an image was a "continuous tone image". The other evidence may establish that this judgement is subjective; and that the ability to discriminate between gray levels varies from person to person and depends on ambient lighting levels. The evidence does not establish that the "skilled addressee" is not a radiologist, or a team which would include a radiologist. It does not establish that a radiologist, viewing an image under standardised conditions, could not unambiguously make this judgement. The state of the evidence is accordingly such that the opponent has not proved, on the balance of probabilities, that this part of the claim is bad for ambiguity.
Mr Yates also submitted that the expression "with an accuracy having a standard deviation of less than 2 percent" lacked clarity. (This expression refers to the accuracy of placement of the scan lines.)
Mr Yates first referred to the definition of "scan line" at lines 28 and 29 on page 5 of the specification, namely:
" `scan line' means a series of spots along the width of a recording medium".
[I note that the following lines, lines 30 and 31, set out a definition of "accuracy of scan line placement", namely:
" `accuracy of scan line placement' means the standard deviation in the position of adjacent scan lines".]
The details of Mr Yates' submissions in this respect were as follows.
"I refer you to the definition of "scan lines" as it appears in the specification at page 5 line 28. Scan line is there defined to mean "a series of spots along the width of a recording medium". Now there are a number of things to note about that. First is that it is a series of spots. It may be 2 spots, 3 spots, 10 spots, a hundred spots or any number of spots, as long as there is more than one spot. The next thing to observe is that the spots may describe a line of any shape. It may be a straight line, it might be a curved line, it may be a combination of curved and straight lines. The other thing to observe is that the line will have a width. Now if one goes to integer (e) and the desiderata one sees that the scan lines have to be placed accurately and they have to be placed with an accuracy having a standard deviation of less than 2%. It is clear from integer (e) that the accuracy has to be gauged at the image plane because integer (e) says "means for positioning said scan lines consecutively and accurately along the length of said image plane". The balance of the claim says "said combination being adapted to enable the formation at the image plane of a continuous tone image".
The image plane and the recording medium are two separate things. That is clear from integer (d). The image plane is where the image is formed and it also happens to be a locus at which a recording medium may be positioned.
When one has an image recording system and one seeks to ask whether there is infringement of claim 1 one is naturally going to ask "are the scan lines being placed at the image plane with an accuracy having a standard deviation less than 2%". Now to what datum is the accuracy measured. Is it to be measured by reference to some point on the image plane? And if so, what is it? Or is it to be measured by some other series of spots? And if so, what series of spots? And how do you carry out this measurement? The specification itself lays down no criteria for measuring accuracy of line placement at the image plane. How is the accuracy to be measured in the absence of that data being recorded? And if you assume that the series of spots is the preceding scan line, and that its relative position to the scan line whose accuracy is to be measured can be determined at the image plane you have to consider how straight is the preceding line, and how parallel to the preceding line is the line whose accuracy is to be measured.
Then you have to consider how those factors are to be taken into account in measuring the standard deviation.
For example the two lines are not straight and parallel, then the following problems arise. If we assume that, by way of illustration, the series of spots describes a line broadly as I have given it [on a sketch which Mr Yates made before me] and if the next line whose accuracy is to be measured is something as I have indicated then at what point - these are two lines which are not straight and are not parallel then at what point along these two lines is the distance to be measured? Is it to be measured in more than one spot? If so, where along the lines are they measured and at what intervals? And how are the set of measurements dealt with? Are they averaged? Or do you take a median? How do you take account of the change in thickness of the line? At what point of the line do you start and end with your measurement? Because we know from the body of the specification that the spot size can be anything from a thousand microns to 5 microns. That's at page 10 lines 32 and 33.
Now then once those measurements have been obtained, how closely do they reveal a normally distributed population, it being remembered that a standard deviation is the statistical measure of a normally distributed population. And if the measurements do not reveal a normally distributed population, then what correctional matters are to be taken into account in obtaining the standard deviation? In other words there is no criterion in the specification for measuring the accuracy of line placement at the image plane.
And then one asks, how is one to carry out the measurements? Including the selection of the measuring points, that is obscure. And then how is one to deal with the results to arrive at a standard deviation. So we submit that the accuracy to which the claimed image recording system is limited cannot be clearly determined and thus anyone willing to exploit a system will not know whether or not he falls within or without the monopoly which is sought."
In reply, Mr Catterns first pointed out that:
"At [page 10] line 34, and this is the answer to one of my friend's ambiguity points, scanning is a means for sweeping the beam across the recording medium in a series of horizontal parallel lines. My friend can't submit that they are circular lines or wavy lines so you don't know where to measure them. They are horizontal parallel lines. And everyone knows what standard deviation is. If the lines have any width to them anyone in the art, I submit, would know and measure them from the middle of the lines."
Mr Catterns also said:
"I am also told, and I think this is evidence from the bar table, and I think my learned friend's was too, that you can calculate the standard deviation even if the intervals that you are measuring are non-normally distributed, it's just a matter of doing those square roots."
Mr Catterns was correct in his observation about "evidence from the bar table". I can speculate that a "person skilled in the art" would understand this part of the claim to have the meaning outlined by Mr Catterns. As a person with very basic skills in this particular art, I would have understood this part of the claim to have that meaning. However, from past experience in the interpretation of claims defining "high technology" arts, I readily appreciate that "a little learning is a dangerous thing". I can also speculate that, to a person skilled in the art, this part of the claim may well present the irresolvable ambiguity which would result in invalidity. I can speculate that the claim may be ambiguous to the skilled addressee, I can equally speculate that it may not be ambiguous. The state of the evidence is such that the opponent has not established, on the balance of probability, that this part of the claim lacks clarity.
When considering the clarity of the concluding words of the claim, Mr Yates also addressed the meaning of the expression "said scan lines being placed at a rate not less than 80 lines per second", in the following terms.
"The claim is unclear in other respects. Another parameter to which the system is limited is the speed at which it operates. It is said that it operates at a rate not less than 80 lines per second. Now the critical consideration here is the length of the scan line because the choice of the length of the line will determine the speed of the apparatus. Now the definition of the scan line in the body of the specification is silent. As a matter of construction it could be any number of dots from 2 to a very large number. And it is not facetious to ask rhetorically and by way of analogy how long is a piece of string because the body of the specification tells us that each line can have any number of pixels. It does say between one to 200 pixels per millimetre would be useful. That is at page 10 line 56. But it is clear, both by express omission and by construing the specification according to its terms that the line can have any number of spots as long as it's got more than one.
And we know also from the body of the specification what is thought to be desirable is something width is a scanning means somewhere between 5 mm and 5 metres and that is on page 11 line 4.
And the characteristics of the line that its size is also a function of its spot size and the specification indicates at page 10 lines 32 and 33 that 1,000 microns ... can be useful but without limiting those ranges. And also that a function of the length of the line is also not only the size of the spots but also whether they are uniformly spaced.
Now to say that the machine can place down the scan lines at 80 lines per second is really quite meaningless because the speed is indeterminate therefore anyone with a given image recording system and wishing to know whether or not their apparatus infringed claim 1 just wouldn't know. There would be some considerable doubt whether the apparatus did or did not."
Mr Catterns replied as follows, when addressing both the clarity and fair basis of this part of the claim.
