Desper Products, Inc. And Spatializer Audio Laboratories, Inc. v. Qsound Labs, Inc.

PLAGER, Circuit Judge. Desper Products, Inc. and Spatializer Audio Laboratories, Inc. (collectively “Desper”) sued QSound Labs, Inc. (“QSound”), owner of U.S. Patent Nos. 5,105,462 (the ’462 patent) and 5,208,860 (the ’860 patent), in the United States District Court for the Central District of California (No. CV 94-7276). Desper sought a declaratory judgment that Desper’s “Spatializer” product does not infringe either patent and that the patents are invalid. QSound counterclaimed for patent infringement. The district court entered summary judgment for Desper, and QSound appeals. Because the district court correctly interpreted the disputed claim language and properly concluded that QSound surrendered subject matter within which Desper’s “Spa-tializer” product falls, we affirm the judgment for Desper. BACKGROUND This case involves two patents, the ’462 patent and the ’860 patent, both entitled *1328 “Sound Imaging Method and Apparatus.” At the trial level, the case was assigned, with the consent of the parties, to a special master. The parties filed cross-motions for summary judgment with regard to the question of infringement. After considering both parties’ submissions, and after conducting an extensive hearing during which both sides put on expert testimony, the special master recommended that judgment of non-infringement be entered in favor of Desper, with one exception in which the special master deemed the infringement of certain claims to involve a disputed question of fact precluding summary judgment. The district court adopted all of the special master’s recommendations, except the one in which the master thought there existed a genuine dispute of material fact. The district court concluded that Desper was entitled to summary judgment of non-infringement on all the asserted claims. The district court entered final judgment of non-infringement in .Desper’s favor under Fed.R.Civ.P. 54(b). QSound now appeals that ruling. A. The Patented Technology Both of QSound’s patents stem from a common application and, as such, have a common written description; only the claims differ between the two. The patents describe an audio sound image location system that processes a monaural sound signal in such a way as to create an illusion that the source of the sound is located someplace in a three-dimensional space. The patented invention gives an audiophile the ability to locate individual sounds at different locations in space in order to create a “virtual symphony.” At the heart of the patented system is a so-called “sound processor.” A block diagram of a sound image location system incorporating two such sound processors is shown in Fig. 16 in the patents and is reproduced below: [[Image here]] A monaural signal is provided at terminal 1502. That signal is separated into two individual “channel signals.” Each channel signal is then processed through a separate “channel.” One channel includes sound processor 1501 while the other includes sound processor 1502. Each sound processor alters the phase and amplitude of the channel signal passed through the respective channel. The output of each sound processor can be coupled to a transducer (not shown in Fig. 16), otherwise known as a speaker, to reproduce the sound. By altering the phase and amplitude of the channel signals, the patented system allegedly can create the illusion that the sound originates from a point in 3-dimensional space away from the speakers, even though the sound actually originates from the speakers. As is well known in the art, a transfer function is the relationship between the out *1329 put of a component in a system to its input. The transfer function in the preferred embodiment of the patents was empirically derived through a laborious process, which is described in detail in the patents. The patents stress the importance of the change in amplitude and phase of one of the channel signals relative to the other. As a result of this relative relationship between the signals, one of the sound processors can be significantly reduced or even eliminated by performing all of the amplitude alteration and phase shifting on one of the channel signals. As described in the patents, the transfer functions, themselves, can be changed by position control parameters that are provided to the sound processors via terminal 1505. These position control parameters give a user the added flexibility of moving the apparent point of origin of a sound, giving a similar effect as the adjustment of “balance” or “fade” controls would in a conventional stereo. A schematic diagram of the preferred embodiment of the patented sound processor is shown in the patents at Fig. 18a, which is reproduced below: [[Image here]] The phase shifting and amplitude alteration is performed in filters 1610 and 1630, which are specifically designed to impart a unique and different change in each of a plurality of successive frequency bands that comprise the audio spectrum (i.e., 20-20,000 Hz). In one embodiment of the sound processor, the bands are 40 Hz in width, thus creating approximately 500 bands within the audio spectrum. It is critical both for the operation and patentability of the invention that the phase shift and amplitude change for each of the frequency bands be individually specified. The ability to separately alter the phase and amplitude of the channel signals for each frequency band is at the heart of the invention. To implement this approach, the patent describes a filter embodiment that uses a plurality of cascaded bandpass filters, with each bandpass filter designed to impart a pre-determined phase shift and amplitude change to a corresponding frequency band. The sound processor design shown in Fig. 18a includes six potentiometers (1651-1654, 1657,1658) to give the system added flexibility. In the preferred embodiment, four of the potentiometers (1651-1654) are controlled by a single two-axis joystick so that an operator can vary the apparent position of the sound in a smooth and continuous