Creo Products Inc. v. Presstek, Inc. (166 F. Supp. 2d 944, 2001)

MEMORANDUM OPINION

SLEET, District Judge.

I. INTRODUCTION

On August 12, 1999, Creo Products, Inc. (“Creo”) filed a declaratory judgment action in which it claimed that two patents belonging to Presstek, Inc. (“Presstek”), United States Patent Nos. 5,163,368 (“the ’368 patent”) and 5,174,205 (“the ’205 patent”) are not enforceable and that it does not infringe them. Presstek filed counterclaims on September 22, 1999, and is currently arguing that Creo induced infringement. The court has issued two Markman orders construing disputed limitations of asserted claims of the patents at issue (D.I. 64-65) and denied two of Creo’s three summary judgment motions in a memorandum opinion (D.I. 146).

The court conducted a five day bench trial from June 25 to June 30, 2001 on the issue of liability. After hearing the evidence at trial, reviewing the record, considering the parties briefs on various issues and their proposed findings of fact and conclusions of law (P.F.F.C.L.), the court finds that both the ’368 and ’205 patents are valid and that Creo did not induce infringement. Additionally, the court will deny Creo’s motion for summary judgment for impermissible claim broadening on the reexamination of the ’368 patent (D.I. 92). Pursuant to Federal Rule of Civil Procedure 52(a), this memorandum opinion contains the court’s findings of fact and conclusions of law.

II. FINDINGS OF FACT

A. Background

1. The Parties

Creo is a Canadian corporation having a principle place of business in British Columbia, Canada. In general, Creo manufactures and sells imaging systems with are designed to be installed in printing presses. Presstek is a Delaware corporation with its principal place of business in Hudson, New Hampshire. Presstek also sells equipment for use in on-press imaging. Presstek is the assignee of the ’368 and ’205 patents.

2. Overview of the Technology

From its inception, the printing press has undergone many changes and improvements. Although technology has made printing faster, cheaper, and more accurate, the underlying concept remains the same. Basically, an image is created on a printing plate with two different surface characteristics. When ink is applied to the printing plate, it attaches to certain areas and is repelled from other areas. When the inked printing plate comes into contact with paper, the image is transferred to the paper. Among the various types of printing processes currently in use, one of the most common methods is offset lithography. This printing procedure can rely on the mutual repulsion of oil and water. Much like a photographic negative, printing plates (sometimes containing a chemical coating) are exposed to a discharge source so that they have ink receptive (oleophilic) areas and ink repellant (oleophobic) areas. The printing plate is mounted on a cylinder, inked, and the image is transferred to paper.

Color printing is a bit more complicated. To make a color image, the image must first be separated and decomposed into one of the four basic colors — yellow, cyan, magenta or black. After color separation, the portion of the image representing one of the four colors is transferred to one of four printing plates, one for each color. Each printing plate is then loaded onto a different cylinder at one of four print stations. When paper passes each print station, the ink is transferred from the printing plate to the paper. The result is a full color image. One of the major problems associated with color offset lithography is that unless the separately printed colors are precisely aligned and overlaid on the paper, the resulting image will appear distorted or discolored in various ways. This phenomenon is referred to as misregistration or “out of register.”

Different types of errors contribute to misregistration of images. Generally, errors can be either uniform (affect the entire image) or non-uniform (affect parts of the image). A non-uniform error happens when there are imperfections on the printing plate itself. Uniform errors occur for various reasons. An axial error may occur when the image from one printing plate is shifted to the left or right vis á vis the image from another plate. An angular offset error may happen when the image from one printing plate is shifted up or down with respect to the image from another plate. A size error may occur when the image from one printing plate is larger or smaller than the image from another plate. A skew error occurs when the image on one plate is “twisted” (i.e., rotated) relative to an image on another plate. The naked eye is extremely sensitive to these problems; misalignment of an image by as little as one hundredth of an inch results in a visibly distorted image.

Although misregistration can be corrected manually by a skilled press operator, doing so is both expensive and time consuming. Further, although it is possible to manually correct uniform (or linear) errors, it is extremely difficult to manually correct non-uniform errors. Electronic correction is the preferred method since, in addition to being faster and cheaper, it can correct both uniform and non-uniform errors. The key to electronic registration is creating the image directly on the plate on the press, (i.e., applying the images on the plates at the print stations rather than creating them off press and loading them onto the print stations). Using electronics, the image on the plate is created in such a way as to account for the imperfections of the press (both linear and non-linear). Therefore, although the image on the printing plate may be distorted, the resulting image, when transferred to paper, is in register. The patents-in-suit both describe ways to do this.

3. The Patents-in-Suit

a. The ’205 Patent

The only independent claims of the ’205 patent asserted by Presstek at trial are apparatus claim 11 and the corresponding method claim 23. Since the other asserted claims are dependent on claims 11 and 23, the court will only discuss these two claims.

The claimed invention of the ’205 patent requires a “controlling means” that has “the ability to cause corrections to be effected on a point-by-point basis.” The court has previously construed the term “controlling means.” The function is “producing on the lithographic plate an array of image spots suitable for reproduction.” The structure is “a microprocessor, micro-controller, or algorithmic state machine which performs the algorithm of (1) receiving image data, (2) receiving correction data defining predefined offsets to the image data, and (3) applying the offsets to adjust the timing of the image data, thereby causing the image dots to be created on the printing plate at the desired location.”

The court has also already construed the clause which states “such that the discharges occur in response to the image information at selected discrete positions on the printing surface as specified by the image information offset by the correction data.” This clause requires the controlling means “to have the ability to cause corrections to be effected on a point-by-point basis.” Similarly, with respect to the counterpart method claim, the “controlling” step includes the limitation of requiring “the ability to control the discharges on a point-by-point basis.” The ’205 patent describes point-by-point correction as independent control of the timing of the discharges from each imaging device in a multi-device writing head.