"Now the next criticism on section 40 related to the 80 lines per second. And my friend submitted that length of the line determined the speed. So far as ambiguity goes in my submission there is no question of ambiguity, 80 lines per second means what it says. It's got to run to the end of the line, come back again, do it again, 80 times per second. And it doesn't matter there that the lines are different lengths. If my friend builds one that is five metres wide, and he does 79 lines a second, leaving aside questions of pith and marrow, he won't infringe. And good luck to him. ... Now on fair basis one of my learned friend's claims is that we only, we claim as low as 80 lines per second but our example was 84 lines per second and so therefore we have fudged a few lines per second. That example is at page 15, about line 20. A continuous tone photographic quality image was printed by the system in 41 seconds. The image had 3300 lines. We think that that is 80.4, not 84, by the way. We respectfully submit that this is not his best point. To put it the other way round, if it was forty-one and a quarter seconds it would be down to 80, and that is possible, its the nearest whole number. The fact that they are a fraction of a whole number above, and when we go on, and when we go on and say "From this, the continuous tone image is formed at a rate of at least 80 lines per second," there is the disclosure of 80 where we do our own rounding, is not a good fair basis. The bit I just read was line 24 on page 15."
I think that there is some merit in Mr Cattern's submission that "80 lines per second means what it says. It's got to run to the end of the line, come back again, do it again, 80 times per second." To quote the paragraph bridging pages 10 and 11 of the specification
"Scanning is a means for sweeping the beam across the recording medium in a series of horizontal parallel lines. ... Scanning is accomplished in the present invention, for example, by a polygon mirror having 1 or more reflecting facets, a scanning galvanometer, acousto-optic deflector, or a holographic deflector, all of which devices are well-known in the art."
The number of lines scanned per second when using a rotating polygon mirror will be precisely determined by the number of facets to the mirror and by the number of times the mirror rotates per second. It is true, but irrelevant, that the linear speed of the flying spot will also depend on the length of the optical path from the rotating polygon to the image plane. The claim defines the invention in terms of the number of scan lines placed per second, not in terms of the linear speed of the scanned spot.
The number of facets to a rotating polygon is invariant. Control of the rotational speed of electric motors is a precise art. I am sure that a person skilled in the present art would be able to build a printer with a rotating polygon scanner with a specified rate of placement of scan lines. More relevantly, the evidence does not show that a skilled addressee would not be able to tell what the line placement frequency would be.
Similar observations apply to the other types of scanners which are referred to, that is scanning galvanometers, acousto-optic deflectors and holographic deflectors. The number of lines scanned per second by a scanning galvanometer is determined by the energising frequency, and control of oscillator frequency is a precise art. The number of lines scanned per second by an acousto-optic deflector is precisely determined by the frequency and wave shape of radio frequency excitation to a piezoelectric transducer. The number of lines scanned per second by a holographic deflector is determined by the speed of rotation of a spinning disk. There is no evidence that the skilled addressee would not be able to build these scanners to perform at a specified frequency of scan line placement. Accordingly, I find that this part of the claim has not been proved to be ambiguous.
The facts of the present case are markedly different from those in the decision referred to earlier by counsel, that is Australian Solar Mesh Sales v Tomlin Industries, (1991) 21 IPR 447 at pages 454 to 456. The claim in that case defined
"a mesh material for use as a screen, said mesh comprising vinyl coated glass fibre yarn of approximately 0.25 mm diameter woven so as to form a matrix of substantially rectangular holes ... "
Justice Wilcox held (at lines 11 and 12 of page 456) that "The description in the claim is referring to the mesh as manufactured". That is, the claim did not define a mesh which had been produced by taking a fibre yarn of 0.25 mm diameter, and weaving it into a mesh. Rather the claim defined a mesh in which the yarn had a diameter of 0.25 mm. However
"The expert evidence shows that the fibreglass yarn used in this mesh was originally round. However, in the hot weaving process it was flattened."
The court then decided, on the basis of expert evidence, that the flattened yarn in the finished product did not have a "diameter in the ordinary sense of that term". The court then went on and found that a highly qualified metrologist using very sophisticated equipment could determine what the diameter of the yarn was before it was flattened. There is no analogous problem of interpretation in the present claim. The present claim is limited by reference to an entity which does exist (the line scanning frequency) and not by reference to a non-existent entity (the "diameter" of flattened yarn).
Accordingly, to use the words of Justice Gummow in Rescare v Anaesthetic Supplies cited above, the opponent has not established that claim 1 does not "recite some limitation (by result) which draws a line between two classes of things". I consider below the issue of whether or not the description adequately instructs the reader "how to go about achieving that result".
NOVELTY
Novelty in the Light of the Ricoh Specification
The single piece of prior art which is most relevant to claim 1 is the Ricoh specification which I referred to earlier. An English language translation of that specification is Exhibit PEC-2 to the declaration by Mr Ciddor, and Exhibit DMS-2 to the declaration by Dr Mark Geoffrey Sceats. That translation describes the Ricoh invention as producing "half-tone" images. The "half-tone" process, as used in newspaper publishing, involves printing dots, all of which have the same colour and tone, but different sizes, to give an overall effect of gray levels. However declarations by Mr C J O'Sullivan and Dr Toshiko Mori establish that, although "half-tone" is one correct translation of the Japanese word which was used in the original, in context the terms "gray scale" or "gradations of shade" may be more appropriate.
Given this modification of the original translation of the Ricoh specification, that specification states at lines 18 to 21 of page 2 of the translation:
"The present invention ... provides a method of driving a semi-conductor laser for reproducing a finer gray scale with greater accuracy than the conventional one."
The claim of the Ricoh specification reads:
"A method of driving a semiconductor laser in a laser printer characterized in that an information signal to be inputted into the laser printer is gamma-corrected, the light output of the semiconductor laser is detected by a photodetector, the output signal from said photodetector is compared with said gamma-corrected information signal by means of a differential amplifier, and the difference signal obtained thereby is inputted into a driver circuit for the semiconductor laser to control the drive current of the semiconductor laser."
The specification says that the invention relates to a method of driving a semiconductor laser in a laser printer. It describes that the light output from the semiconductor laser "is modulated according to the information signal inputted into" the printer. This light forms a latent image on the surface of photosensitive material. A visible image is obtained by developing this latent image. To quote the specification:
" ... is desirable that the visible image which is finally obtained has image density proportional to the level of the information signal. However, in the conventional printer of this type, it was not possible to have the image density proportional to the information signal level. Accordingly, the reproducibility of the so-called half-tone image was low and the quality of the visible image was inevitably lowered.
The present invention was made based on the above observation and provides a method of driving a semiconductor laser which has a half-tone reproducibility of higher accuracy than the conventional one."
The specification then goes on to describe a preferred embodiment. Figure 2 in the drawings shows a graph of "The relationship between the output light (amount of light) L from the semiconductor laser 2 and the drive current I flowing through the semiconductor laser". The specification then says:
"In this case, in order to utilize the linear portion of this characteristics, [sic] the information signal has been subjected to a bias to obtain variation in the light output L0 - L10 which is proportional to the change in the drive current I0 - I1 based on the information signal. By doing so, it is possible to expose the photosensitive material to the amount of light proportional to the level of the information signal. However, the relationship between the amount of exposure of the photosensitive material and the density of the developed visual image is not linear as shown by curve B in Fig. 3. Therefore, if the information is simply subjected to a bias as heretofore been conducted [sic], it is not possible to obtain a visible image of density which is proportional to the level of the information signal, and the aforesaid drawback in which the reproducibility of half-tome image is lowered cannot be avoided.
The present invention originates from an idea that the relationship between the level of the information signal and the light output of the semiconductor laser is gamma-corrected as shown by curve C of Fig. 4 and the relationship of the information signal and the image density is approximated by a linear relationship as shown by curve D of Fig. 5, thereby improving the quality of the visual image. If the relationship as shown in Fig. 5 is obtained, a visual image having the density proportional to the information signal level can be obtained and consequently the reproducibility of half-tone is enhanced.