manner. The four potentiometers operate as follows. The monaural signal is simultaneously fed to the potentiometers 1651 and 1652. The two potentiometers 1651 and 1652 work differentially, so that an increase in one produces a corresponding decrease in the other and vice versa. The output of filter 1610 is fed to a differentially controlled potentiometer, 1653, which attenuates the output of the filter. *1330 The companion to potentiometer 1658 is potentiometer 1654, which is coupled to the output of filter 1630 in order to attenuate its output. The two potentiometers, 1653 and 1654, operate in the same complimentary manner as potentiometers 1651 and 1652. The output of potentiometer 1653 is passed through a buffer amplifier 1655 to another potentiometer 1657 and from there to reversing switch 1659, “which allows the filter signals to be fed directly or interchanged, to first inputs of summing elements 1660 and 1670.” ’462 patent, col. 13,11. 12-14. Assuming that switch 1659 is in the position shown in Fig. 18a, the output of potentiometer 1657 is fed to summing element 1670 while the output of potentiometer 1658, which acts in unison with potentiometer 1657, is fed to summing element 1660, where it is summed with the attenuated monaural signal. From the summing elements, the channel signals proceed to their respective output terminals 1688 and 1689. The embodiment shown in Fig. 18a permits the operator to individually select different signals by changing the settings of switches 1682, 1685, and 1691, the details of which are not relevant to the present dispute. B. The District Court Proceedings Before the district court, QSound asserted the four independent claims of the two patents: claims 1 and 7 of the ’462 patent; and claims 1 and 5 of the ’860 patent. The special master, in a thorough opinion, interpreted a number of disputed words and phrases in the asserted claims. 1 The district court adopted most of the special master’s findings and recommendations. While QSound challenges many of these conclusions, we need only consider a few of the disputed phrases because they are disposi-tive. Each of the independent claims is reproduced below with the disputed claim language emphasized. 1. The Claims a. The %62 Patent: 1. A method for producing and locating an apparent origin of a selected sound from an electrical signal corresponding to the selected sound in a predetermined and localized position anywhere within the three-dimensional space containing a listener, comprising the steps of: separating said electrical signal into respective first and second channel signals; altering the amplitude and shifting the phase of the signal in both said first and second channel signals while maintaining said phase and amplitude differential ther-ebetween for successive discrete frequency bands across the audio spectrum and each successive phase shift being different than the preceding phase shift, relative to zero degrees, thereby producing first channel and second channel' modified signals and creating a phase differential and an amplitude differential between the two channel signals; maintaining the first channel signal separate and apart from the second channel signal following the step of altering the amplitude and shifting the phase; and respectively applying said first and second channel modified signals that are maintained separate and apart and that have said phase and amplitude differential therebetween to first and second transducer means located within the three-dimensional space and spaced part [sic: apart] from the listener to produce a sound apparently • originating at a predetermined location in the three-dimensional space that may be different from the location of said sound transducer means. 7. A system for conditioning a signal for producing and locating, using two transducers located in free space, an auditory sensory illusion of an apparent origin for at least one selected sound at a predetermined localized position located within the three-dimensional space containing a listener from a single electrical signal corresponding to the selected sound, comprising: first and second channel means both *1331 receiving the same single electrical signal, said first and signal [sic: second] channel means including respective first and second sound processor means each for altering the amplitude and shifting the phase angle of the respective electrical signal on a frequency dependent basis for successive discrete frequency intervals across the audio spectrum to produce a respective modified signal wherein the amplitude alteration differential and the phase angle shift differential occurring between the two channels are respective predetermined values for each said successive frequency interval of the audio spectrum, said sound processor means shifting the phase angle such that each successive phase angle shift is different and independent of a preceding phase angle shift relative to zero degrees, and said first and second channels being maintained separate and apart prior to being fed to the two transducers. b. The ’860 Patent: 1. A method for producing and locating an apparent origin of a selected sound from an input monaural signal corresponding to the selected sound in a predetermined and localized position anywhere within the three-dimensional space containing a listener, comprising the steps of: separating said input monaural signal into respective first and second channel signals; providing a sound position control signal derived independently of the input monaural signal for making amplitude and phase adjustments at each of a number of discrete frequency bands over the audio spectrum and relating to a sound location determined by azimuth, height, and depth; altering the amplitude and shifting the phase of the signal in at least one of said first and second channels in response to said sound position control signal, both altering and shifting being done on a predetermined frequency dependent basis for successive discrete frequency bands across the audio spectrum and each successive phase shift being different