Both claims 11 and 23 of the ’205 patent also include the claim limitation of “a series of axially sequential, circumferential imaging swaths.” Under the terms of the court’s Markman order, this limitation means “a series of image swaths formed sequentially along the axis of the plate cylinder.” According to the system disclosed in the ’205 patent, the imaging head moves in discrete steps parallel to the axis of the plate cylinder during the imaging process. This limitation was added during prosecution of the patent to overcome a rejection based on the prior art. In a memorandum opinion denying Creo’s various motions for summary judgment the court ruled that, during the reexamination of the ’205 patent, Presstek explicitly disclaimed the concept of continuous movement of the imaging head with respect to the plate cylinder. See Section III. C.2.b.1.b, infra (discussing court’s previous memorandum opinion regarding doctrine of equivalents).

The court has also construed the clause “controlling the discharges in accordance with said image, correction, and position information so that the discharges occur at selected positions on the printing surface” as used in claim 23. This clause means “controlling the timing and, therefore, location of the discharges in accordance with image, correction, and position information.” Further, this clause, in combination with the clause “correction information specifying offsets to the image information” as used earlier in claim 23, requires the ability to control the discharges “on a point-by-point basis.”

b. The ’368 Patent

The only independent claim of the ’368 patent asserted by Presstek is claim 1. Since all the asserted dependent claims rely on claim 1, the court will only discuss the various contested limitations of claim 1.

According to the court’s Markman order, the term “discharge source,” as used in claim 1, paragraph a, means “a digitally operated and controlled device utilizing lasers, electromagnetic radiation, electron beams, ink-jet, and/or spark discharge technologies for altering the surface of the plate at selected points to form an image.” Claim 1 also discloses a printing apparatus containing several print stations. Each print station includes a plate cylinder, at least one discharge source for applying an image to the plate, a means for moving each discharge source relative to the plate cylinder and a means for rotating each cylinder.

To correct for imaging errors, the ’368 patent explains that the printing apparatus has a control means that includes a “dot position look-up table for storing the x and y coordinates corresponding to substantially all dot positions on each plate.” The ’368 patent abstract describes the dot position look-up table as part of the controller: “The controller includes a dot-position lookup table for storing the x and y coordinates of substantially all dot positions on the plate.” The dot position look-up table is created during the calibration step that is “performed only once at the factory during the final check-out phase of press manufacture” and is stored “as the pedigree for each of the print stations.” The ’368 patent also describes the dot position look-up table as containing the dot pattern of each plate cylinder so that the locations of all dot positions of each plate are known.

As the court has already stated, the control means also includes a “means for actuating said discharge source to form image dots at selected ones of said dot positions when said electronic signals are present.” The court’s Markman order defined the limitation “means for actuating,” as used in claim 1, paragraph c.ii. The function is “actuating the discharge source to form image dots at selected dot positions when electronic signals representing an original document are present.” The structure is “a computer or controller associated with the discharge source that simultaneously supplies an image signal and a dot-position to the writing head.” The dot position, which is supplied to the writing head, is the correction data that is obtained from the dot position look-up table.

Claim 1 of the ’368 patent also includes a limitation requiring a discharge source. The control means repeatedly actuates the discharge source momentarily based on various inputs. To describe the controller for the spark discharge source, the ’368 patent refers back to the ’205 patent. This spark discharge controller “repeatedly actuates each discharge source momentarily” for each dot.

Although the ’205 and ’368 patents both teach corrections in both the axial (or x) direction and circumferential (or y) direction, both patent specifications describe corrections which are made in the circumferential direction (i.e., around the cylinder). The patents-in-suit do not, however, exclude the possibility of other corrections. For example, claim 1 of the ’368 patent corrects for angular inconsistencies and alters the length of the image scan, both in the circumferential direction.

4. The Accused Products

In 1997, Creo began developing an on-press imaging system (the “DOP System”) for use in a printing press. Portions of the hardware and software used in the DOP System were modified from a previous Creo system called Computer-to-Plate. The DOP System is comprised of several electronic components that interact with each other and with other mechanical press components to permit the imaging of printing plates on press in such a way as to correct for many of the above described imaging errors.

Creo does not incorporate its DOP System into its own printing presses. Rather, Creo sells the DOP System as a kit to other printing press manufacturers. The DOP System is used on the SpeedMaster SM 74 DI (the “SM 74 DI”) which is manufactured, offered for sale, and sold by Heidelberg. The parties dispute whether the SM 74 DI has been or is currently in use in the United States. To support its contention that the SM 74 DI is in use in the United States, Presstek relies entirely on the deposition testimony of Bradley Palmer, an employee of Creo and a former manager of the DOP System. Given the importance of the issue, the court will repeat the relevant portion of the deposition:

Q: Do you know where the Heidelberg DOP presses are installed?

A: We know where they are by country, yes.

Q: Do you have a list of the [amp] users?

A: No, we don’t.

Q: How many are there in the United States?

A: I believe there’s approximately six presses in the United States.

Q: Do you have any more information on where they are located?

A: Yeah. I know from meeting people or public sources where one or two of them are.

Q: And where are they?

A: There’s one in Maryland. There’s another at a company called Photo Effects, which is in the Eastern U.S. somewhere. I believe there was one that is [sic] recently installed in Nevada. I may have heard of others but I don’t recall the specifics at this time.

Q: Do you know if one is installed in Minnesota? ‘

A: Yes. There is one in Minnesota.

Q: Do you know if there is one installed in Delaware?

A: I believe that may be the Photo Effects one.

Palmer Dep. at 74:20-75:19 (emphasis added). Creo argues that Bradley’s knowledge regarding the existence and use of the SM 74 DI in the United States is based on hearsay. Creo also argues that Presstek has not established that the SM 74 DI was in operation in the United States since at least September, 1999.