Fig. 6 shows an example of a circuit for the laser device 1 shown in Fig. 1. In Fig. 6, the part enclosed by a broken line is a gamma-correction circuit 10 for gamma-correcting the information signal. The information signal inputted into the laser printer is gamma-corrected by the gamma-correction circuit 10. On the other hand, part (light output L2) of the laser beam emitted by said semiconductor laser 2 is detected by a photodetector comprising, for example, a photodiode 11, and the detected output is compared with the gamma-corrected information signal by use of a differential amplifier OP3, and the differential signal thus obtained is fed back to a driver circuit 12 for a semiconductor laser to control the driving current of the semiconductor laser ... "
The specification then goes on and gives a detailed description of the gamma-correction circuit, and how it operates. Following that description, the specification says:
"According to the above-described structure of the present invention, since the gamma-corrected information signal as shown in fig. 7 is inputted into the light output detecting feedback loop of the semiconductor laser as a standard signal and the light output of the semiconductor laser is modulated thereby, the relationship between the information signal inputted into the laser printer and the image density of the visible image can be made close to the proportional relationship as shown in Fig. 5, whereby a half-tone presentation of high accuracy can be effected."
The preamble of the specification says that the invention "provides a method of driving a semi-conductor laser for reproducing a finer gray scale with greater accuracy," [my emphasis]. The detailed description of the invention is about "gamma-correction". In the photographic arts, "gamma" is a rough measure of the contrast of a photographic medium. For example, in the field of photographic negative film, "gamma" is the slope of the linear part of the graph of density of the negative against the logarithm of exposure. On my interpretation of the Ricoh specification, it is about the compensation of non-linearities in the signal path to more "accurately" reproduce, in the printed image, the contrast of the original image. The feedback arrangement of Ricoh does not specifically teach the elements which are recited in paragraph (b) of claim 1 of the present specification. That is, the element that the feedback circuit be:
"responsive to the instantaneous intensity of said beam of light for continuously providing a feedback control signal ... for adjusting said laser diode to stabilize its light output, allowing virtually instantaneous changes in output intensity in response to an input signal while maintaining the output constant regardless of changes in the characteristics of said laser diode". [My emphasis]
The principles set out by Mr Blanco White QC in paragraph 4-104 of the fourth edition of Patents for Inventions are directly in point. That paragraph is as follows.
"A claim for a method of using certain apparatus is not lacking in novelty merely because the apparatus itself is old: in addition the method of use must be old. That is to say, either the old apparatus must actually have been made and used within the claim, or there must have been publication of `clear and unmistakable instructions so to use it.' [Citing Flour Oxidising v Carr (1908) 25 RPC 428 at 457.]
In a case where the apparatus was never made and used before the patent, but was merely the subject of a paper proposal containing no express instructions for use in the patented way, the test for anticipation may be phrased in the following way: If at the priority date of the claim a person skilled in the art concerned and grappling with the patentee's problem had had the earlier document put into his hand, would he have said: `This gives me what I want'? [Citing British Thomson-Houston v Metropolitan-Vickers (1928) 45 RPC 1 at 22.] In other words, to deprive the claim of novelty the earlier proposal must `deal with the same problem and get the same answer' as the patent. [Again, citing British Thomson-Houston v Metropolitan-Vickers.]
It is to be understood that such considerations do not apply where the earlier document contains express instructions to do or make something within the claim, as, for instance, in the case where the claim is for the apparatus previously proposed itself and not for a particular way of using it ... [Citing Molins v Industrial Machinery (1938) 53 RPC 31 at 41.] It would seem also that this test would not apply if the earlier proposal were for apparatus that could only be used within the claim, for this would seem to amount to the giving (by implication) of instructions for such use."
It is true that the present case is not about a claim to an allegedly new method of using an old apparatus. It is however about an analogous issue. It is about modifying an integer of an allegedly old apparatus so that the apparatus can achieve a different result. I appreciate that this statement may over simplify the analogy. It may be more accurate to describe a diode and feedback arrangement which "maintains output constant" as a different integer from a diode and feedback arrangement which operates to maintain relative accuracy of representation of tones. However the analogy be stated, it is an accurate analogy. The relevant question to ask is:
either the old apparatus must actually have been made and used within the claim, or there must have been publication of `clear and unmistakable instructions so to use [or make] it.'
There are no `clear and unmistakable instructions' in Ricoh to make a laser printer with a feedback circuit to stabilise the output of a laser diode to maintain "the output constant regardless of changes in the characteristics of said laser diode". Accordingly, a textual analysis of Ricoh by itself does not establish that claim 1 lacks novelty.
However, a mere textual analysis of Ricoh independently of the evidence does not conclusively settle the issue of novelty of claim 1 in the light of Ricoh. It may be that the skilled addressee, in implementing Ricoh, would inevitably produce something which is within claim 1 although that result is not transparent from a textual analysis of Ricoh. I refer to the decision in General Tire & Rubber Co v Firestone Tyre & Rubber Co Ltd, [1972] RPC 457 at 485-486:
" ... if carrying out the directions contained in the prior inventor's publication will inevitably result in something being made or done which, if the patentee's patent were valid, would constitute an infringement of the patentee's claim, this circumstance demonstrates that the patentee's claim has in fact been anticipated." [My emphasis]
Ricoh has been the subject of a number of "reconstructions" and other evidence. That evidence, in so far as it is material to the present issue raised by General Tire v Firestone, is as follows.
In paragraphs 8 and 10 of his declaration of 22 August 1989, Dr Sceats says:
" ... the Ricoh specification ... with the substitution of a laser diode ... for an ion laser ... represents the conventional practice at that time.
...
10. It is clear to me that the Ricoh specification is concerned with the accurate reproduction of images, and the apparatus it describes would be capable of producing a high quality image in which the gray level appears to the eye to change smoothly, and which has a large number of measurable gray levels. There is no reason to believe that the apparatus described in the Ricoh specification would not give a high quality image with as many gray levels as the apparatus described in the 3M specification; and every reason to believe that the 3M apparatus would be less accurate through neglect of the non-linearity addressed by the Ricoh specification."
Hisatoyo Kato, Yuji Ohara and Takashi Shoji are all well qualified scientists or engineers employed in research or development work by Fuji Photo Film Co Ltd. The relevant parts of their declarations are similar, and to the following effect:
"3. In the course of employment by Fuji Photo Film Co., Ltd., I was asked to reconstruct and test apparatus according to the Ricoh specification (hereinafter referred to as "the test apparatus") are as set out in Exhibit FPF-2.
4. The purpose of the test was to determine the number of gray levels which could be produced using apparatus constructed according to the RICOH specification. Attached hereto and marked Exhibit FPF-3 are test results comprising two sheets of transparent film imaged by the test apparatus and two tables giving the measured densities of the various shadings provided by the test apparatus. The densities were measured using a Macbeth TD 932 density measuring instrument.
5. I declare that the tests of the apparatus made in accordance with the Ricoh specification shows that the Ricoh specification teaches the construction of an image recording system which allows continuous tone images to be formed creating in excess of 64 gray levels when imaged onto a transparent medium."
In reply to this evidence, Professor Orr's declaration of 21 December 1990 says:
"Ohara, Kato & Shogi [sic]
These declarants aver to testing the circuit the subject of Ricoh Japanese Patent Application No. 56-205582. As I understand it, this Japanese patent application had no foreign counterpart and its availability in Australia was limited to the accession of a Japanese language copy of the specification to the Australian Patent Office Library before the priority date of the opposed application. Certainly, the subject matter of that specification could not be said to be part of the knowledge which might be expected of the average skilled worker in this field in Australia on or before August 27, 1984.
Indeed, I had no knowledge of the existence or of the disclosure of this specification at that date, nor do I expect that the document would have been located on a reasonable search conducted by a reputable searching authority.
In exhibit "FPF-2" the Opponent's simulation test of the Ricoh patent apparatus is said to have been conducted by use of elements now used in the Opponent's laser printer Model CR-LP414, except for the LD drive circuit. To the best of my knowledge and belief, that particular model laser printer by the Opponent Fuji was not available in Australia on or before the priority date of the claims of the opposed application. To the best of my knowledge and belief, that particular printer did not become available at all until well after 1984.