than the preceding phase shift relative to zero degrees, thereby producing at least a first channel or a second channel modified signal and creating a continuous phase differential and a continuous amplitude differential between the two channel signals that varies for each of said discrete frequency bands; maintaining the first channel signal separate and apart from the second channel signal following the step of altering the amplitude and shifting the phase; and respectively applying at least first and second channel modified signals that are maintained separate and apart and that have said phase and amplitude differential therebetween to first and second transducer means located with [sic: within] the three-dimensional space and spaced apart from the listener to produce a sound apparently originating at a predetermined location in the three dimensional space that may be different from the location of said sound transducer means. 5. A system for conditioning a signal for producing and locating, using two transducers located in free space, an auditory sensory illusion of an apparent origin for at least one selected sound at a predetermined localized position located within the three-dimensional space containing a listener from a single monaural input signal corresponding to be selected sound, comprising: first and second channel means both receiving the same single monaural input signal, one of said first and signal channel means including sound processor means connected to receive a sound position control signal derived independently of the input monaural signal for controlling amplitude and phase adjustments at each of a number of selected frequency intervals of the audio spectrum and relating to a sound location determined by azimuth, height, and depth for altering the amplitude and shifting the phase angle of the respective signal on a frequency dependent basis for successive discrete frequency intervals across the audio spectrum to produce a respective modified signal there *1332 from in response to said sound'position control signal, wherein the amplitude alterations and the phase shifts form a continuous differential between the two channels that varies for each said successive frequency interval of the audio spectrum, said sound processor means shifting the phase angle such that each phase angle shift is different relative to zero degrees, and said first and second channels being maintained separate and apart prior to being fed to the two transducers. 2. The Claim Construction The first claim at issue of each of the patents is a method claim, while the second claim at issue of each of the patents is a claim to a “system.” After examining the prosecution history, the special master gave the phrase “being maintained separate and apart” its ordinary, literal meaning. He distinguished, however, between the term “following” that appears in claim 1 of the ’462 and the ’860 patents and the phrase “prior to” that appears in claim 7 of the ’462 patent and claim 5 of the ’860 patent. The special master interpreted “prior to” as simply requiring that the two signals be kept separate and apart “merely immediately before the loudspeakers [i.e., the ‘transducers’] where the signals are clearly separate and apart.” “Following,” however, was deemed more restrictive. The special master interpreted that term to require the two signals to be kept “separate and apart” from each other immediately after the two signals are initially altered. Because the two channel signals in the accused Spatializer device are combined after their phase and amplitudes are altered, but are separated immediately before they are fed to the speakers, the special master recommended that the Spatializer does not infringe the “following” claims but potentially could infringe the “prior to”- claims. As indicated above, the district court adopted the special master’s recommendations as to the “following” claims but not the “prior to” claims. The district court held that Desper was entitled to summary judgment of non-infringement on the “prior to” claims as well as the “following” claims. On appeal, QSound challenges the district court’s interpretation of both “following” and “prior to.” 2 DISCUSSION “The law governing summary judgment is well established.” C.R. Bard, Inc. v. Advanced Cardiovascular, Inc., 911 F.2d 670, 672, 15 USPQ2d 1540, 1542 (Fed.Cir.1990). “Summary judgment is as appropriate in a patent case as it is in any other case.” Id. Under the Federal Rules of Civil Procedure, a motion for summary judgment should properly be granted when “there is no genuine issue as to any material fact and ... the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). The determination of whether an accused product or process infringes a claim of a patent is universally understood to involve two steps. First, we construe the asserted claim to determine its meaning and scope. See Markman v. Westview Instruments, Inc., 52 F.3d 967, 976, 34 USPQ2d 1321, 1326 (Fed.Cir.1995) (in banc), affd, 517 U.S. 370, 116 S.Ct. 1384, 134 L.Ed.2d 577, 38 USPQ2d 1461 (1996). In the second step, we compare the accused product or process to the properly construed claim. See id. Whether a product or process infringes the properly construed claims of a patent, literally or under the doctrine of equivalents, is a question of fact. See Tanabe Seiyaku Co. v. United States Int’l Trade Comm’n, 109 F.3d 726, 731, 41 USPQ2d 1976, 1981 (Fed.Cir.1997). Often, as in this case, the composition of the allegedly infringing process or product is undisputed. In such a case, literal infringement collapses into claim construction — a matter of law — amenable to *1333 summary judgment. See Athletic Alternatives, Inc. v. Prince Mfg., Inc., 73 F.3d 1573, 1578, 37 USPQ2d 1365, 1370 (Fed.Cir.1996). To determine the meaning and scope of a claim, we first examine the claim language, the specification, and, if in evidence, the prosecution history. See Vitron ics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1581, 39 USPQ2d 1573, 1576 (Fed.Cir.1996). Extrinsic evidence, that is, evidence