Having considered the record and Creo’s arguments, the court concludes that Palmer’s statement at the deposition does not stand for the proposition for which it was cited by Presstek. First, there is nothing in the record about the specific timing of the installation of any presses (either in September, 1999 or thereafter). Second, Bradley’s knowledge of the SM 74 DI is not well explained. Palmer states that there are six SM 74 DI presses in the United States. His knowledge of “one or two of them” is based entirely on “meeting people or public sources.” This statement is classic hearsay since it (1) is being offered for the truth of the matter, (2) relies on out of court statements and documents not in the record, and (3) there is no recognized exception to the hearsay prohibition. As to the other three or four SM 74 DI presses, Palmer offers no explanation as to how he knows where in the United States the machines are located. Indeed, he admits that he does not have a list of users of the press. Even giving Presstek the benefit of the doubt — that he learned about the presses from “meeting people or public sources” — this information is still hearsay for the above mentioned reasons. Thus, the court finds that the record does not contain any evidence that the SM 74 DI was made, used, or sold in the United States in September, 1999 or thereafter.

The DOP System is also used in the Komori Project D Press (the “Komori Press”) which is manufactured by Komori Corporation (“Komori”). The Komori press is the result of a joint development effort by Creo and Komori, and it is uncontested that the press exists. At trial, there was a dispute over whether the Ko-mori Press has been offered for sale in the United States. Similarly, there was no evidence presented at trial, either through live testimony or by deposition designations, that the Komori Press has been sold or is currently in operation in the United States. There is similarly no evidence in the record that the Komori Press is currently in use in the United States. Rather, the Komori Press was shown at a printing industry trade show in Chicago, Illinois in 2000. Prior to the trade show, Creo was aware that Komori’s goal was to show the Komori Press at the show and that the “Demonstration Press” was in Chicago one month before a trade show.

a. Imaging in the DOP System

The DOP System is connected to a computer. The computer, although not part of the DOP System, is called the Creo Data Node Interface (CDI). The CDI contains software provided by Creo (the “service shell”). The service shell is a graphical user interface which allows the user to enter five offset parameters to register images on the printing plates (main scan shift, sub scan shift, main scan scale, sub scan scale and track rotation). A sub scan shift relates to the placement of pixels (the smallest element of a picture) along the length of the printing plate (the axial or x-direction). Main scan shifts are movements of the image around the circumference of the printing plate (circumferential or y-direction). Track rotation rotates the axis of the image to compensate for the orientation of the track on which the imaging head moves. These geometric correction parameters are used to register images on the printing plates. The image corrections must be uniform since the DOP System cannot (1) independently control a point or pixel in the image, (2) affect individual points, or (3) perform non-uniform corrections.

The first step in imaging with the DOP System is to send a file of the image to be printed from the CDI to the raster image processor (RIP), located in the Data Node Electronics (DNE). The RIP then breaks the image down into pixels and builds a image for each color. Each image corresponding to a color is saved as a “bitmap” for use in imaging a separate plate on a separate print station. The image bitmap is then stored on a hard disk in the DNE. To account for the fact that the DOP System creates an image on a plate rotating on a cylinder, each bitmap image undergoes an orthogonality/corner turning process. This is done so that the imaging head can create a series of “swaths” on the printing plated. The bitmap is stored in the DNE until the imaging process begins.

The DOP System determines when to begin imaging based on its geometric correction parameters (which are entered into the service shell located in the CDI). Upon reaching the appropriate spot on the plate, the CDI commands that the image bitmap information be read out of the DNE and sent to a 128 byte first in/first out (FIFO) memory located on the Image Control Electronics (ICE) board. Each print station has its own ICE board. As the image data is read out of the FIFO to the imaging head, new information is sent from the DNE to the FIFO memory. Thus, image information is not stored in the FIFO — it is in memory for an instant before being fed to the imaging head.

The DOP System also includes a center frequency calculation (CFC) to maintain uniform spacing of the pixels regardless of speed changes of the cylinder. The DOP System computes a CFC by analyzing the current speed of the cylinder and the acceleration or deceleration of the drum at a given moment, comparing it to data from the previous drum rotation, and then estimating how fast the drum will be moving during the next few clock pulses. Since the CFC is based on present speeds and future estimations, it is not entirely based on entry by an operator, nor does it constitute predetermined data. Since it is merely a timing clock based on present and estimated drum speed, the CFC does not make any non-linear corrections, is not affected by non-uniformities on the plate cylinder, and does not correct on a point-by-point basis. The CFC is updated a fixed number of times during imaging. For example, in the SM 74 DI (the Heidelberg system), the CFC is adjusted 256 times per cylinder rotation — once every 0.1 inches of rotation. Further, there are approximately 240 strokes, corresponding to 50,000 points affected by each CFC.

b. The DOP System’s Imaging Head

The DOP System has a Thermal Exposure Head (TEH) at each print station. The TEH contains a circuit board with microprocessors which manipulate information from the DOP System to produce the image on the printing plate. The discharge source in the TEH is a laser diode and a light valve with up to 240 separate laser channels or beams. The laser remains on throughout the entire imaging process and the light valve channels— which are parallel to the cylinder axis— either open or close to allow laser light to pass through to the printing plate.

Depending on the image to be produced, each pixel is either “on” or “off.” If a pixel is on (a digital “1”), the corresponding channel in the light valve will open, the laser beam will pass through it, and image a pixel on the printing plate. Correspondingly, if the pixel is off (a digital “0”), the corresponding channel will close and the laser beam will not image a pixel on the printing plate. This process results in the image being “painted” onto the printing plate as a series of lines (not discrete dots). Although each pixel may be either on or off, the position of each channel relative to each other remains fixed; the position of a pixel cannot be varied independently on any other pixel on the plate. Consequently, the DOP System cannot perform non-linear (i.e., non-uniform) corrections that result from variations between pixels or a group of pixels.

The DOP System reads the groups of pixels out of the FIFO in groups of up to 240 pixels at a time. This group of pixels is referred to as a “stroke.” The “stroke clock” determines when a stroke of information is imaged on the printing plate. Each pulse of the stroke clock loads one stroke into the light valve. The TEH sits on a track that moves continuously along the length of the plate cylinder (the axial direction). Since the plate cylinder is continuously rotating, the image created traces a helical path around the plate cylinder. The DOP System refers to the helical image created in one rotation of the plate cylinder as a “swath.” If a printing plate is removed from a plate cylinder and laid flat, each swath on the plate is angled.