Further, Fuji CR Film Type 780 was not, to the best of my knowledge and belief, available in Australia before the priority date of the opposed application and indeed, as I understand the situation, has never been commercially available in this country.
It is pertinent to point out that any equipment designed to meet the criteria of amended claim 1 of the opposed application must operate at very high speed. With this in mind, appreciating the componentry available in Australia on and before August 27, 1984, I would have thought it extremely difficult if not impossible to have put together a successfully operating piece of apparatus as now claimed in claim 1 of the opposed application and as depicted and described in the specification of the opposed application with reference to either of the examples given therein. I further note that the Ricoh Japanese laid-open patent specification does not teach a recording system, rather circuitry for a gamma correction circuit and that circuit's utility. In the Fuji reconstruction of the Ricoh circuitry, there would seem to be major differences in optical components between those depicted in the figures of the Ricoh specification and those utilised in the Fuji reconstruction. The Fuji reconstruction uses a far more sophisticated system than that disclosed in the Ricoh laid-open patent specification. Again, the media transport system of the Ricoh laid-open patent specification on the one hand and the Fuji reconstruction on the other differ significantly.
There are three differences between the circuits of the Ricoh disclosure and of the Fuji reconstruction. These are as follows:
1. the protection diode
2. the feedback capacitors
3.the gamma correction circuit, which is not explicitly shown in Fig.3 of FPF-2, nor elsewhere described in that exhibit.
In the fourth paragraph of exhibit FPF-2, under the heading "Results of the Test" there is reference to the fact that visual observation might be used to estimate 150 grey levels. In my view, the human eye may not be sufficiently quantitative to effect such estimation."
In his declaration of 28 February 1991, Dr Peter Jamieson (one of the three inventors of the present application) says:
"The Ricoh document appears to disclose primarily a method for correcting the emission of the laser diode to adjust for the non-linear response of the photosensitive drum media. We do not require such a circuit since we use a silver halide film as the photosensitive medium which exhibits a long straight-line section on the photometric curve. We make any necessary corrections to the exposure conditions by use of preloaded look-up tables, a different method from that described by Ricoh. The Ricoh document is thus irrelevant to our invention, and had we known about it at the time, it would not have been of any help."
John Dennis Xanthos is a director of Optical and Photometric Technology Pty Limited. He is a declarant for the applicant. The relevant parts of his declaration are as follows:
"2. My company has been briefed in this matter to consider the specification of RICOH K.K. Japanese patent application No. 56-205582 and to determine the parts and equipment available in Australia as of August 24, 1984 were a reconstruction of the apparatus described and depicted in that specification to be attempted. At the time of this briefing my company was provided with no other documentation concerning these proceedings and, in particular was not provided with a copy of the declarations of any witness in the proceedings.
3. My company first conducted a literature search to determine whether the device depicted and described in the said RICOH specification was exemplary of what was known in Australia at that time. This search involved international computer database searches conducted both by my company and by outside searching authorities in Australia and the United States, including patent abstracts and indexing from journals covering computers, telecommunications and electronics.
The searches also embraced the USA based National Technical Information Service database which includes government sponsored research, development and engineering, this database including in the range of 1.5 million records from 1964 to date. A further search was carried out at the Bailliue Library, Melbourne University using the Library's on-line catalogue.
None of these searches included any reference to the RICOH specification. Some of the references cited by those searches concerned laser beam printers, laser beam control, image sensing and printing and optical scanning systems. However none of the cited references was in any way relevant to the disclosure of the RICOH specification, the subject matter of which could not, to the best of my knowledge and belief, be said to have been commonly known to persons with expertise in this field (electro-optical technology) in Australia in August 1984.
4. I have also ascertained details of the componentry available in Australia on or before August 25, 1984, either physically or by import order, which might be used to reconstruct the RICOH device with a view to using the reconstructed device to produce an exposure according to the following specifications:
1)a transparent output medium and 64 grey levels,
2)format size 8.5 x 11 inches (216 x 280 mm),
3)pixel size 0.1 mm x 0.1 mm,
4)uniformity of exposure to be no more than 0.005 peak-to-peak in density with a scan line placement with an accuracy having standard deviation less than 2%,
5)scan line placement not less than 80 lines per second.
These standards are reasonable as minimum performance standards for use in quality radiographic images.
5. Having ascertained those details my company then proceeded to build and test a laser printing device which was an effective reconstruction of the RICOH device using componentry available in Australia in August, 1984 ...
6. Thus, having been provided with a translated copy of the Ricoh Japanese specification No. 56-205582, an instruction to reconstruct the laser printing device depicted and described therein utilising equipment available in Australia on or before August 24, 1984 and to use that reconstructed device to form an image subject to the criteria specified in paragraph 4 above of this declaration, I conclude that it is, and at all material times was, impossible to achieve the specified results in any acceptable form.
7. Having completed my company's report on the aforesaid reconstruction project, I was provided with a copy of each of the declarations of Takashi Shoji, Yuji Ohara and Hisatoyo Kato and the exhibits FPF-1, FPF-2 and FPF-3 hereto.
8. In respect of the said declarations:
(i)There is considerable difference between the optical systems of the FUJI reconstruction and the Ricoh patent. These differences lie after the laser beam leaves the polygon scanner travelling towards the film. FUJI use an F-theta lens, whereas Ricoh does not. FUJI also use a negative cylindrical lens and a negative cylindrical mirror whereas Ricoh does not;
(ii)There is no detailed specification of the optical system used in the FUJI reconstruction with respect to componentry dimensions, focal lengths and powers of lenses;
(iii)In paragraph 5 of each of the said declarations, the declarants state " .... the Ricoh specification teaches the construction of an image recording system .... ." This is not the case. In order for the Ricoh specification to teach the construction of a working commercial image recording system, it would need to:
(a) incorporate a very different optical system from that shown, as my company has conclusively proven that the optical system as shown in Ricoh has resolution constraints from approximately 10 mm either side of the vertical centre line (optic axis) of the recording medium (film) to the outer edges of the film. The Fuji reconstruction did not use the Ricoh optical system to achieve its alleged results; and
(b) give a very detailed specification of the entire optical system as shown, together with details of gamma-correction and modulation circuitry.
(iv)The FUJI CR film type 780 was not available in Australia in August 1984."
Philip Sydney Francis is a declarant for the applicant. He is a Professional Officer employed by the Royal Melbourne Institute of Technology, Applied Physics Department as a research assistant. The parts of his declaration which are material to the present issue are as follows.
"5.The following comments are appropriate in relation to the FUJI reconstruction of the RICOH device as attested to in the declarations of Shoji, Ohara and Kato:
(i)The FUJI "reconstruction" utilises a substantially different optical system from that shown in the Ricoh specification, viz.
(a)The Ricoh specification shows an optical system with a positive cylindrical lens between the scanning polygon mirror and the main focussing lens. However, the FUJI reconstruction does not have a positive cylindrical lens between the scanning polygon mirror and the main focussing lens;
(b)The Ricoh specification shows no optical components between the main focussing lens and the scanned medium. The FUJI reconstruction uses a negative cylindrical lens and a negative cylindrical mirror between the main focussing lens and the scanned medium.
The variation from the optical system of the Ricoh specification used in the FUJI reconstruction would have resulted in a better image resolution across the scan line than would have been possible using the optical system shown in the Ricoh specification, since the optical system shown in the Ricoh specification cannot possibly produce a high definition circular spot that is uniform across the width of the scanned medium.
This is due to the changing optical path length between the scanning polygon mirror and the second cylindrical lens, as the beam is scanned. It may be possible for the optical system of the FUJI reconstruction to produce the appropriate high resolution, given it uses a different optical system and given appropriate focal lengths and element spacing in that different system, details of which features are not indicated in the FUJI drawing.