outside the record before the United States Patent and Trademark Office (“PTO”), such as expert testimony about how those skilled in the art would interpret certain language in the claim, may also be considered when appropriate as an inherent part of the process of claim construction and as an aid in arriving at the proper construction of the claim, but may not be used to vary or contradict the otherwise unambiguous meaning of the claim. See id. at 1582, 90 F.3d 1576, 39 USPQ2d at 1577. The proper interpretation of a claim is a legal conclusion, over which we exercise plenary review. See Cybor Corp. v. FAS Tech., Inc., 138 F.3d 1448, 1456, 46 USPQ2d 1169, 1174 (Fed.Cir.1998) (in bane) (“[W]e review claim construction de novo on appeal including any allegedly fact-based questions relating to claim construction.”). I. “Following” We start first with the claim language itself. See Thermalloy, Inc. v. Aavid Eng’g, Inc., 121 F.3d 691, 692, 43 USPQ2d 1846, 1848 (Fed.Cir.1997) (“[T]hroughout the interpretation process, the focus remains on the meaning of claim language.”) The first claims of both the ’462 patent and the ’860 patent require that the system “maintain[] the first channel signal separate and apart irom the second channel signal following the step of altering the amplitude and shifting the phase” (emphasis added). The plain meaning of “following” is “subsequent to, after in time” or “next after,” depending on the context. Webster’s Third New International Dictionary 883 (1986). The problem in this case is that this definition does not tell us how close in time to the “altering” step the signals must be maintained separate and apart. To answer that question we consult the specifications. Though not definitive, the specifications support an interpretation that requires the two signals to be maintained separate and apart beginning immediately after the amplitude is initially altered and the phase is initially shifted. The word “following” appears only twice in each of the written descriptions and is not used with reference to the channel signals. (The same is true of the phrase “prior to,” discussed infra.) The drawings, Fig. 16 and Fig. 18a (pictured supra), however, graphically demonstrate that once the two channel signals are separated they are thereafter kept separate. Thus, the drawings support the interpretation that, immediately after the amplitude and phase of the channels signals are initially altered by the filter, they thereafter must be maintained separate and apart. QSound points to an isolated passage in the written description to support its proposed claim construction that it is only after the alteration of the amplitude and the shifting of the phase by filters and summers are complete that the signals are required to be maintained separate and apart. The passage upon which QSound relies simply indicates that the sound processor “may include some or all of the following circuit elements: filters, delays, [inverters], summers, amplifiers, and phase shifters.” ’462 patent, col. 11, 11. 11-12. The passage sheds no light as to when the signals must be maintained separate and apart. In contrast, the prosecution history confirms our claim construction. The limitation that the channel signals be maintained separate and apart was not in the original claim language. The original claims of what is now the ’462 patent were rejected as obvious under § 103 in light of British patent No. 942,-459. Tha original application (Ser. No. 07/398,988) was then abandoned in favor of a File Wrapper Continuation (“FWC”) (Ser. No. 07/696,989). See 37 C.F.R. § 1.62 (1997). The limitation that the channel signals be *1334 maintained separate and apart was then added in a preliminary amendment dated June 12, 1991. In the remarks accompanying that amendment, the patent applicants’ attorney stated: In fact, according to the present invention, once the monaural signal is split into the two channel signals, those signals are forever after isolated one from another and are never combined or commingled. Various operations may take place in each channel provided by the present invention, yet the signals being operated upon are never recombined or mixed. This is in direct distinction to the teaching of the British patent.... (Emphasis added.) In his closing remarks, the attorney further stressed in urging reconsideration: [I]n view of the amendments to the claims hereby, as well as the above remarks, it is respectfully submitted that a method and apparatus for producing two sound signals derived from a single monaural input signal in which an amplitude and phase shift is provided with such signals being maintained separate and apart up until the actual production of the sounds at the speakers, as taught by the present invention and as recited in the amended claims, is neither shown nor suggested in any of the cited references, alone or in combination. (Emphasis added). Rather than wait for a formal, written response, the applicants conducted a personal examiner interview. In the resulting “Examiner Interview Summary Record,” the examiner indicated that the amended claims “appear to distinguish over art of record in prior case (FR 1,512,059 & B 942,459).” However, the examiner did come up with a new reference on which to base another prior art rejection. Despite this new reference, the examiner indicated that two dependent claims would be allowable if they were to be rewritten in independent form. In response, the applicants abandoned the original independent claims and took what the examiner allowed. The rewritten claims were subsequently allowed and issued as part of the ’462 patent. The prosecution history of the ’860 patents tells much the same story. The ’860 patent was originally filed as a Rule 60 divisional of the FWC (Ser. No. 07/696,989). See 37 C.F.R. § 1.60 (1997). In a preliminary amendment dated October 31, 1991, after the parent application had been rejected in view of the British patent, the applicants amended all the independent claims to include the limitation that the channel signals be maintained separate and