`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`SONOS, INC.
`Petitioner
`
`v.
`
`IMPLICIT, LLC
`Patent Owner
`
`Case: To be assigned
`Patent No. 8,942,252
`
`
`
`
`
`
`
`
`
`PRELIMINARY RESPONSE TO PETITION FOR INTER PARTES
`REVIEW OF U.S. PATENT NO. 8,942,252
`
`
`
`
`
`
`
`
`
`
`TABLE OF CONTENTS
`
`INTRODUCTION ............................................................................................... 1
`I.
`II. THE ’252 PATENT .......................................................................................... 3
`III. THE PRIOR ART ............................................................................................. 6
`IV. CLAIM CONSTRUCTION ...........................................................................11
`V. THE APPLICABLE LAW .............................................................................11
`VI. PETITIONER’S GROUNDS .........................................................................13
`A. Janevski coupled with the knowledge of a person of ordinary skill does not
`yield the claimed inventions .................................................................................14
`B. Janevski coupled with the alleged prior art does not yield the claimed
`inventions ..............................................................................................................16
`1. The prior art does not teach “smoothing a rendering time differential” .....16
`2. Petitioner fails to demonstrate a reason to combine any prior art
`“smoothing” with Janevski, or explain how such a combination would work .17
`3. Baumgartner and Janevski are incompatible ...............................................19
`4. Combining “smoothing” with Janevski would not result in an operable
`method, much less the claimed methods ...........................................................20
`VII. CONCLUSION ...........................................................................................23
`
`
`
`
`
`i
`
`
`
`TABLE OF AUTHORITIES
`
`Cases
`
`Beckson Marine, Inc. v. NFM, Inc.,
`292 F.3d 718 (Fed. Cir. 2002) ..............................................................................13
`Dynamic Drinkware, LLC v. National Graphics, Inc.,
`800 F.3d 1375 (Fed. Cir. 2015) ............................................................................11
`Graham v. John Deere Co.,
`383 U.S. 1 (1966) ..................................................................................................12
`Harmonic Inc. v. Avid Tech., Inc.,
`815 F.3d 1356 (Fed. Cir. 2016) ............................................................................11
`In re Wilson,
`424 F.2d 1382 (C.C.P.A. 1970) ............................................................................12
`In re: Hodges,
`882 F.3d 1107, 1111 (Fed. Cir. 2018) ..................................................................22
`In re: Stepan Company,
`868 F.3d 1342 (Fed. Cir. 2017) ............................................................................22
`KSR Int’l Co. v. Teleflex Inc.,
`550 U.S. 398 (2007) ..............................................................................................12
`Personal Web Technologies, Inc. v. Apple Inc.,
`848 F.3d 987 (Fed. Cir. 2017) ....................................................................... 13, 16
`Statutes
`
`35 U.S.C. § 103 ......................................................................................... 1, 2, 12, 13
`35 U.S.C. § 103(a) ...................................................................................................12
`35 U.S.C. § 314 ........................................................................................................11
`
`
`
`
`
`ii
`
`
`
`I.
`
`INTRODUCTION
`
`Patent Owner Implicit, LLC (“Implicit”) opposes institution of Inter Partes
`
`Review on all grounds because Petitioner Sonos, Inc. (“Petitioner”) cannot show a
`
`reasonable likelihood of demonstrating invalidity of any challenged claim.
`
`Petitioner alleges invalidity under 35 U.S.C. § 103, relying primarily on a
`
`single prior art reference, U.S. Patent No. 7,269,338 (“Janevski”) (Ex. 1007).
`
`Petitioner fails to demonstrate, however, that Janevski, either alone or in
`
`combination with other alleged prior art cited in the Petition, discloses every
`
`element of any challenged claim.
`
`The patent at issue, U.S. Patent No. 8,942,252 (“the ’252 Patent”) (Ex.
`
`1001), discloses and claims methods of synchronizing the rendering of a single
`
`content stream on multiple devices. In order to do that, the ’252 Patent teaches the
`
`use of “master and slave” devices that track and use two separate and distinct
`
`elements: 1) “device” time, and 2) “rendering” time. The ’252 Patent further
`
`teaches “smoothing” a “rendering time differential that exists between the master
`
`rendering device” and “slave device.” All of the claims at issue in this Petition
`
`require smoothing of rendering times.