(ii)The Ricoh specification shows a cylindrical recording surface whilst the FUJI reconstruction uses a flat recording surface;
(iii)The FUJI reconstruction is absent any detail of the gamma-correction circuit, thus failing to allow comparison with the Ricoh specification;
(iv)The declarants provide no specific detail concerning numerous components of the reconstruction, specifically the frame memory, the interface between the frame memory and the RAM, and the digital/analogue converter."
Colin James Richard Sheppard is a declarant for the opponent. He is Professor of Physics (Optics) and Head of Department of Physical Optics, University of Sydney. The material parts of his declaration are as follows.
2.I have read Exhibit CJS-2 which is an English translation of Japanese unexamined Patent Publication No. 58-106953; a typed document with numerous manuscript amendments. This document addresses the problem of forming a continuous tone image from a modulated semiconductor laser, and in particular concerns the production of a linear relationship between an information signal and the resulting image density. It is of course clear to anyone with any darkroom experience that it is necessary for the photosensitive material to be sensitive to wavelengths in a range which overlaps the range output from the light source, and that a linear relationship between the output power of the source and the resultant image density is desirable if continuous tone images are to be formed.
CJS-2 describes a laser printer with reference to Figure 1, and I note that the optical system is not described in any detail. Figure 1 schematically shows mirrors, a collimator, a pair of cylindrical lenses, a rotating polygonal mirror and a focusing lens between the semiconductor laser and the photosensitive material, but presumes that the reader of the specification would be familiar with the details of a practical implementation of the optical system; or at least would be able to find this information from published sources as indeed I would.
3.I was subsequently shown Exhibit CJS-3 which is Exhibit JDX-3 "A final report on reconstruction of a laser printing device". The first thing I noticed about this report is that it is attempting to prove a negative, in other words that a particular result cannot be achieved by the apparatus, and in my view no such attempt can be scientifically conclusive.
I note that the system is said to have "performed moderately well" before the cylindrical lenses were included, but that although the inclusion of the cylindrical lenses eliminated the grouping of the scan lines, the second cylindrical lens caused a change in spot size across the scan line. If there were some reason why this were undesirable a number of possible solutions occur to me; these include the use of corrective optics, or the use of an alternative scanning means which would allow the second cylindrical lens to be dispensed with.
4.There are a number of factors in any production system which will influence the quality of continuous tone images; for instance the sensitivity of the film, and the quality of the optical components and any mechanical parts. An increased number of grey levels will require a higher sensitivity and a greater accuracy in the line and pixel placement. This is self-evident from the nature of the image being an array of grey dots."
I do not think that this evidence establishes that:
" ... carrying out the directions contained in the prior inventor's publication [ie, in Ricoh] will inevitably result in something being made or done which ... would constitute an infringement [of the opposed] claim";
and lack of novelty has not been established under the test set out in General Tire v Firestone.
Hisatoyo Kato, Yuji Ohara and Takashi Shoji do give evidence that their reconstruction of Ricoh will
"allow continuous tone images to be formed, creating in excess of 64 gray levels when imaged onto a transparent medium"
("Continuous tone images with at least 64 gray levels when imaged onto a transparent medium" is part of the "definition by result" in the opposed claim 1.)
However, Professor Orr gives evidence to the effect that
"In the Fuji reconstruction of the Ricoh circuitry, there would seem to be major differences in optical components between those depicted in the figures of the Ricoh specification and those utilised in the Fuji reconstruction. ..."
John Xanthos gives evidence that
"I conclude that it is, and at all material times was, [by reconstructing Ricoh] impossible to achieve the specified results in any acceptable form."
The "specified results" which Mr Xanthos referred to in this passage are
1)a transparent output medium and 64 grey levels,
2)format size 8.5 x 11 inches (216 x 280 mm),
3)pixel size 0.1 mm x 0.1 mm,
4)uniformity of exposure to be no more than 0.005 peak-to-peak in density with a scan line placement with an accuracy having standard deviation less than 2%,
5)scan line placement not less than 80 lines per second.
Items 1), 3), 4) and 5) are based on, although not textually identical to, the "defining by result" limitations in the opposed claim 1.
Mr Xanthos also declares that
"The Fuji reconstruction did not use the Ricoh optical system to achieve its alleged results ...".
On the other hand, Professor Sheppard, a declarant for the opponent, observes that Mr Xanthos reported that the [Xanthos] reconstruction of Ricoh "performed moderately well" before the cylindrical lenses were included, but that the inclusion of cylindrical lenses caused a change in spot size across the line. Professor Sheppard then declares that "a number of possible solutions" occur to him.
There is thus some evidence that the prior Ricoh specification teaches a construction which infringes the opposed claim 1. There is other evidence, critical of the accuracy of the Fuji reconstruction of Ricoh, which contradicts this evidence. There is also evidence, based on the Xanthos reconstruction of Ricoh, to the effect that Ricoh does not teach a construction which would infringe the opposed claim 1. The opponent has not demonstrated, on the balance of probability, that claim 1 lacks novelty under the test set out in General Tire v Firestone.
Novelty in the Light of Other Art
Mr Yates made submissions to the effect that each of integers (a) to (e) of claim 1 was known in the prior art, and that various combinations, including the combination of all of integers (a) to (e) was known. Those integers are:
(a)"light means including a solid state laser diode ...";
(b)"a feedback circuit ...";
(c)"means for at least one of collimating, focusing, and controlling the path of said beam of light";
(d)"means for scanning said light beam across the width ..."; and
(e)"means for positioning said scan lines ... along the length ... ".
I accept that the evidence establishes, in a broad sense, that in this art laser diodes, feedback circuits, means for collimating, etc, means for scanning across the width of an image, and means for positioning scan lines along the length of an image are all known. However, to establish lack of novelty in claim 1, it is not sufficient to establish merely that broad forms of these integers exist in the prior art. It would be necessary for the opponent to establish:
(1)that there has been prior disclosure of the combination of these integers;
(2)where those integers are each in the specific forms recited in claim 1; and
(3)that combination is "adapted" to achieve the result which is recited in the concluding lines of the claim.
I do not think that I need to decide whether or not the evidence establishes item (1). This is because I am not satisfied that the opponent has established either item (2) or (3).
So far as concerns item (2), it would be necessary for the opponent to establish, inter alia, that the prior art discloses integer (b), that is:
"a feedback circuit responsive to the instantaneous intensity of said light beam for continuously providing a feedback control signal coupled to said light means for adjusting said laser diode to stabilize its light output, allowing virtually instantaneous changes in output intensity in response to an input signal while maintaining the output constant regardless of changes in the characteristics of said laser diode".