apart. In an office action dated February 5, 1992, the examiner rejected the amended claims as obvious in light of another reference. The applicants’ attorney responded with further amendments to the claims, and accompanying remarks in which the attorney distinguished the claimed invention from the cited reference. This amendment was sufficient to overcome the rejection, and the amended claims were issued as part of the ’860 patent. Thus, in the file history of both patente, the limitation that the channel signals be maintained separate and apart was added to overcome the prior art rejection based on the British reference, and in both cases it succeeded. QSound does not argue on appeal that its attorney’s comments do not, on their face, limit the scope of the asserted claims. Rather, QSound argues that the prosecuting attorney’s remarks should, not be used to interpret the disputed claim language because the remarks, according to QSound, were both erroneous and extraneous — erroneous because they are contrary to the express teaching of the patent itself and extraneous because they did not become part of the claim. We disagree on both accounts. Fig. 18a (pictured supra), according to QSound, “proves ... that counsel’s statement was erroneous.” Brief for Appellant at 33. In particular, QSound argues that “[a]f-ter the signals pass through the points labeled 1657 and 1658, figure 18[a] shows that the signals cross, and are commingled, at the point labeled 1659.” Brief for Appellant at *1335 34. This post-hoc argument flies in the face of the entirety of the written description and the operation of a switch as well understood in the art. The symbol for 1659 shows a broken, diagonal line with a continuous, diagonal line passing through the opening therein. This notation is used, as is well known in the art, to indicate that the two conductors within the switch are not connected. The switch, according to the written description, “allows the filter signals to be fed directly or interchanged, to first inputs of summing elements 1660 and 1670.” ’462 patent, col. 13, lines 12-14. Thus, reversible switch 1659 does not, as QSound alleges, “commingle” the two channel signals, but, instead, re-routes the respective signals to either summing element 1660 or 1670 depending on the switch setting. Switch 1659 therefore gives the operator the ability to create a mirror image of the sound illusion by simply flipping the switch. This reading is consistent with the normal operation of a reversible switch. Counsel’s argument to the contrary is disingenuous at best and blatantly false at worst. Following oral argument, during which counsel’s characterization of the switch was challenged, QSound attempted to bolster its untenable position with a submission of testimony from its expert, ostensibly pursuant to Fed. R.App. P. 28<j). Rule 28(j), however, permits a party to bring supplemental authorities to the court’s attention, not supplemental argument. Accordingly, QSound’s submission was improper. Even were we to consider the proffered testimony, it does not prove QSound’s point. None of the cited testimony establishes that reversible switch 1659 commingles the two channel signals. QSound’s attempt to sacrifice the prosecuting attorney who wrote and prosecuted the application before the PTO, and who presumably was in the best position at the time to understand the true nature of the invention — for the sake of its case here — is unavailing. QSound also argues that the prosecuting attorney’s remarks were “extraneous” in that “they did not become part of the claim.” In essence, QSound argues our interpretation would import into the claim the attorney’s “erroneous” remarks. In support of this proposition QSound relies primarily on Intervet America, Inc. v. Kee-Vet Labs., Inc., 887 F.2d 1050, 12 USPQ2d 1474 (Fed.Cir.1989). Intervet provides no such support. In Inter-vet, the trial court read two limitations into the claims (“single administration” and “attenuation”) based on the remarks of the prosecuting attorney. The “single administration” limitation was, in fact, an express limitation in one of the independent claims but not the others. During prosecution, “the attorney made the unqualified and now admittedly untrue statement that ‘the claims are restricted to a single vaccination scheme.’ ” Id. at 1054, 12 USPQ2d at 1477. This court, not surprisingly, held that the independent claims in which the “single administration” limitation did not appear were not limited to a “single administration,” the attorney’s remarks notwithstanding. Id. This court also reversed the trial court’s construction that required the claims to include an “attenuation” limitation based on the attorney’s remarks, even though that word did not appear in any of the claims. Id. at 1055, 12 USPQ2d at 1478 (“We have set forth the asserted claims in full above and it is clear that they make no reference whatever to attenuation.”). The differences between Intervet and this case are apparent. First, the dispute in this case involves the proper interpretation of language that is in the claims, in particular, the words “following” and “prior to.” We are interpreting explicit claim language, not importing limitations into the claim. Second, the attorney’s remarks in this case were not “erroneous.” As we have discussed above, the attorney accurately described the claimed invention and the cited prior art. There is no “admittedly untrue statement” in this case as in Intervet. Thus, QSound’s reliance upon Intervet is misplaced. QSound further argues that the attorney’s remarks should not be used to interpret the claim language because the amendment that precipitated the remarks did not end the prosecution. In effect, QSound argues that the remarks were inconsequential because *1336 they did not result in allowance of the claims. Again, QSound relies on Internet. What QSound fails to acknowledge, however, is that the amendment and accompanying remarks were made for the purpose of overcoming the outstanding rejection based on the British patent. That the prosecution shifted