`
`Petitioner does not identify anywhere in the prior art the smoothing of a
`
`rendering time differential. Petitioner concedes Janevski does not teach smoothing
`
`of a rendering time differential. Pet., at 42. Petitioner argues instead that it would
`
`
`
`1
`
`
`
`have been obvious to modify Janevski, based on a statement by its expert that
`
`smoothing functions were known in the prior art.
`
`Petitioner’s expert testimony does not satisfy the requirements to prove
`
`obviousness under 35 U.S.C. § 103. Petitioner’s expert argues that smoothing
`
`functions were known in the prior art, citing technical articles and textbooks dating
`
`to as early as 1971. But Petitioner does not rely on these sources as prior art, much
`
`less demonstrate a reason, suggestion, or motivation to adapt prior art smoothing
`
`functions and combine them with prior art synchronization methods. In the end,
`
`Petitioner and its expert are left with nothing more than conclusory testimony that
`
`Janevski “could be modified” to achieve the claimed invention.
`
`In a series of alternative arguments, Petitioner argues that Janevski could
`
`have been combined with various prior art references that may teach the use of a
`
`smoothing function. But again, Petitioner does not identify the claimed
`
`requirement of “smoothing a rendering time differential” in any of those
`
`references. Petitioner merely demonstrates, at most, that smoothing algorithms
`
`have been used in the prior art for various other purposes. Petitioner’s claimed
`
`combination thus does not meet all of the requirements of the claims for which the
`
`Petition seeks review.
`
`
`
`2
`
`
`
`II. THE ’252 PATENT
`
`The ’252 Patent is directed to a method and system for synchronizing the
`
`rendering of content on multiple networked devices. It allows, for example,
`
`simultaneous playback of audio and/or video across multiple devices. Over the
`
`past several years, the patented technology has achieved remarkable commercial
`
`success. Because the patented technology allows a single content stream to be
`
`rendered with extraordinarily tight synchronization at multiple networked devices,
`
`it has been widely adopted in the wireless, multi-room audio space. Examples
`
`include Sonos’s popular speaker systems, marketed and sold by the Petitioner.
`
`The problem of synchronizing the rendering of content is an old one.
`
`Multimedia presentations, for example, confronted the problems of synchronizing
`
`video, audio, and text. Ex. 1001, at col. 1:23-28). In the case where the rendering
`
`device is a single device, operating on a single clock (such as a television,
`
`computer, or smartphone), various solutions are possible.
`
`The problem becomes more complicated, however, where multiple devices
`
`are used to render content simultaneously. This is because each rendering device
`
`has its own system clock, which means that each rendering device typically
`
`operates in its own “time domain.” Ex. 1001, at col. 1:40-44. The clocks on each
`
`of the system’s devices typically operate on slightly different frequencies, such that
`
`the “time domains” of each rendering device can, over time, drift out of alignment.
`
`
`
`3
`
`
`
`A viewer would perceive content to be out of synchronization. Ex. 1001, at col.
`
`1:44-46.
`
`The ’252 Patent’s solution
`
`to
`
`the problem of rendering content
`
`simultaneously on multiple devices (the “rendering devices”) includes methods and
`
`systems for exchanging information or “messages” between the various devices.
`
`The messages may, depending on the embodiment, include “device times,”
`
`“rendering times,” and/or other information that indicates one or both of these
`
`elements. Ex. 1001, at col. 2:15-17. A device time is a time indicated by that
`
`device’s system clock. Ex. 1001, at col. 2:20-21. A “rendering time” identifies a
`
`specific position in the content stream. It can be the time represented by the
`
`amount of a specific item of content that a rendering device has rendered. Ex.
`
`1001, at col. 2:18-19. Or, it can be an idealized “rendering time” that identifies a
`
`content position that “may not correspond to the actual rendering time of any of the
`
`rendering devices.” Ex. 1001, at col. 5:65-6:1.