Mr Yates submitted that feedback circuits were well known. Mr Yates said:
"The next integer, a feedback circuit ... was also known. One sees for example in the inventor's specification at page 7 line 20 to page 8 line 9 that there are problems with the laser diode. Problems included mode hopping, internal heating and changes due to aging. And feedback techniques for correction of the signal intensity are disclosed in the publications referred to on page 8 lines 2 to 9. Now once again Ciddor paragraph 6 says that such feedback circuits were well known in Australia at the relevant date and this is agreed to by Professor Paul in his declaration. Professor Orr is a witness called for the applicant and on page 16 of his declaration he says "it is agreed that feedback circuits were available in Australia at the priority date of the opposed application" and he does however go on to say that "not informed or optimized for application to the system of the current application or combined with suitable other components to achieve continuous tone high quality images" ... obviously going to the promised performance of this particular system"
However, the precise details of this admission are critical. For that reason, I set out a full quotation from the specification:
"Several problems exist with laser diodes which must be overcome to obtain 64 gray levels as defined above. These include mode hopping, internal heating, and changes due to aging. In order to correct for these defects the preferred embodiment uses a beam splitter to split off a fraction of the output beam to a photodiode which continuously monitors the light output of the laser diode and feeds back a control system [sic] which adjusts the laser diode to stabilize the operation of the laser diode, allowing virtually instantaneous changes in the output intensity in response to the input signal, while maintaining the output constant regardless of changes in the light intensity versus current characteristics of the laser diode. Such a continuous feedback circuit can correct for long and short term changes in laser diode operating characteristics. It can provide for an analog signal formed from a digital signal of several million samples per second with precise control over the power output that is provided in part by a feedback arrangement wherein corrections are made in less than 50 nanoseconds (i.e. about 20 MHz). This high speed correction is especially helpful in overcoming problems such as mode hopping which can occur over very short periods of time; it is not enough to control the intensity from line to line. It is therefore preferred to provide feedback correction of the signal intensity for each pixel or spot. Techniques to accomplish this type of continuous feedback control are disclosed by M. Lutz, B. Reiner, and H.P. Vollmer in "Modulated Light Source for Recording with GaAIAs-Lasers", presented at First International Congress on Advances in Non-Impact Printing Technology, Venice, Italy (July 22-26, 1983), and D.R. Patterson and R.B. Childs in "Semiconductor Lasers Reach for Maturity: Applications in Fiber Optic Communications", Photonics Spectra, pages 83-87, April 1982. Preferably, the circuitry and apparatus for stabilizing the output of a laser diode and allowing the laser diode to be modulated by a wideband analog signal comprises:
means operatively arranged for receiving the signal for modulating the output of a laser diode, the means including a high gain amplifier having two inputs, one input operatively connected for receiving the signal for modulating the laser diode, the other input for receiving a feedback signal;
a compensating network connecting the output of the high gain amplifier to the laser diode;
a feedback circuit connected to provide a feedback voltage signal to the other input of the laser diode including a pin diode;
means optically coupling the pin diode to the laser diode for providing a portion of the light output from the laser diode to the pin diode.
Fig. 3 shows one embodiment of such a continuous feed back circuit, which is also disclosed in U.S. Patent No. 4,625,315, having the same priority date as this application. Voltage input bias network 116 is provided by ... ."
In summary, this quoted passage recites existing problems with laser diodes which must be overcome "to achieve 64 gray levels". It then goes on to say that a particular type of feedback arrangement "can correct for long and short term changes in laser diode operating characteristics". It also says that "This high speed correction is especially helpful in overcoming problems such as mode hopping which can occur over very short periods of time; it is not enough to control the intensity from line to line".
The specification then states that:
"Techniques to accomplish this type of continuous feedback control are disclosed by ... ". [My emphasis.]
There is a real issue as to just what type of continuous feedback is meant where this quotation refers to "this type". This is because the preceding lines comprehend more than one type of "continuous feedback" control. Does it merely mean feedback where the feedback signal is derived by splitting "off a fraction of the output beam to a photodiode"? Or does it mean that, in the prior art, in addition to beam splitting, that there is "an analog signal formed from a digital signal of several million samples per second with precise control over the power output"?
In attempting to answer this question, I think that Mr Yates' submission attempts to read too much into paragraph 6 of Mr Ciddor's declaration; particularly when that paragraph is read in the context of the rest of the declaration.
Paragraph 6 says:
"6.Claim 1 of the 3M specification continues: "b. a feedback circuit responsive to the instantaneous intensity of said light beam for continuously providing a feedback control signal ... allowing virtually instantaneous changes in output intensity in response to an input signal while maintaining the output constant regardless of changes in the characteristics of the laser diode," ... Such feedback circuits were well-known in Australia to workers in the field of laser imaging, and to me, before August 1984, and are acknowledged as known by the writer of the 3M specification on page 8 lines 13 to 22."
However, the breadth of this paragraph must be considered in the light of paragraph 23 of the same declaration. In that paragraph, Mr Ciddor declares:
"23.Claim 3 of the 3M specification characterises the feedback circuit of Claim 1 by requiring it to comprise: "means operatively arranged for receiving the signal for modulating the output of said laser diode, said means including a high gain amplifier having two inputs, one input operatively connected for receiving the signal for modulating said laser diode, the other input for receiving a feedback signal; a compensating network connecting the output of said high gain amplifier to said laser diode; a feedback circuit connected to provide a feedback voltage signal to said other input of said laser diode including a pin diode; and means optically coupling said pin diode to said laser diode for providing a portion of the light output from said laser diode to said pin diode". But such feedback circuits are [sic] commonly known in Australia before August 1984, as exemplied [sic] by the abstract of Stark H. et al. Application of laser diodes to laser xerographic printers. In: 1st International Congress on Advances in Non-impact Printing, Venice 1981. Soc. Photographic. Sci. Eng., Washington, DC, page 110, a copy of which is attached and marked Exhibit PEC-4; and the abstract of Lutz, M. Reimer, B. In: 1st International Congress on Advances in Non-impact Printing, Venice, 1981. Soc. Photographic. Sci. Eng,. Washington, DC, pages 108 and 109, a copy of which is attached hereto and marked Exhibit PEC-5; both of which were available in Australia before August 1984. They are all generally acknowledged as known by the writer of the 3M specification on page 8 lines 14 to 21. All the means are also exemplified in Figure 6 of the Ricoh specification, see particularly OP3, OP1 TR and R8, and element 11."
Mr Christopher James Murray is a patent attorney for the opponent. A number of pieces of prior art are exhibited to his declaration of 4 April 1989.
Exhibit CJM 2 is a copy of an article entitled Laser Beam Recorder for Color Television Film Transfer by L Beiser, et al, published in the September 1971 Journal of the SMPTE, Volume 80 pages 699 to 703. That paper reports the research and development of a colour film recorder, using gas lasers. The input to the recorder is a colour television signal, in NTSC format, from a magnetic tape. In respect of that paper, Mr Murray says:
" ... it can be seen that work was being conducted with laser scan line placement accuracies of less than 1% (see ultimate paragraph of page 699)."
Exhibit CJM 3 is a copy of an article entitled Laser Picture Recording by Norio Motoki and Yukio Matsuura, from Movie Television Techniques, Volume 25 pages 292 onwards, and an English language translation. Relevant passages from that paper are as follows.
" ... In addition, previously existing mechanical type optical deflection technologies such as rotating mirrors and galvanometers have been reexamined and their performances have been improved with an eye to applying them to lasers. Thus, the foundations have been firmly laid for image-recording equipment using lasers, and they have today come into practical application in various areas. ...
There are also film supply mechanisms of various types. Feeding systems include continual feed and intermittent feed. In addition, depending on the manner in which the optical deflector and the optical system are combined, the surface on which the images are formed may be a flat surface or may have a cylindrical or a spherical shape, and corresponding methods for holding and fastening the photo-sensitive materials will be needed. If one is dealing only with the film feeding techniques for laser picture recording, it will be sufficient for 16 mm film to be fed intermittently in flat form. Therefore, the cameras used in the past in kinescope recording can be applied without any change. However, there will usually be far-reaching changes in the types of film supply mechanisms depending on the purposes for which the equipment is intended, and the mechanisms used are specific to each type of devices.
The auxiliary component elements which are needed include electrical circuits for compensation for the gradation of the recorded images, light detectors for stabilizing and monitoring the light output by the lasers, and dichroic mirror systems for combining three-colored laser beams into a single beam when recording color images. However, such elements of good performance can easily be acquired, and we will confine ourselves here to merely pointing them out. ...
An immense number of oscillation wavelengths have been reported thus far for solid, gas, liquid and semiconductor lasers. ...
Attempts have been made recently to use feedback for stabilization purposes. In some cases, some of the light output by the lasers is separated out and input into a light detector, where any power fluctuations and noise are converted into electrical signals and fed back into the discharge current. In other cases, internal modulation is used. That is, feedback is introduced into the optical modulator inserted into the laser resonator, and modulation signals are applied simultaneously, so that intensity-modulated light is obtained directly from the laser. No doubt it will be possible in the future to obtain quite stable lasers with little noise. However, since internal modulation cannot be performed at high speeds, it will be difficult to apply it to laser picture recording, although it is being used in applications such as facsimile machines.