to a different focus does not blunt the impact of those remarks made to overcome the prior rejection. The significance of the remarks in this case is no different than in a case in which the claims are allowed in response to an amendment. We conclude that the claim construction for claim 1 of the ’462 patent and of the ’860 patent adopted by the district court is correct. II. “Prior To” Claim 7 of the ’462 patent and claim 5 of the ’860 patent, the two other independent claims that are at issue, both claim a “sound processor means.” These elements are written in means-plus-function language, as 35 U.S.C. § 112, ¶ 6 (1994) permits, since they recite a function without corresponding structure, see id.; Fonar Corp. v. General Elec. Co., 107 F.3d 1543, 1551, 41 USPQ2d 1801, 1807 (Fed.Cir.1997). Interpreting the function of an element written in means-plus-function language is a question of law subject to complete and independent review on appeal. See Chiuminatta Concrete Concepts, Inc. v. Cardinal Indus., Inc., 145 F.3d 1303, 46 USPQ2d 1752, 1755 (Fed.Cir.1998). QSound argues that the district court misconstrued the functional language — “said first and second channels being maintained separate and apart prior to being fed to the two transducers” — appearing in claims 5 and 7. In particular, QSound argues that the court erred in its construction of “prior to.” The district court construed “prior to,” based on the written description and the prosecution history, consistent "with its interpretation of “following.” In light of all the intrinsic evidence, we conclude that the district court properly construed this phrase. The plain meaning of “prior” supports QSound’s argument that “prior to” should be construed more broadly than “following.” QSound offers a definition for “pri- or,” which Desper does not dispute: “preceding in time; earlier; previous; former,” quoting Webster’s NewWorld Dictionary 1131 (2d ed.1984). Accord Webster’s II New College Dictionary 879 (1995). All the plain language requires, under this definition, is that the two channel signals must be maintained separate and apart before they are fed into the transducers (i.e., the speakers). There is some appeal to this argument. Common words, unless the context suggests otherwise, should be interpreted according to their ordinary meaning. See York Prods., Inc. v. Central Tractor Farm & Family Ctr., 99 F.3d 1568, 1572, 40 USPQ2d 1619, 1622 (Fed.Cir.1996) (“Without an express intent to impart a novel meaning to claim terms, an inventor’s claim terms take on their ordinary meaning.”) The problem with QSound’s approach in this case is that the context of the specifications strongly suggests that “prior to” should not be given its ordinary meaning. The drawings' and corresponding discussion in the written description show that the two channel signals are initially separated and thereafter maintained separate and apart. The phase and amplitude of each channel signal is altered and then supplied to one of the speakers. The district court construed the “prior to” language as requiring the channel signals to be maintained separate and apart after the phase and amplitude are initially altered. Obviously, this construction is supported by the written description because this is precisely how the preferred embodiment of the invention works. Moreover, maintaining the channel signals separate and apart after the phase and amplitude are initially altered is still before the signals are fed into the transducers. The district court concluded that the prosecution history tips the balance in favor of construing “prior to” in the same manner as “following” in claim 1. We agree. Prosecution history is an important source *1337 of intrinsic evidence in interpreting claims because it is a contemporaneous exchange between the applicant and the examiner. This is true whether the claim element in question is written pursuant to § 112, ¶ 6 or not. See Alpex Computer Corp. v. Nintendo Co., 102 F.3d 1214, 1220, 40 USPQ2d 1667, 1671 (Fed.Cir.1996) (prosecution history is relevant to interpreting an element written in means-plus-function language); United States v. Telectronics, Inc., 857 F.2d 778, 782, 8 USPQ2d 1217, 1220 (Fed.Cir.1988). During the prosecution of claim 7 of the ’462 patent, QSound treated claim 1 and claim 7 (then claim 8) together even though one used “following” language and the other used “prior to” language. In successfully overcoming the prior art rejection based on the British patent, QSound argued that “claims 1 and 8 have been amended hereby to make it more clear that the first channel signal is always maintained separate and apart from the second channel signal following the step of altering the amplitude and shifting the phase in at least one of those signals.” (Emphasis added). Thus, even though only claim 1 included the “following” language, QSound itself considered the scope of claims 1 and 8 commensurate in this regard. Because the public has the right to rely on the applicants’ remarks in seeking allowance of their claims, see Alpex, 102 F.3d at 1221, 40 USPQ2d at 1673 (“Just as prosecution history estoppel may act to estop an equivalence argument under the doctrine of equivalents, positions taken before the PTO may bar an inconsistent position on claim construction under § 112, ¶ 6.”), we agree with the district court that “prior to” should be construed to mean after the altering of the amplitude and shifting of the phase has begun, see Tandon Corp. v. United States Int’l Trade Comm’n, 831 F.2d 1017, 1023, 4 USPQ2d 1283, 1288 (Fed.Cir.1987) (“[T]wo claims which read differently can cover the same subject matter.”). 