B. Validity

Creo has asserted that the ’205 and ’368 patents are invalid because they are obvious, violate the on-sale bar, and do not disclose the best mode. Below are the court’s findings of fact on each of these issues.

1. Obviousness

Creo has asserted that the ’368 and ’205 patents are invalid as obvious. At trial, Creo’s expert conceded that the ’368 patent and various claims in the ’205 patent are not anticipated under 35 U.S.C. § 102. Yet, Creo’s expert concluded that the patents-in-suit should be considered obvious in light of a combination of references. Creo seeks to rely on two alleged prior art references to support its claim of obviousness. One reference is U.S. Patent No. 4,911,075 (the “Lewis ’075 patent”). The Lewis ’075 patent is one of Presstek’s first patents and covers its plate technology. The Lewis ’075 patent was cited during prosecution of both the ’368 and ’205 patents. There is no evidence to suggest that, during reexamination, the Examiner failed to consider the Lewis ’075 patent as prior art. Indeed, the P.T.O. regulations on prior art required the Examiner to do so.

The other reference relied upon by Creo was an alleged sale or use of the Optrotech Image 5008 (the “Image 5008”), a printed circuit board film setter, prior to January 9, 1990. There is no documentary evidence in the record which describes the design, structure, or operation of the Image 5008. Rather, the only evidence which suggests that the Image 5008 was on sale, sold, or used in the United States is the testimony of Daniel Gelbart, President of Creo. Gelbart testified the product was made in Israel using Creo components made in Canada. Gelbart stated that the Image 5008 was in operation at various locations in the United States in 1989. He also stated that an Optrotech technician was stationed at Creo to do service calls in the United States and that he would assist in such service calls. Creo does not, however, have any sales documents, import/export documents, invoices, shipping records or other documents of any sort in the record to substantiate any such sale in the United States. According to Gelbart, these types of documents should exist some place. His only memory of documentation was seeing an operator manual many years ago but he does not possess a copy.

The trial testimony revealed that initial demonstration of the Heidelberg GTO-DI (the “GTO-DI”), the first press to incorporate the inventions in the ’205 and ’368 patents, was at the Print ’91 trade show in Chicago in September, 1991. The GTO-DI was very well received by representatives of the printing industry at the trade show. Indeed, the GTO-DI was demonstrated to a “packed house” and there was so much interest that additional stands were erected to permit people to stand behind those seated to view the press demonstration. The Heidelberg QuickMaster 46 DI, which also incorporated the inventions of the ’205 and ’368 patents, was likewise well received when it was demonstrated at the DRUPA trade show in 1995.

Creo’s own documents provide laudatory comments about the GTO-DI and Quick-Master 46 DI presses. United States Patent No. 5,713,287, owned by Creo and naming Gelbart as the inventor (the “Creo ’287 patent”), states that a major advantage of the GTO-DI press compared to presses using plates made off-press is much better registration between printing units when printing color images. This is precisely the function of the patents-in-suit. The Creo ’287 patent states that the GTO-DI press eliminates inefficient and expensive bottle-necks in printing operations associated with using plates that are made off-press. Further, Creo’s own marketing documents state that the QuickMaster 46 DI is a “huge success.” Finally, Creo—through its CreoScitex division— has a royalty bearing license under the patents-in-suit for the manufacture and sale of its imaging system used on its 74 Karat press. In addition to the license to CreoScitex, the inventions covered by the patents-in-suit have also been licensed to third-party press manufacturers including Heidelberg, Adast, and Ryobi.

2. On-Sale Bar

a. Presstek’s Technology

Presstek’s first device for imaging plate material was a handheld device with a spark discharge needle which was manually moved over the surface of the plate material as a way to test the effect of sparks on different configurations of materials to form the plates. The second device was a small wheel on which spark images were deposited by a single electrode on a thin strip of plate material. In mid-1988, Presstek created an imaging head which had multiple spark discharge needles and began assembling a “feasibility press” configured with a central impression cylinder for the purpose of continuing its feasibility tests.

The feasibility press was assembled by Presstek from parts taken from used printing presses and pieces built in the Presstek laboratory. The feasibility press was never designed as a product nor intended as a product in itself. Instead, it was created as an experimental device (albeit more sophisticated than the prior handheld device) to image Presstek’s continually evolving printing plate designs. According to the trial testimony, the “main purpose” of the feasibility press was to test and finalize the development of the imaging system and a printing plate. Presstek only made one feasibility press which was used to generate test images on the variety of printing plate materials with which Presstek was experimenting. The feasibility press is still in Presstek’s laboratory today.

At no time during 1988 or 1989 did the feasibility press include all the imaging systems or controls which are described and claimed in the ’205 or ’368 patents. Creo’s argument regarding the disclosures of U.S. Patent No. 4,936,211 (“the ’211 patent”) does not demonstrate that Pres-stek was aware of what controls and corrections the ’205 and ’368 patents would ultimately claim. Rather, the poor gearing of the feasibility press made it too imprecise. The feasibility press did not have individual encoders on individual plate cylinders; it had a single magnetic encoder on the central impression cylinder. Since the central impression cylinder was geared to the plate cylinders, there was only one reading indicating the expected position of all the plate cylinders from this one central source.

The feasibility press was unable to do several things required by the ’205 and ’368 patents. The presence of one magnetic encoder on the feasibility press did not provide sufficient individual information to determine angular inconsistencies among the plate cylinders or for making angular adjustments among the cylinders to bring the images into register. On the contrary, image adjustments in the circumferential direction done on the press were done manually and mechanically. The angular inconsistencies among the cylinders on the press were also corrected by mechanically adjusting the cylinders as in a conventional press. Further, the feasibility press had no ability to change the size of the image in the circumferential direction to achieve registration. In this sense, the feasibility press (and the imaging for that press) was not a commercially saleable product.

b. Potential Development Partners

At trial, Creo asserted that two business relationships between Presstek and potential joint development partners violate the on-sale bar. The court, however, finds that Presstek did not offer to sell any product incorporating all the inventions of the ’368 and ’205 patents prior to January 9, 1990 (the “critical on sale bar date”). The court will discuss these associations in turn.