...
We list below the characteristics of laser recording:
(1) High resolving power. ... since microspots at the diffraction limits can be formed, an extremely high resolving power can be obtained if the optical deflector is suitably designed. A resolving power on the order of some 1,000 television lines in laser picture recording can be realized easily. ......
(4) Good deflection linearity. When the optical deflector has a large deflection angle, a better linearity than in electron beam deflection can be obtained, although lens distortion may sometimes appear.
...
We have described chiefly laser picture recording, i.e. recording television signals on 16 mm film using lasers as the light sources. However, there have been very few attempts to apply lasers in broadcasting technologies. As far as the authors are aware, there seem to have been only the laser picture recording devices of CBS and NHK described here. Research on application of lasers to picture recording has been carried on widely, and these technical developments will no doubt have an important influence on broadcasting technologies in the future.
The direct picture recording onto color print film mentioned here as an example of an experiment conducted by the authors is probably the first such attempt, and this attempt probably would not have been achievable without the use of lasers. In this way, lasers have supplied us with a means so powerful with respect to the light quantity and resolving power as to be unthinkable in terms of the common sense of the past. Thus, when considering the applications of lasers to broadcasting, including laser picture recording, it will be necessary to change our viewpoint and begin to think in a new way. For example, it will be necessary to take up again and rethink ideas which were considered in the past to be impossible or were not studied seriously because it was believed that they would have little effect if implemented. Such ideas include picture recording on film by means of red, green and blue signals from the studio and optical sound recording onto film dyes."
Exhibit CJM 4 is a copy of an article entitled High-Resolution Graphics Using a HeCd Laser to Write on Kelvar Film by A D Berg et al, published int Journal of the SMPTE, Volume 83 pages 588 onwards. The abstract reads:
"A machine is described which was developed at Bell Laboratories at Holmdel, N.J., in connection with the PICTUREPHONE program to provide high resolution graphics transmission capability. This particular machine was designed using commercially available components. A low-power, 10-mW, continuous-wave helium cadmium laser is used to scan simultaneously an 8 1/2- by 11-in (21 1/2- by 28-cm) page original and a 24X reduced image on vesicular diazo heat developable film. A 1400-picture element by 2000-line image, with full gray scale, is transmitted over a 0.5-MHz line in 4 seconds, and developed in less than one second. Each machine is a transmitter-receiver capable of local copy. Applications, operating modes and results are presented."
I have previously quoted paragraph 7 of Professor Orr's declaration, where he says:
" ... to the best of my knowledge and belief, it was nowhere suggested in Australia prior to August 28, 1984 that high quality medical electronic images could be recorded directly onto film, in a reasonable length of time, using a method different from conventional photographing of a video screen. Had it been suggested to me at that time that a device of that nature might be constructed, I would have regarded the prospect as being a very considerable challenge involving extensive research and development work involving great cost and a considerable period of time."
In summary, there is evidence to the effect that the combination of integers (a) to (e), in their broad or general forms, were known. There is also evidence that the prior art illustrates concerns with image resolution and accuracy of scan line placement.
However, I think that Mr Cattern's submissions in respect of this evidence are generally convincing. He said:
"We submit that although each of these per se may have been known, we don't accept that they were when so limited, or in combination.
... I think it's important though to look at the evidence that my friend relied on the very first question because if we were in court we would object to this and my submission is that they are of very little weight. Now the first one was Ciddor, paragraph 17. These sort of bald statements without documentary support are virtually useless.
For example in paragraph 17 Ciddor says "The use of elements a, b, c, d and e of Claim 1 of the 3M specification is a common sense way of arranging a laser image recording system; it was commonly known in Australia before August 1984, to workers involved in the field of laser imaging...". It's our submission that there was no such field so called. "... that laser image recorders were organised along these lines, and this is exemplified by the Ricoh specification ..." The Ricoh specification, let me re-iterate, is in Japanese, four months in the Patent Office and nowhere else, four months before the priority date.
And notice he doesn't say "I knew it, we all knew the Ricoh". ... I don't think he relies on Ricoh as part of common general knowledge because he hasn't done what the 3M Case requires him to do. Namely get something out of the Patent Office and into the common general knowledge. I submit that's of little weight or no weight a sentence like that and Dr Paul is just as bad in his paragraph 13. I think it's even in identical words. "In my view, the use of elements a, b, c, d and e of claim 1 of the 3M specification is a common sense way of arranging a laser image recording system; it was commonly known that laser image recorders were organised along these lines before August 1984, and the Ricoh specification, insofar as it describes such a system would have been common general knowledge ..." Again they don't say "I knew it". "... common general knowledge to those with practical interest in the design of laser image recorders before August 1984."
Now so I submit that's why you have to be very careful about this state of the art here. In fact most of the art cited relates to gas lasers with their different optical characteristics though the Ricoh one is not gas. So we do not accept, you should not accept, that Ricoh was common general knowledge on the basis of those two paragraphs.
My learned friend also tried to get some help out of two of our witnesses on this, namely Orr and Wilksch. Now I submit that the Canon device which is referred to by Orr was not show, he doesn't say it and nobody else says it does, to have these five integers. I'm not just talking about limited by results, I'm talking about per se, the bald integers themselves.
He describes the Canon device in very general terms at the beginning of paragraph 5 and I think that the part my learned friend relied on most was at page 12 where at the bottom referring to paragraph 16 where he says "It is not true that all of the elements specified by the witness were known at the priority date of the opposed application. In general terms it can be said that the elements were known, but in less developed and less sophisticated form ... Each has been more highly refined by the 3M inventors than was previously described. In addition, each element was not known in combination with the remaining features of the system." This is what Orr says. And if my friend wanted to prove that the Canon had all five integers they should have bought one and pulled it apart. And that would have been a lot cheaper than some of the other reconstructions that each side has done here."
I think that Mr Catterns is not correct in submitting that the combination of the "broad forms" of integers (a) to (e) was not part of the prior art. I do however accept that the opponent has not demonstrated that the combination of integers (a) to (e), in the specific forms necessary to meet the "definition by result" in claim 1, would be obvious to the skilled addressee.
DOES THE SPECIFICATION FULLY DESCRIBE THE INVENTION?
When I was considering the issue of "definition by result", I foreshadowed that it would be necessary to decide whether the description "sufficiently instructs the reader in how to go about achieving that result". This issue can best be summarised by the following quotation from Mr Yates' submissions:
"It is trite that the specification must put the skilled addressee fairly in possession of the invention. It has to contain such instructions as would enable all those who have an interest in the art and to whom the specification is addressed to produce something within each of the claims by following the directions in the specification. Now in addressing you on ... claims to various features I have already made reference to passages in the specification which are relevant. Cryptic references to the need for special means, precisely and carefully controlled transport systems, combinations, without exactly telling you what they are. And as to that we make these observations.
There is only one possible part of the specification which overcomes that problem and that is in relation to examples 1 and 2. Of course claim 1 and its dependent claims are not so limited. And in so far as and to the extent that claim 1 and its dependent claims seek to broadly claim a image recording system which has the twin desiderata of greater speed and greater accuracy it is not limited to the embodiment of examples 1 and 2 then of course the specification is not sufficient."
Earlier in his submissions, Mr Yates had said:
"So at this part of our submissions what we say is that, in summary, the provision of elements (a) to (e) talks about nothing but the most fundamental way of constructing a laser image recording system, well known, exemplified by the Canon system. The promise of accuracy and speed are the obviously reasonable desiderata. The evidence of the expert witnesses shows that there must be some special relationship between these elements. Indeed there must be in order to distinguish whatever is being claimed over the prior art. That it involves a consideration of the interrelationship of many elements, that the patent specification itself says that, and yet the claim itself is entirely silent as to the feature which will enable that promise to be secured. All that it says is that the said combination must be adapted. How it is to be adapted finds no expression in the claim and the claim is bad for failing to define the invention.