`
`In order to exchange information to synchronize rendering of content, one
`
`device is identified as the “master” device, while the remaining devices are
`
`identified as “slave” devices. Ex. 1001, at col. 2:32-36. Messages are exchanged
`
`between the devices to clearly identify the master device and the slave devices. Id.
`
`In order to synchronize the devices, in one embodiment, the master device
`
`sends a message to each slave device, including: 1) the master’s rendering time,
`
`
`
`4
`
`
`
`and 2) the master’s device time. Ex. 1001, at col. 2:34-36. Each slave device then
`
`determines whether it is synchronized with the master’s rendering time. If not, the
`
`slave adjusts its rendering time to adjust for the difference between the master
`
`rendering time and its own rendering time. Ex. 1001, at col. 2:39-42.
`
`To make an adjustment, in one embodiment, the slave device exchanges
`
`device time and rendering time information for the master and slave devices, along
`
`with time stamps on the messages exchanging the information, allowing the slave
`
`device to determine by how much its rendering time is out of alignment with the
`
`master device’s rendering time. The slave device can, for example, compare the
`
`slave rendering time to the master rendering time at a given slave device time. Or,
`
`the slave device can compare the slave device time at a given rendering time to the
`
`master device time at the same rendering time. Ex. 1001, at col. 2:42-48.
`
`The ’252 Patent teaches, in addition to calculating latency using rendering
`
`times and device times, using information from multiple time points to further
`
`assist synchronization. A component can use the stored information to “smooth”
`
`the time domain differentials using various techniques. Ex. 1001, at col. 7:16-17.
`
`Time domain differentials can be smoothed, for example, by averaging the last
`
`several differentials. Ex. 1001, at col. 7:17-19. In one embodiment, the last eight
`
`pairs of time domain differentials are stored, and the system calculates the average
`
`of the minimum value of the set of eight larger differentials and the maximum
`
`
`
`5
`
`
`
`value of eight smaller differentials. Ex. 1001, at col. 7:21-26, Fig. 7. A decaying
`
`function can be used to limit the impact of the oldest time differentials. Ex. 1001,
`
`at col. 7:18-21. The smoothed time domain differentials are then used to
`
`synchronize the master and slave rendering times.
`
`All claims of the ’252 Patent require smoothing a time differential and, in
`
`particular, the rendering time differential.
`
`III. THE PRIOR ART
`
`Petitioner primarily relies on Janevski. Patent Owner does not dispute, for
`
`purposes of this Preliminary Response only, that Janevski is prior art to the ’252
`
`Patent.1
`
`Janevski is directed to the problem of simultaneous playback of digital
`
`content at remote locations that have each recorded and stored their own copy of
`
`the content to be played back. Ex. 1007, at col. 5:3-10. The digital content
`
`streams are pre-recorded on and played back from Personal Video Recorders
`
`(“PVRs”). Janevski purports to allow multiple viewers to simultaneously view
`
`content, that they have each previously recorded, in their respective homes
`
`
`1 Patent Owner does not believe that Janevski is prior art, and reserves its right to
`
`challenge the prior art status of Janevski should review be instituted or in any other
`
`proceeding.
`
`
`
`6
`
`
`
`simultaneously with one another, and to rewind or fast-forward content at remote
`
`locations while maintaining simultaneous viewing for all viewers. Ex. 1007, at
`
`col. 5:16-19; col. 15:22-26.
`
`The primary manner in which Janevski addresses the problem of
`
`simultaneous playback is content-based synchronization. “To achieve precise
`
`synchronization, the present invention compares corresponding content or
`
`‘landmarks’ of pairs of video playbacks to be synchronized, determines video
`
`replay ‘distance’ between the landmark pairs, and slows downs or speeds up
`
`selected playbacks in accordance with these distances.” Ex. 1007, at col. 3:52-57.
`
`Janevski allows remote devices to exchange information about rendering time to
`
`facilitate its content matching. All calculations are based solely on rendering
`
`times.
`
`Two or more users may participate in a viewing session. The first user to
`
`play given content is the “initiator” device, while the remaining devices are
`
`“participant devices.” Ex. 1007, at col. 6:16-21. An initiator device can be a
`
`participant device, and a participant device can be an initiator device. Ex. 1007, at
`
`col. 6:22-23. If, for example, a user of a participant device performs an operation
`
`on the streamed content, such as fast-forward, the participant device becomes the
`
`“initiator” device for that operation. Ex. 1007, at col. 6:23-25. Accordingly,
`
`
`
`7
`
`
`
`“initiator/participant” is not a persistent designated status, but rather a status that is
`
`conditional on a given operation.