(1) Multigraphics

In late 1988, Presstek sought a partner or partners who would be willing to devote resources to the joint development of a press with on-press imaging; it was not trying to sell a press or imaging equipment. Every potential development partner was required to sign a very stringent Non-Disclosure Agreement so that whatever Presstek did with such companies remained secret and confidential and could not be published, disclosed, or used by them. Presstek entered into such an agreement with Multigraphics. In November, 1988 the parties began discussions to jointly develop the Presstek system. While Multigraphics was interested in the concept, it ultimately decided not to proceed with Presstek. As a result, the development agreement with Multigraphics terminated and no business or sale was ever consummated.

In late January or early February, 1989, Presstek signed a series of letters with Multigraphics laying out a proposed business relationship between the parties. As part of the business relationship, Pres-stek hoped to supply its imaging systems to Multigraphics for incorporation into Multigraphics’ one and two color Eagle Presses. According to the terms of the letter, Presstek agreed to furnish Multi-graphics with an imaging system for such one and two color presses. The letter explicitly states that the imaging system is to be further described in specifically referenced exhibits. None were attached to the letter. Other than a Confidentiality Agreement, no such exhibits were otherwise created. The state of the letter agreement with Multigraphics was summarized in the Registration Statement for Presstek’s public offering of stock in 1989 as follows:

The Company has entered into an agreement with Multigraphics pursuant to which it will furnish Multigraphics with the Prepress System and the Direct Imaging System, integrate such systems into Multigraphics’ line of one and two color Eagle presses and license Multigraphics to use the associated proprietary software in such modified Eagle Presses at prices and on terms which the parties have agreed to negotiate and use their best efforts to complete - as soon as the relevant information becomes available.

The Registration Statement also makes clear that:

There can be no assurance that development of the Presstek Printing System will be successfully completed, that it will satisfactorily perform all of the functions for which it has been designed or that it will meet current price/performance objectives.

As stated above, the definition of what would be supplied and all the actual significant terms and conditions of a potential agreement were never determined. Specifications, pricing, and other schedules for the agreement were never prepared.

Further, the development proposed in the January, 1989 letter does not appear to encompass the color image printing system Presstek later invented, disclosed, and claimed in the ’368 and ’205 patents. Rather, a one-color Eagle press would print one color only, normally black. In a one-color press, there is only a single plate cylinder and a single plate; there is no need or use for control systems of the type in the ’368 and ’205 patents. Similarly, a two-color press is a “spot color” press; it is designed for printing black text with, for example, a one colored border or logo. In such a press, the dimensional factors corrected by the ’368 and ’205 patent inventions for color images have no applicability.

The companion January 31, 1989 letter (PTX 102), which proposed that Multi-graphics would supply Presstek with parts and components from its Eagle press for Presstek to develop a four-color press with on-press imaging of its own, was also never consummated. The schedules referred to in that agreement were never created. In any case, such a transaction would have involved only Presstek’s purchase of parts and components from a vendor, under a strict Confidentiality Agreement, for its own internal development of a four-color press.

(2) Heidelberg

In the fall of 1989, Presstek did a demonstration of its feasibility press for a representative of Heidelberg. After the demonstration, Presstek entered into a strict Confidentiality Agreement with Heidelberg and agreed to do a demonstration of Presstek’s plates and spark discharge energy for Heidelberg in Germany. Williams and four other Presstek engineers traveled to Germany to demonstrate Presstek’s plate and imaging technology to Heidelberg. In advance of this trip, Pres-stek shipped to Germany a laboratory fixture designed to image plates. That fixture was a single imaging device with a single plate cylinder, which could image Presstek’s specially prepared plates for testing. Not only was it not a printing press, it was neither sold nor leased to Heidelberg. The Presstek engineering team spent approximately three weeks at Heidelberg. Upon completion of the demonstration, the fixture was shipped back to Presstek.

In early 1990, Presstek began developing a press for Heidelberg embodying Presstek’s imaging technology. In late spring of 1990, Heidelberg sent Presstek a five color GTO press. Unlike the feasibility press, the GTO press was a commercial product with commercially acceptable print quality. During this time, Presstek converted the GTO press into an on-line Direct Imaging press (which became the GTO-DI). Several of the aspects of the claimed invention were not incorporated into the GTO press until this time. The GTO-DI was shown to the public for the first time at Print ’91 in September, 1991. This marked the first time a product embodying the ’368 and ’205 patents was ever offered for sale.

3. Best Mode

There is no evidence on the record that Presstek withheld the best mode of practicing the claimed inventions of the patents-in-suit. Both Williams and Pensa-vecchia testified that the inventions of the ’205 and ’368 patents were not incorporated in the feasibility press. The court finds the testimony of both witnesses to be credible and trustworthy. Williams did not personally design the feasibility press. At the time, however, he was the chief operating officer who oversaw the technical developments of the engineering and design teams. Although he may not have been intimately familiar with the details of the feasibility press, he had sufficient knowledge to testify about whether the inventions of the patents-in-suit were included in the feasibility press. At minimum, the ’211 patent specifications do not contradict the credible and trustworthy testimony of Williams and Pensavecchia on this subject.

C. Inequitable Conduct

Creo argues that the facts demonstrate that Presstek’s change in inventorship of the ’368 patent constitutes inequitable conduct since doing so removed “material” subject matter contained in the ’211 patent from the P.T.O. examination. The facts, however, are open to a different, and more plausible interpretation. The original ’368 patent was filed with 94 pending claims. In a first Office Action, the P.T.O. stated that the claims fell into one of four distinct categories and were, therefore, subject to a restriction requirement (i.e., Presstek could continue to prosecute only one category of claims in the pending application). In response to the restriction requirement, Presstek elected to prosecute claims 1-24 and 68. This restriction requirement also appears to have resulted in the change of inventors in the ’368 patent. Thus, there is nothing in the record to indicate that the inventorship change showed that the inventors were aware that the ’211 patent was material to the patentability of the ’368 patent application.