3 III. Non-Infringement The second step in an infringement analysis is determining whether the properly construed claim reads on the accused device. See Markman, 52 F.3d at 976, 34 USPQ2d at 1326. In order to literally infringe a method claim, the accused device must literally meet each and every one of the claim limitations. To find infringement under § 112, ¶ 6, the accused device must, inter alia, perform the identical function to that claimed in the means-plus-function element. See Cybor, 138 F.3d at 1457, 46 USPQ2d at 1175. A. The structure of the Spatializer is not in dispute. The Spatializer, 4 as in the claimed invention, splits the monaural signal into two channel signals (“L¡„” and “Rin”). A third signal (“C”) is created by summing these two signals together. The phase and amplitude of this third signal are then altered by a bandpass filter. The output of the filter is passed through an amplifier. The output of the amplifier is then added to one of the two channel signals (i.e., LjJ by a summer to form one of the output signals (“Lout”), which is then fed to a speaker. The third signal C is subtracted from the other channel signal (i.e., Rfa) to form the other output signal (“Rout”)- In view of the undisputed structure and operation of the accused device, we agree with the district court that the two channel signals in the Spatializer are not “main- *1338 tainted] ... separate and apart., following the step of altering the amplitude and shifting the phase,” as we have properly interpreted that phrase. The two channel signals are mixed by the summer, thereby altering the amplitude after the phase has been shifted by the bandpass filter. Thus, the two channel signals are not maintained separate and apart after the step of altering the amplitude and shifting the phase has begun. Accordingly, the Spatializer cannot and does not literally infringe either claim 1 of the ’462 patent or claim 1 of the ’860 patent. 5 Since we have construed the phrase “prior to” in claim 7 of the ’462 patent and claim 5 of the ’860 patent to be commensurate with the “following” limitation, these claims also are not literally infringed because the Spatializer does not perform the identical function. Accordingly, we affirm the district court’s summary judgment that the accused “Spatializer” device does not literally infringe. B. An accused device that does not literally infringe a claim may .nonetheless infringe under the doctrine of equivalents if differences between the accused device and the claimed invention are “insubstantial.” See Fonar, 107 F.3d at 1555, 41 USPQ2d at 1810. However, “[p]rosecution history estop-pel precludes a patentee from obtaining in an infringement suit patent protection for subject matter which it relinquished during prosecution in order to obtain allowance of the claims.” Mark I Mktg. Corp. v. R.R. Donnelley & Sons Co., 66 F.3d 285, 291, 36 USPQ2d 1095, 1099-1100 (Fed.Cir.1995); Southwall Techs., Inc. v. Cardinal IG Co., 54 F.3d 1570, 1580, 34 USPQ2d 1673, 1680 (Fed.Cir.1995). Amendments made to overcome a prior art rejection can create such an estoppel. See Litton Sys., Inc. v. Honeywell, Inc., 140 F.3d 1449, 1462, 46 USPQ2d 1321, 1330 (Fed.Cir.1998). The scope of the estoppel in such eases includes “features that the applicant amended his claim to avoid” or “trivial variations of such prior art features.” Id. (citing Southwall, 54 F.3d at 1580, 34 USPQ2d at 1680). Moreover, prosecution history estoppel cannot be avoided by filing a continuing application with narrowed claims rather than responding directly to an outstanding rejection. See Mark I, 66 F.3d at 292, 36 USPQ2d at 1100. Unequivocal assertions or arguments made during prosecution may also create an estoppel. See Haynes Int’l, Inc. v. Jessop Steel Co., 8 F.3d 1573, 1579, 28 USPQ2d 1652, 1657 (Fed.Cir.1993) (“Thus, an estoppel can be created even when the claim, which is the basis for the assertion of infringement under the doctrine of equivalents, was not amended during prosecution.”); Texas Instruments, Inc. v. United States Int’l Trade Comm’n, 988 F.2d 1165, 1174, 26 USPQ2d 1018, 1025 (Fed.Cir.1993). The application and scope of prosecution history estoppel is a legal question, which is reviewed without deference to the trial court. See LaBounty Mfg., Inc. v. United States Int’l Trade Comm’n, 867 F.2d 1572, 1576, 9 USPQ2d 1995, 1998 (Fed.Cir.1989). In determining the scope of what, if any, subject matter has been surrendered, the standard is an objective one: what would a reasonable competitor reading the prosecution history conclude has been surrendered. See Mark I, 66 F.3d at 291, 36 USPQ2d at 1100. In this case,- the amendments made to overcome the prior art rejection based on the British patent coupled with the accompanying unequivocal assertions made by the applicants before the Patent and Trademark Office to obtain allowance of the claims create an estoppel that precludes infringement under the doctrine of equivalents as a matter of law. The British patent, entitled “Improvements Relating to Apparatus for Deriving Pseudo-Stereophonic Signals,” describes an electrical circuit used to produce two *1339 pseudo-stereophonic signals from a single monaural signal. This invention takes the monaural signal and splits it into two “channel signals.” The amplitudes of these two channel signals are then altered by manipulating the two signals to form a signal equal to their difference. This difference signal is then added to an attenuated version of the monaural signal to form one of the pseudo-stereophonic signals (e.g., a right signal) and subtracted from the attenuated monaural signal to form the other pseudo-stereophonic signal (e.g., a left signal). These two pseudo-stereophonic signals are then fed to respective speakers to reproduce the sound. As explained above, the claims of the original patent application (Ser. No. 07/398,988) were rejected under 35 U.S.C. § 103 (1994) as being obvious in view of, inter alia, the British patent. That action was made final. Rather than appeal this final decision, the applicants chose instead to file a File Wrapper Continuation under 37 C.F.R. § 1.62. In addition, the applicants filed a Preliminary Amendment which added the “maintaining ... following” language to claim 1, and “maintained ... prior to” language to then claim 8. In the accompanying remarks, after describing the teaching of the British patent, applicants’ attorney distinguished the claimed invention (including claims 1 and then claim 8) from that reference by the