Another way of putting the point is this. If certain results are obviously desirable, then a claim to any method of attaining it would be too wide, and there is discussion in the English authorities, particularly the English texts about "free beer" claims. And clearly the defect of such a claim is obvious. There would no doubt be many people who would wish to claim an apparatus which performs more accurately, or performs faster, or perhaps performs more accurately and faster, but a patent will not be granted merely to cover any combination or any apparatus which achieves that without giving some consideration for the monopoly. And if this claim were to be allowed in its present form, it would in fact foreclose to others seeking to devise an image recording system which was faster or more accurate from the endeavours of their enterprise so by claiming all methods of achieving such a machine the claim is bad and lacks fair basis. But of course the claim has the fundamental defect of not even identifying one feature which secures the invention."
In considering these submissions, I think it is important to keep in mind the third of my conclusions from the decision of Mr Justice Gummow in Rescare Ltd v Anaesthetic Supplies Pty Ltd, (1992) 25 IPR 119, set out above, namely:
"It is not the function of the claim, in a situation of "defining by result", to instruct the reader how to go about achieving that result. This is the function of the rest of the specification."
The most optimistic statement of the case for the applicant is as follows. There is invention in appreciating the "idea" of claim 1, and there is invention in appreciating that the combination of integers (a) to (e) can be refined or optimized to achieve that idea. The specification then describes specific embodiments of the optimized or refined versions of integers (a) to (e) to achieve this idea. In particular, the refinement or optimization of the feedback arrangement appears to be critical. One example of such a suitable feedback arrangement is described in detail in the specification, from line 24 on page 8 to line 21 on page 9, and by cross-reference to US Patent Number 4,625,315.
To succeed against this case, the opponent has to prove a "negative" on the balance of probability. That is, the opponent must be able to point to evidence which proves that a skilled addressee cannot, merely by non-inventive experimentation, "work" the invention on the basis of the disclosure in the specification. As one of the witnesses observed in another context, it is very difficult to prove such a negative. There is certainly no direct evidence of that negative. There is evidence that it is difficult to produce both working prototypes and successful commercial products which embody the "idea" of the invention and which contain integers (a) to (e). See, for example, the applicant's evidence in paragraphs 8 and 9 of Dr Jamieson's declaration. This falls short of proof that further invention is required to go from the disclosure of the specification to a working embodiment of claim 1.
The specification also describes a number of "examples". Example 1 is particularly detailed. It even gives manufacturers' catalogue numbers for most of the components used in the example. Although the applicant did not file any evidence that this example actually produces an embodiment which falls within claim 1, there is also no evidence to the contrary.
I accordingly find that the opponent has not succeeded in proving that the specification does not fully describe the invention.
CLARITY OF THE CLAIMS
Mr Murray's declaration of 4 April 1989 is about the specification as accepted, not about the specification in its current form. In that declaration, Mr Murray makes some criticisms of the clarity of the accepted claims.
In paragraph 5 of his declaration, when referring to claim 1, Mr Murray says:
"The claim is then limited by a functional "whereby" clause which requires the system to be able to form a continuous tone image of at least 64 gray levels on transparent mediums, or 32 gray levels on opaque mediums. It is to be noted that there is no antecedent for "said elements" in line 25 of claim 1 but I assume it refers to the items labelled (a) to (e) in the claim."
The claim as amended now recites a "combination" comprising integers (a) to (e), and then refers to "said combination" rather than to "said elements". I think that Mr Murray himself implies that the accepted form of the claim did not suffer from the irresolvable ambiguity which leads to invalidity. If the accepted claim did so suffer, the amended claim does not.
In paragraph 7 of his declaration, Mr Murray says:
"It should be noted that it is unclear whether the feedback circuit of claim 3 line 11 is the same as the feedback circuit of claim 3 line 2 (claim 1 line 5) as this claim currently states there is a feedback circuit comprising a feedback circuit. Also, there is no antecedent for "said input of said laser diode". It is not stated nor clear that the laser diode does have two inputs."
These criticisms have been addressed in the amended claims.
Paragraph 6 of Mr Ciddor's declaration of 7 August 1989 reads:
"Claim 1 of the 3M specification continues: "b. a feedback circuit responsive to the instantaneous intensity of said light beam for continuously providing a feedback control signal coupled to said generating means for adjusting said laser diode to stabilize its light output, allowing virtually instantaneous changes in output intensity in response to an input signal while maintaining the output constant regardless of changes in the characteristics of the laser diode," and I note that there is no antecedent for the "generating means" mentioned in this clause, but I assume it to refer to the means of clause a."
I think it clear that this "generating means" referred to by Mr Ciddor has its antecedent in the "means ... for generating" of integer (a) of the accepted claim. However, the wording of the amended claim avoids any ambiguity in this respect.
FAIR BASIS OF THE CLAIMS
In the course of his submissions, Mr Yates said:
" ... the specification as lodged we would have said lacked clarity because the definition of gray levels was such that you couldn't determine what those levels were. It was defined by reference to a densitometer but the evidence is that you could set a densitometer to any desired level. Now the definition of gray levels has been amended to claim matter appearing on page 5 lines 20 to 23, that is to say that a specific kind of densitometer is provided to a certain level which gives the desired differences in gray levels.
Now presumably it is intended that the claim be read in the light of the definition and so that the claim by incorporation now is claiming matter contained on page 5 lines 20 to 23 of the specification. We submit that the claim now includes as a characteristic of the invention a feature as to which the specification as lodged was wholly silent and in accordance with the criteria laid down in re Mond Nickel is not fairly based."
The amended definition which Mr Yates referred to reads:
" `gray levels' means controlled variations in density of an exposed recording medium which can be distinguished by measurement with a Tobias Model TBX diffuse densitometer of sensitivity capable of measuring steps of 0.01 optical density units, that being the threshold of average human eye perception".
The underlining shows the text which was added by the amendments after acceptance of the application.
Although there is evidence on file concerning perceptions of "gray levels", I do not think that that evidence conclusively proves that this added material changes the scope of the definition. I can legitimately speculate that the specification of the make and model of the densitometer is irrelevant, given that the densitometer defined is one which is "of sensitivity capable of measuring steps of 0.01 optical density units". I can also speculate that a skilled addressee, faced with the original definition, would assume that the "diffuse densitometer" had a sensitivity comparable with that of the average human eye, rather than an above or below-average human eye. Although speculation is not evidence, there is no evidence on file from which I can conclude that the amended definition is different in scope from the original definition.
SUMMARY
The evidence has not established any of the grounds of opposition. I accordingly dismiss the opposition.
COSTS
The specification was amended significantly after the notice of opposition was filed. In respect of costs, Mr Catterns submitted:
"This is a case where there has been an amendment to the specification, but the evidence didn't really change. If my friends had abandoned their case after the amendment then there would have been a case perhaps for us to pay their costs, or at least not to have our costs up to the amendments. Where there has been an amendment and the opponent has batted on fully forward, including use of all the evidence filed before amendment, I submit that all of the costs should be ours, and if that is wrong, then all of the costs after the amendment."
The amendments to claim 1 were critical to my finding in favour of the applicant. Accordingly, I award costs against the applicant and in favour of the opponent up to the date of advertisement of those amendments, that is 23 August 1990. I award costs against the opponent and in favour of the applicant after that date.
Allen Evans
Delegate of the Commissioner of Patents
Patent attorneys for the applicant : Spruson & Ferguson, Sydney
Patent attorneys for the opponent : Griffith Hack & Co, Sydney
0