In addition, Creo’s a fortiori logical chain that the P.T.O. Examiner would have considered the ’211 patent material is not supported by the record. The court does not agree that the ’211 patent describes a product (the feasibility press or another imaging prototype) that was on-sale or offered for sale before the critical on-sale bar date. The trial testimony by various Presstek employees does not prove otherwise. Thus, there was no “material” withholding of information. Even if there were a product on-sale embodying the ’211 patent, it may not have been considered material. Indeed, the P.T.O. Examiner considered the ’211 patent during the prosecution of the ’205 patent but did not reject any claims on this basis.

D. Infringement

In its counterclaim, Presstek asserts that the Creo DOP System induces infringement of (1) independent claim 1 of the ’368 patent (and the dependent claims) and (2) independent claims 11 and 23 (and the dependent claims) of the ’205 patent. Each claim contains several limitations. Some limitations are stipulated as uncontested, some are contested in passing, and some are bitterly contested. Questions of infringement require the court to make both legal determinations (i.e., construing the claims of the patents-in-suit) and factual determinations (comparing the accused products to the construed claims). Rather than engage in this hybrid two step analysis twice, the court will both repeat its claim construction and discuss the accused devices below. See Section III.C, infra.

III. CONCLUSIONS OF LAW

The court will first address the issue of validity, including Creo’s motion for summary judgment on claim broadening on reexamination. The court will then examine whether the patents-in-suit are unenforceable. Next, the court will discuss whether Creo induced infringement of the patents-in-suit. Upon reviewing the parties’ arguments and the relevant case law, the court finds that (1) Creo has not shown, by clear and convincing evidence, that the patents-in-suit are invalid, (2) Creo has not adduced sufficient evidence to show that the inventors of the patents-in-suit engaged in inequitable conduct, and (3) Presstek has failed to demonstrate, by a preponderance of the evidence, that Creo induced infringement of the patents-in-suit. The court will explain the bases for its rulings.

A. Validity

1. Claim Broadening On Reexamination Of The ’368 Patent

On January 5, 2001, Creo filed a motion for summary judgment in which it argues that claims 1-20 of the reexamined ’368 patent are invalid because the claims were broadened on reexamination by the P.T.O. (D.I. 92). In a previous memorandum opinion, the court took the motion under advisement. See Creo, 2001 WL 637397, at *12. The court has fully considered the motion and concludes that claims 1-20 of the ’368 patent were not improperly broadened upon reexamination. In explaining its decision, the court will first outline the ’368 patent prosecution history and then discuss the applicable law.

a. Patent Prosecution History

Presstek filed an application for what was to become the ’368 patent on January 9, 1991. In the original application, claim 1 did not include any limitations regarding image correction. The P.T.O. rejected claim 1, among others, under 35 U.S.C. § 103 as being obvious in light of U.S. Patent Nos. 4,718,340 (the “Love Patent”), 4,524,364 (the “Bain Patent”), or 4,835,544 (the “Winterburn Patent”). See Def. Ans. Br. Sum. J. Claim Broadening, App. C at C1873.

In response to the P.T.O.’s initial rejection of claim 1, Presstek substantially amended it by, among other things, adding limitation c.iii which states “means for offsetting, with respect to said x and y coordinates, the action of the discharge-source actuation means to correct imaging errors.” See id. In justifying its amendment, Presstek stated:

We have added new limitations to claim 1 to overcome the rejections over the primary references Love, III and Lewis et al., and the secondary references cited in connection therewith. Specifically, the primary art references, alone or in combination with the secondary references do not discuss or suggest offsetting control signals in order to correct image length, registration or skew errors.

Id. at C1995. The parties dispute the exact meaning of this language and what it suggests as to why Presstek proffered the amendment. In any event, the P.T.O. allowed the amendment and issued the ’368 patent on November 17,1992.

On July 28, 1998, Creo filed a Request for Reexamination of the ’368 patent with the P.T.O. Upon reexamination, the P.T.O. rejected claim 1 of the ’368 patent in light of the prior art. In response, Presstek amended claim 1 by adding language to limitation c.iii and adding limitation c.iv. Perhaps the best way to demonstrate the variations of limitation c of claim 1 of the ’368 patent is to reproduce the relevant portions of both the original and the reexamined language. The original language of limitation c.iii of claim 1 read:

means for offsetting with respect to said x and y coordinates, the actions of the discharge-source actuation means to correct imaging errors.

After the P.T.O. rejected the above language, Presstek added language to limitation c.iii and added a new limitation e.iv which teaches:

iii means for offsetting with respect to said x and y coordinates, the actions of the discharge-source actuation means in accordance with the angular offset parameters to correct the angular inconsistencies; and

iv means for altering the length of the scan in accordance with the size difference parameters to correct the image size inconsistencies.

The court’s Markman ruling construed several phrases in the ’368 patent claims. Among the terms it construed was a clause in limitation c.ii which reads “angular inconsistencies among plate cylinders” to mean “inconsistencies in the rotational positions of the plate cylinders that cause printed images to be out of alignment in the y direction.” See D.I. 64 at ¶ 6. The court also added a footnote which briefly outlined the contours of the present dispute but declined to reach the issue at that time. See id. at n. 1. In the footnote, the court stated:

The term angular is consistently used in the patent to refer to the rotational or circumferential position [but that] .... there is some tension between [the aforementioned] construction of ‘angular’ inconsistencies and the claim language that requires offsetting ‘with respect to said x and y coordinates’ .... This tension is somewhat mitigated when one considers that ‘x’ and ‘y’ coordinates work together (i.e., as a unit) to define a single location on a plate. Thus, moving a particular image spot in the ‘y’ direction could reasonably be said to require an adjustment to its ‘x and y coordinates’.