`
`To synchronize playback, an initiator device sends a message to participant
`
`devices. The message includes a time stamp from the “video timer,” which is the
`
`rendering time of the video or digital stream. Ex. 1007, at col. 7:41-47, col. 9:15-
`
`29. The participant notes the time the message was received on its own video
`
`timer, and sends a reply including the time of the reply message according to that
`
`video timer. Ex. 1007, at col. 9:21-17, col. 10:10-17. The initiator notes the time
`
`of receipt as measured by its own video timer. Ex. 1007, at col. 10:17-18. Using
`
`this set of rendering times, the initiator device can calculate time delays in the
`
`rendering of content between initiator and participant devices. Ex. 1007, at col.
`
`12:18-29.
`
`The initiator device, having calculated the time delay between the rendering
`
`times of the initiator and participant devices, also selects a “query frame,” which is
`
`a “frame that the initiator device has just played or has recently played, so that the
`
`content of the query frame and its respective time stamp represent where the
`
`playback is in the content at a particular time which is current.” Ex. 1007, at col.
`
`12:5-11.
`
`The initiator then sends messages to the participant devices including the
`
`time misregistration, query frame time stamp, and other information. Ex. 1007, at
`
`
`
`8
`
`
`
`col. 12:30-36. Upon receiving the message including time misregistration, the
`
`participant PVR may advance or rollback the playing of video to compensate for
`
`the time misregistration. Ex. 1007, at col. 12:50-56.
`
`To accomplish synchronization, the participant device uses the information
`
`sent by the initiator device to calculate “frame misregistration,” which is the
`
`difference in the number of frames between the initiator’s query frame and the
`
`participant device’s frame at or near the rendering time of the initiator device’s
`
`selected query frame. Ex. 1007, at col. 13:24-28. To synchronize the content, the
`
`participating device, based on the frame misregistration, searches for an
`
`appropriate frame to fast-forward or rewind toward. See generally Ex. 1007, at
`
`col. 13:29-15:2.
`
`Janevski does not teach, anywhere, smoothing a rendering time differential.
`
`Nor does any of the other prior art cited in the Petition. The Azevedo
`
`publication (Ex. 1010) is directed to convergence algorithms used in clock
`
`synchronization. Ex. 1010, at 1. Azevedo teaches smoothing of “correction
`
`terms” used in one algorithm. Ex. 1010, at 4. Azevedo does not teach, and
`
`Petitioner notably does not allege that Azevedo teaches, smoothing a rendering
`
`time domain differential.
`
`Like Azevedo, none of the other alleged prior art references (Mills, Ex.
`
`1011, challenge #2; Berthaud, Ex. 1012, challenge #3; and Eidson, Ex. 1013,
`
`
`
`9
`
`
`
`challenge #4) teaches smoothing of a rendering time domain differential. Rather,
`
`like Azevedo, they teach other uses of smoothing functions.
`
`Mills (Ex. 1011) teaches smoothing clock offset values to “select the peer
`
`clock(s) used for synchronization [of computer clocks] and to maximize the
`
`accuracy and stability of the indications.” Ex. 1011, at 36.
`
`Berthaud (Ex. 1012) also addresses computer clock synchronization,
`
`specifically, correcting errors in the measurement of the offsets of computer
`
`clocks. Ex. 1012, at Abstract. Synchronization of the rendering of content is not
`
`addressed.
`
`Eidson, U.S. Patent No. 6,278,710 (Ex. 1013), is yet another disclosure
`
`related to synchronization of computer clocks. Ex. 1013. It teaches averaging the
`
`differences in time measurements for messages sent and received in computer
`
`networks, but does not address the synchronization of the rendering of content.