following: In fact, according to the present invention, once the monaural signal is split into the two channel signals, those signals are forever after isolated one from another and are never combined or commingled. Various operations may take place in each channel provided by the present invention, yet the signals being operated upon are never recombined or mixed. This is in direct distinction to the teaching of the British patent.... In.order to emphasize the above distinction, claims 1 and 8 have been amended hereby to make it more clear that the first channel signal is always maintained separate and apart from the second channel signal following the step of altering the amplitude and shifting the phase in at least one of those channels. It is clear from this record that the applicants surrendered subject matter that .includes the specific technique employed in the accused Spatializer device. The British patent teaches the use of a conditioning signal derived from the monaural input signal. That conditioning signal is then added to the monaural input signal to form- one of the output signals and subtracted from the input signal to form the other output signal. Thus, in the British patent, the conditioning signal is not kept separate and apart from the input signal after altering of the amplitude and shifting of the phase has begun. The applicants explicitly surrendered such a topology when they distinguished the claimed invention from the teaching of the British patent. The Spatializer topology falls within the subject matter that applicants explicitly disclaimed. In the Spatializer, the monaural input signal is split into two input signals. A conditioning signal is generated by processing the input signal. This conditioning signal is ádded back to one of the two input signals to form one output signal and subtracted from the other input signal to form the other output signal. Thus, as in the British patent, the conditioning signal in the Spatializer is not maintained separate and apart from the input signals after altering of the amplitude and shifting of the phase has begun. Whatever differences there are between the sound processor in the British patent and the Spa-tializer with regard to this limitation are “trivial.” See Litton, 140 F.3d at 1462, 46 USPQ2d at 1330. Because the Spatializer falls within the subject matter surrendered by the applicants during the prosecution of the patents, infringement under the doctrine of equivalents is precluded as a matter of law. 6 QSound’s response is to argue that estop-pel should not apply since there are other *1340 bases upon which the claimed invention and the British reference could have been distinguished besides the way the signals are maintained separate and apart. QSound cites Conroy v. Reebok Int'l, Ltd., 14 F.3d 1570, 1576, 29 USPQ2d 1373, 1378 (Fed.Cir.1994) in support. That case has no application here. In Conroy, the district court granted summary judgment of non-infringement, purportedly using the “hypothetical claim” analysis under Wilson Sporting Goods Co. v. David Geoffrey & Assoc., 904 F.2d 677, 14 USPQ2d 1942 (Fed.Cir.1990), simply upon finding one of the elements of the accused device in the prior art. See Conroy, 14 F.3d at 1577, 29 USPQ2d at 1378. We held that that was improper because it failed to consider whether the prior art would render the “hypothetical claim” invalid. See id. In the case at bar, the district court did not merely find the accused structure in the prior art. Instead, it held that the applicants affirmatively distinguished their claimed invention from that in the prior art. This is a classic case of prosecution history estoppel. The fact that the British reference could have been distinguished, standing alone, on different grounds is immaterial. See Southwall, 54 F.3d at 1583, 34 USPQ2d at 1682 (“[A]ny argument made regarding the need to distinguish the prior art ... does create a separate estoppel, regardless of other distinctions made.” (citation omitted)). Accordingly, summary judgment of non-infringement under the doctrine of equivalents was proper. The public has a right to rely on the assertions made by a patent applicant to secure allowance of its claims. Post-hoc, litigation-inspired argument cannot be used to reclaim subject matter that the public record in the PTO clearly shows has been abandoned. CONCLUSION Accordingly, the summary judgment of non-infringement is AFFIRMED. Footnotes [1] Specifically, the special master interpreted "locating,” “localized,” "predetermined,” "maintaining," "channel,” "successive,” "discrete frequency bands,” "separate and apart,” "altering the amplitude ... for successive discrete frequency bands,” and "independent.” [2] QSound also challenges the interpretation of . “discrete” and "independent,” as it must to ultimately prevail. Given ohr interpretation of "maintained separate and apart,” "following," and "prior to,” however, we need not reach these issues. [3] QSound invokes the rule that different words should be interpreted differently, yet it concedes that in other parts of the claims different words should be interpreted the same. For example, the special master found and, QSound concedes, that the phrase "frequency bands” in Claim 1 of the '860 patent and the phrase "frequency intervals” in claim 5 mean the same thing. QSound's argument based on constructional rules therefore loses much of its force. [4] Although there is more than one version of the Spatializer, the general description given applies equally to all versions. A protective order prevents us from giving more details on the accused devices. [5] QSound does not argue the remaining dependent claims separately, and because those claims necessarily must be narrower than the independent claims, they too cannot be infringed. [6] This estoppel applies with equal force to the identical language appearing in the ’860 patent, which was a divisional of the common parent application. See Mark I, 66 F.3d at 291, 36 *1340 USPQ2d at 1100 (applying prosecution history estoppel to claims in a patent from related application).