Id. It appears that in its Markman order, the court accepted Presstek’s construction of “angular inconsistencies” but allowed Creo to argue at a later date that such a constriction is an impermissible broadening of claim 1 of the ’368 patent.

b. Discussion

Neither party disputes that claims that are broadened on reexamination are invalid. See 35 U.S.C. § 305; Quantum v. Rodime, PLC, 65 F.3d 1577, 1580 (Fed.Cir.1995). Further, the parties agree that the proper method for the court to determine whether a claim was broadened on reexamination is (1) to determine the scope of the original claim and (2) determine the scope of the reexamined claim. See In re Freeman, 30 F.3d 1459, 1464 (Fed.Cir.1994). As with traditional Markman determinations, the court must first look to the intrinsic evidence (the claims, the written specification, and the prosecution history) and then may consider the extrinsic evidence. See, e.g., Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582-84 (Fed.Cir.1996). A reexamined claim is considered impermissibly broad if it “is broader in any respect [than the original claim] ... even though it may be narrower in other respects.” See Freeman, 30 F.3d at 1464 (citation omitted). Rather than debating the contours of the law, the parties focus their attention on the interpretation of the original and the reexamined claim 1.

The crux of the issue, therefore, is how to construe the “means plus function” language of original claim 1, limitation c.iii as permitted by 35 U.S.C. § 112, ¶ 6. To interpret “means for offsetting ... to correct imaging errors” .the court must (1) identify the function defined in the claim element and (2) identify the structure in the specification that corresponds to that function. The parties dispute both steps but agree that both require resort to the patent specifications.

The parties differ whether the court should read the patent specifications into limitation c.iii of claim 1. Creo maintains that the original claim 1, limitation c.iii required correction for all imaging errors identified in the patent specification while reexamined claim lc.iii and c.iv merely require correction of a “subset” of the imaging errors — they do not require the ability to correct for errors in the axial direction. According to this line of reasoning, “a device that has no capacity of correcting for imaging errors in the axial direction would not infringe the original claim, but would infringe the reexamined claim.” See PI. Op. Br. Sum. J. Claim Broadening at 2. Pres-stek, on the other hand, argues that the specifications describing imaging errors are “merely examples” of various problems and how to correct for them. See Def. Ans. Br. Sum. J. Claim Broadening at 11-12 (citing language from patent specification).

When the court looks to patent specifications to assist in interpreting claims, it must refrain from reading a limitation from the written description into the claim. See Tate Access Floors, Inc. v. Maxcess Technologies, Inc., 222 F.3d 958, 966 (Fed.Cir.2000) (citing, inter alia, Renishaw PLC v. Marposs Societa’ per Azioni, 158 F.3d 1243, 1248 (Fed.Cir.1998)); Kemco Sales, Inc. v. Control Papers Co., Inc., 208 F.3d 1352, 1362 (Fed.Cir.2000) (cited in Tate); see also Sjolund v. Musland, 847 F.2d 1573, 1581 (Fed.Cir.1988) (“[WJhile it is true that claims are to be interpreted in light of the specification and with a view to ascertaining the invention, it does not follow that limitations from the specification may be read into the claims _”); SRI Intn’l v. Matsushita Elec. Corp. of Am., 775 F.2d 1107, 1121 (Fed.Cir.1985) (“[C]laims are construable ... in light of the specification, yet ‘[t]hat.. .does not mean that everything expressed in the specification must be read into all the claims.’ ”) (quoting Raytheon Co. v. Roper Corp., 724 F.2d 951, 957 (Fed.Cir.1983)) (citations omitted).

Given the above, Creo’s motion for summary judgment must fail. In its motion, Creo argues that the “means for offsetting” limitation of original claim 1 must correct for all the imaging errors discussed in the ’368 patent. This argument, however, falls into the trap of redefining the stated function of the claim element in light of the preferred embodiment. As previously discussed, the purpose and effect stated for performing the offsetting function of original claim 1 is “to correct imaging errors.” The term “imaging errors” is a general reference to all types of errors associated with imaging and is not limited to those discussed in the ’368 patent specifications.

Further, the patent prosecution history demonstrates that Creo’s interpretation of claim 1 is contrary to the other claims of the ’368 patent. Dependent claim 2 of the originally issued ’368 patent was directed to image size correction in the circumferential direction. This is one of the imaging errors that Creo asserts is covered by the language in original claim 1. If, as Creo asserts, the “means for offsetting” language in original claim 1 was already limited to require correction for all imaging errors, then dependant claim 2 would be redundant and meaningless. This interpretation would violate the doctrine of claim differentiation. See Kraft Foods, Inc. v. Int'l Trading Co., 203 F.3d 1362, 1368 (Fed.Cir.2000) (stating that claim differentiation creates presumption that each patent claim has different scope).

The limitation “means for offsetting” is best understood by looking at the patent prosecution history. As originally filed, claim 1 was did not include the capability to shift or adjust the length of images; such capabilities were found in dependent claims 2 and 3. As outlined above, the P.T.O. rejected original claim 1 in light of the Love, Bain, or Winterburn patent. To overcome the rejection, Presstek amended the claim (and specifically limitation sub-part iii) to include imaging errors generieally, something not disclosed in the cited prior art references. Indeed, Pres-stek stated that “the primary references do not discuss or suggest offsetting control signals in order to correct image length, registration or skew errors.” Thus, the original claim 1 did not require all image corrections. Since original claim 1 did not require all image corrections, it follows that the reexamined claim 1 is not impermissibly broad. The court, therefore, will deny Creo’s motion for summary judgment on this ground.

2. Obviousness

Section 103 of the Patent Act prohibits the patenting of an “obvious” invention. This section provides, in relevant part:

A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.

35 U.S.C. § 103(a). The relevant considerations were announced by the Supreme Court over thirty years ago:

Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobvi-ousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness, these inquiries may have relevancy.

Graham v. John Deere Co., 383 U.S. 1, 17-18, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966); see also Apple Computer, Inc. v. Articulate Sys., Inc., 234 F.3d 14, 26 (Fed.Cir.2000) (stating that considerations include “(1) the scope and content of the prior art; (2) the differences between the claimed invention and the prior art; (3) the level of ordinary skill in the art; and (4) objective evidence of non-obviousness, such as commercial success, long-felt but unsolved need, failure of others, copying, and unexpected results.”). Because a patent and its claims are presumed valid, see 35 U.S.C. § 282, the party challenging a patent — in this case, Creo — must establish the facts supporting a determination of invalidity by clear and convincing evidence. See Richardson-Vicks, Inc. v. Upjohn Co., 122 F.3d 1476, 1480 (Fed.Cir.1997).