`
`While Baumgartner, U.S. Patent No. 5,642,171, does address
`
`the
`
`synchronization of audio and video content, it still cannot support even a prima
`
`facie case of invalidity. Ex. 1014. Baumgartner compares video frame and audio
`
`frame “positions” to determine a “synchronization error value,” which is the
`
`number of frames by which a video frame is ahead or behind an audio frame. Ex.
`
`1014, at col. 6:50-55. The synchronization error value is used to assign a “tempo
`
`value” which is used to adjust the video “tempo” (frame speed). Ex. 1014, at col.
`
`
`
`10
`
`
`
`6:56-58. A smoothing is then applied to the tempo value, which is a weighted
`
`average of prior tempo values. Tempo values (frame speeds) are not, however,
`
`time differentials.
`
`IV. CLAIM CONSTRUCTION
`
`While Patent Owner does not agree with the Petition’s proposed claim
`
`constructions, for the purposes of this Preliminary Response only, Patent Owner
`
`does not contest those claim constructions.2
`
`V. THE APPLICABLE LAW
`
`The PTAB may institute Inter Partes Review only if the information
`
`presented in the Petition “shows that there is a reasonable likelihood that the
`
`petitioner would prevail with respect to at least 1 of the claims challenged in the
`
`petition.” 35 U.S.C. § 314.
`
`In an Inter Partes Review, “the petitioner has the burden from the onset to
`
`show with particularity why the patent it challenges is unpatentable.” Harmonic
`
`Inc. v. Avid Tech., Inc., 815 F.3d 1356, 1363 (Fed. Cir. 2016). This burden “never
`
`shifts” to the Patent Owner. Dynamic Drinkware, LLC v. National Graphics, Inc.,
`
`800 F.3d 1375, 1378 (Fed. Cir. 2015).
`
`2 Patent Owner reserves the right to contest any of Petitioner’s proposed claim
`
`constructions and propose its own constructions in this proceeding, should review
`
`be instituted, or in any other proceeding.
`
`
`
`11
`
`
`
`The only alleged statutory basis for challenging patentability in an Inter
`
`Partes Review petition in this matter is obviousness under 35 U.S.C. § 103. An
`
`obviousness analysis requires comparison of the patent’s claims to the prior art.
`
`“All words in a claim must be considered in judging the patentability of that claim
`
`against the prior art.” In re Wilson, 424 F.2d 1382, 1385 (C.C.P.A. 1970).
`
`A claims is unpatentable under 35 U.S.C. § 103(a) only 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.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 405 (2007). The question
`
`of obviousness is resolved on the basis of underlying factual determinations,
`
`including (1) the scope and content of the prior art; (2) any differences between the
`
`claimed subject matter and the prior art; (3) the level of ordinary skill in the art;
`
`and (4) objective evidence of non-obviousness. Graham v. John Deere Co., 383
`
`U.S. 1, 17-18 (1966). A claim is not proved obvious merely by demonstrating all
`
`of the claim limitations are present in the prior art; rather, obviousness requires
`
`Petitioner to demonstrate that a person of ordinary skill in the art would have some
`
`reason to combine the references. KSR, 550 U.S. at 401-402.
`
`As a threshold matter, obviousness cannot be found unless Petitioner
`
`demonstrates that each and every claim limitation is taught in one or more prior art
`
`
`
`12
`
`
`
`references. See Personal Web Technologies, Inc. v. Apple Inc., 848 F.3d 987, 993
`
`(Fed. Cir. 2017) (ob
`
`viousness analysis first requires determination that prior art teaches all
`
`elements of the claim); Beckson Marine, Inc. v. NFM, Inc., 292 F.3d 718, 727
`
`(Fed. Cir. 2002) (for claimed subject matter to be obvious, prior art must expressly
`
`teach each claim limitation exactly, or else the record must disclose a reason,
`
`suggestion, or motivation for a person of ordinary skill in the art modify the prior
`
`art teachings to obtain the claimed invention).
`
`VI. PETITIONER’S GROUNDS
`
`Petitioner asserts five challenges, all based on 35 U.S.C. § 103. All five
`
`challenges allege obviousness based on combinations including Janevski and
`
`either: 1) the knowledge of a person of ordinary skill in the art, or 2) other alleged
`
`prior art.