When making an obviousness analysis based on prior art, courts must not fall prey to a “hindsight syndrome” by reasoning backward from the teaching of the patent itself. See In re Kotzab, 217 F.3d 1365, 1369 (Fed.Cir.2000). According to the Federal Circuit, “the best defense against the subtle but powerful attraction of a hindsight-based obviousness analysis is rigorous application of the requirement for a showing of the teaching or motivation to combine prior art references.” See In re Gartside, 203 F.3d 1305, 1319 (Fed.Cir.2000); see also B.F. Goodrich Co. v. Aircraft Braking Sys. Corp., 72 F.3d 1577, 1582 (Fed.Cir.1996). In other words, something in the prior art, considered as a whole, must “suggest the desirability, and thus the obviousness, of making the combination” of different elements to create the invention. See Fromson v. Advance Offset Plate, Inc., 755 F.2d 1549, 1556 (Fed.Cir.1985) (citation omitted).

A device that was in public use or on sale in the United States more than one year prior to the application date for a patent — in this case January 9, 1990—may be prior art under 35 U.S.C. § 102(b). In Finnigan Corp. v. Int’l Trade Comm’n, the Federal Circuit described how courts have long looked with disfavor upon invalidating patents merely on the basis of testimonial evidence. 180 F.3d 1354, 1366 (Fed.Cir.1999) (stating that “[t]he Supreme Court recognized over one hundred years ago that testimony concerning invalidating activities can be ‘unsatisfactory’ due to ‘the forgetfulness of witnesses, their liability to mistakes, their proneness to recollect things as the party calling them would have them recollect them, aside from the temptation to actual perjury.’ ”) (quoting The Barbed-Wire Patent, 143 U.S. 275, 284, 12 S.Ct. 443, 36 L.Ed. 154 (1892)). The Finnigan court also further clarified that corroboration is required. See Finnigan 180 F.3d at 1366 (finding that “the need for corroboration exists regardless whether the party testifying concerning the invalidating activity is interested in the outcome of the litigation (e.g., because that party is the accused infringer) or is uninterested but testifying on behalf of an interested party.”); see also Woodland Trust v. Flowertree Nursery, 148 F.3d 1368, 1371 (Fed.Cir.1998) (stating that “[c]orroboration of oral evidence of prior invention is the general rule in patent disputes.”).

In this case, Creo asserts that the Lewis ’075 patent and the Image 5008 are two prior art references that make the patents-in-suit invalid as obvious prior to January 9, 1990. Given the court’s findings of fact and the above discussed case law, the court need not tarry long on this issue. Contrary to Creo’s assertion, there is no clear and convincing evidence in the record to suggest that, during reexamination, the P.T.O. Examiner failed to consider the Lewis ’075 patent. See note 14 & accompanying text, supra. As for the Image 5008, Creo relies entirely on the testimony of Gelbart to prove that the product was sold or was in operation in the United States prior to January 9, 1990. This type of oral and uncorroborated testimony is precisely what the Federal Circuit in Finnigan cautioned the courts against using to find obviousness. Not only is Gelbart, as President of Creo, an interested party, but his testimony is vague as to products themselves. For example, Gelbart did not state where in the United States the Image 5008s were located, who used them, or how they were operated.

On the other side of the coin, the record indicates the GTO-DI and the QuickMas-ter 46 DI, both of which incorporated the patents-in-suit, were commercial successes. Not only were both products very well received when they debuted at trade shows, but third parties, including Creo, have said complimentary things about the products and even licensed the patents-in-suit. This evidence of commercial success of the first two products incorporating the ’368 and ’205 patents somewhat counterbalances Creo’s evidence regarding obviousness. Since the court finds that Creo has failed to prove, by clear and convincing evidence, that the Image 5008 and the Lewis ’075 rendered the patents-in-suit obvious, the court declines to invalidate them on this basis.

3. On-Sale Bar

The court has already set forth the legal standard governing the on-sale bar of 35 U.S.C. § 102(b). See Creo, 2001 WL 637397, at *2 (citing and discussing cases). Importantly, “[w]hether an invention was on sale more than one year before the patent’s application date is a question of law for the court to decide based on underlying factual determinations.” See id. Put simply, Creo has the burden to prove, by clear and convincing evidence, that prior to January 9, 1990(1) a product embodying the claimed invention was the subject of a commercial offer for sale and (2) the claimed invention was ready for patenting. See id; see also Pfaff v. Wells Elecs., Inc., 525 U.S. 55, 57, 119 S.Ct. 304, 142 L.Ed.2d 261 (1998). Both conditions must be satisfied. See id. The court finds that Creo has failed to meet its burdens under Pfaff. Therefore, the court declines to find the patents-in-suit invalid in violation of 35 U.S.C. § 102(b). The court will address these issues in turn.

a. First Pfaff Prong

To satisfy the first condition stated in Pfaff, Creo “must demonstrate by clear and convincing evidence that there was a definite sale or offer to sell more than one year before the application for the subject patent, and that the subject matter of the sale or offer to sell fully anticipated the claimed invention or would have rendered the claimed invention obvious by its addition to the prior art.” See STX, L.L.C. v. Brine, Inc., 211 F.3d 588, 590 (Fed.Cir.2000) (quotation and citation omitted); Vanmoor v. Wal-Mart Stores, 201 F.3d

1363, 1366 (Fed.Cir.2000) (stating that invalidating sale must involve device that “actually embodied or rendered obvious the patented invention”) (quoting Evans Cooling Sys., Inc. v. General Motors, Corp., 125 F.3d 1448, 1451 (Fed.Cir.1997)); Tec Air, Inc. v. Denso Mfg. Michigan Inc., 192 F.3d 1353, 1358 (Fed.Cir.1999). Thus, “[t]he first determination in the § 102(b) analysis must be whether the subject of the barring activity met each of the limitations of the claim, and thus was an embodiment of the claimed invention.” See Sc