`
`The challenged claims are the same for each challenge and include claims 1-
`
`3, 8, 11, and 17. Claims 1 and 11 are independent claims. All claims are method
`
`claims.
`
`Each challenged claim requires “smoothing” a rendering time differential.
`
`Claim 1 requires, for example, a “first slave device” which is “configured to
`
`smooth a rendering time differential that exists between the master rendering
`
`device and the first slave device.” Claim 11 requires a slave device that determines
`
`
`
`13
`
`
`
`“a smoothed rendering time differential that exists between the master rendering
`
`device and the slave device.”
`
`Each of Petitioner’s challenges relies on Janevski as its principal reference.
`
`In each case, Petitioner concedes that Janevski does not teach “smoothing” of
`
`rendering time differentials, as required by each of the challenged claims. To
`
`supply the missing element, Petitioner combines Janevski with 1) the knowledge of
`
`a person of ordinary skill in the art (challenge #1), or 2) various alleged prior art
`
`references which purport to apply smoothing functions in various synchronization
`
`contexts (all challenges).
`
`A.
`
`Janevski coupled with the knowledge of a person of ordinary skill
`does not yield the claimed inventions
`
`In challenge #1, Petition argues that a person of ordinary skill in the art
`
`would have known that Janevski could be modified to include smoothing.
`
`Petitioner contends that a person of ordinary skill would have been motivated to
`
`improve Janevski because the goal of Janevski was improve the accuracy of
`
`synchronization, and smoothing any of the “periodic” calculations of Janevski
`
`would have accomplished that goal. Pet., at 45-46.
`
`Petitioner’s argument is illogical on its face. If the objective of Janevski was
`
`to minimize synchronization error, and if a person or ordinary skill in the art would
`
`have known that smoothing any of the periodic calculations would reduce that
`
`
`
`14
`
`
`
`error, then Janevski should have disclosed, or at least pointed to, use of techniques
`
`such as smoothing in its synchronization method. But it did not.
`
`To support the argument that a person of ordinary skill in the art would have
`
`known to use a smoothing function, Petition cites several references, dating as
`
`early as 1971, which purport to discuss smoothing functions. Pet., at 43; Exs.
`
`1015-18. But Petitioner does not contend or show that any of these references
`
`actually discloses smoothing a rendering time differential. Nor does Petition
`
`contend that there would have been any reason, suggestion, or motivation to
`
`combine these references with Janevski, much less that the combination
`
`necessarily would have yielded the claimed invention.
`
`Petitioner’s argument that a person of ordinary skill in the art would have
`
`known to modify Janevski based solely on a person of ordinary skill’s knowledge
`
`is, therefore, an improper attempt to avoid its burden of demonstrating a reason,
`
`suggestion, or motivation to combine prior art references to achieve the claimed
`
`invention.
`
`Given Petitioner’s implicit concession that none of the alleged prior art
`
`teaches smoothing a rendering time differential, the general “knowledge of a
`
`person of ordinary skill in the art” is not a substitute for a missing claim limitation.
`
`Unless all of the limitations of a claim are taught somewhere in the prior art,
`
`
`
`15
`
`
`
`obviousness cannot be shown. Personal Web, 848 F.3d at 993; Beckson Marine,
`
`292 F.3d at 727.
`
`B.
`
`Janevski coupled with the alleged prior art does not yield the
`claimed inventions
`
`Petitioner’s alternative obviousness theories rely on combinations of
`
`Janevski with various prior art references that Petitioner contends teach smoothing
`
`some variable in the context of synchronization.
`
`1.
`
`The prior art does not teach “smoothing a rendering time
`differential”
`Nowhere does Petitioner contend that any of these prior art references
`
`discloses smoothing a rendering time differential, as required by the claims. The
`
`fact that smoothing may have been applied in other synchronization methods does
`
`not imply that it would have been obvious to combine these references with
`
`Janevski or, more importantly, to combine them with Janevski so as to achieve the
`
`claimed requirement of smoothing a rendering time differential.
`
`Of the five references relied on in the Petition, four of them—Azevedo,
`
`Mills, Berthaud, and Eidson—do not discuss synchronization of the rendering of
`
`content at all, much less rendering time differentials. Rather, each of these
`
`references broadly relates to synchronization of clocks in computers and computer
`
`networks, and the various functions that are used to determine the error, or offset,
`
`in the various clocks used in them. See Section III, supra.
`
`
`
`16
`
`
`
`To the extent smoothing functions are used in these references, they are to
`
`smooth variables other than rendering time differentials. Neither Petitioner its
`
`expert contends that any of the alleged reference teaches smoothing a rendering
`
`time differential.
`
`At most, Azevedo (challenge #1) teaches smoothing of correction terms. As
`
`noted above, Mills (challenge #2) teaches smoothing clock offset values to “select
`
`the peer clock(s) used for synchronization [of computer clocks] and to maximize
`
`the accuracy and stability of the indications.” Ex. 1011, at 36. Berthaud
`
`(challenge #3) also teaches (at most) smoothing of measured computer network
`
`clock offsets. Ex. 1012, at 265-66. U.S. Patent 6,278,710 “Eidson” (challenge
`
`#4) also addresses smoothing measurements of the differences between sent and
`
`received messages for purposes of synchronizing computer network clocks. Ex.
`
`1013, at col. 1:42-col.2:6.
`
`2.
`
`Petitioner fails to demonstrate a reason to combine any
`prior art “smoothing” with Janevski, or explain how such a
`combination would work
`In addition to failing to show how any given reference actually supplies the
`
`claim limitation that Petitioner concedes is missing from Janevski, Petitioner does
`
`not attempt to address how combining any of the above prior art references
`
`actually would result in the inventions of the challenged claims. There is no
`
`
`
`17
`
`
`
`explanation as to how (or why) a person of ordinary skill would adapt any of these
`
`references so that they would teach smoothing a rendering time.
`
`This
`
`is particularly problematic because
`
`Janevski addresses
`
`the
`
`synchronization of rendering pre-recorded content, not the synchronization of
`
`clocks of participating devices. Petitioner does not explain how any of the various
`
`references for teaching clock synchronization would relate or be applied to pre-
`
`recorded content synchronization. Petitioner offers the conclusory statement that
`
`“Janevski’s synchronization techniques are not ‘limited to any particular time
`
`synchronization’” (Pet., at 59, quoting Ex. 1007), but that is not a reason,
`
`suggestion, or motivation to combine as alleged by Petitioner. At best, it is an
`
`open-ended statement that additional synchronizations methods beyond those
`
`taught by Janevski could be developed.
`
`Similarly, Petitioner argues that Janevski teaches that any available “time
`
`synchronization” mechanism could be used to determine the “time count”
`
`difference. Pet., at 63-64. But how Janevski measures time differences differs
`
`markedly from how the ’252 Patent uses time domain differentials to accomplish
`
`the synchronization of the rendering of content.
`
`Without addressing how a person of ordinary skill in the art actually would
`
`combine Janevski with “smoothing” prior art, Petitioner argues that a person of
`
`ordinary skill in the art nonetheless would have combined smoothing with Janevski
`
`
`
`18
`
`
`
`because smoothing had known benefits and because increased clock accuracy was
`
`desired. E.g. Pet., at 43-45. These generalized “reasons” to combine do not
`
`address why or how Janevski would be combined with such art. Janevski
`
`synchronizes content only after calculating “frame misregistration” to search for an
`
`appropriate frame to fast-forward or rewind toward. It does not teach actually
`
`synchronizing any clocks of any devices, so it remains unclear how “smoothing”
`
`would apply to a reference like Janevski, or why anyone with ordinary knowledge
`
`of Janevski would be motivated to add a “smoothing” function.
`
`3.
`
`Baumgartner and Janevski are incompatible
`
`Unlike the other alleged prior art, the final reference relied on by Petitioner,
`
`Baumgartner, does
`
`relate
`
`to content synchronization,
`
`in
`
`this case
`
`the
`
`synchronization of audio and video content. However, Baumgartner teaches
`
`synchronization using frame position differences, not time or time domain
`
`differences. Baumgartner calculates the number of frames by which the video
`
`presentation is ahead or behind the audio presentation, and uses that difference in
`
`frame position to determine a “tempo value,” or frame speed, to synchronize the
`
`content. These “tempo values” may be averaged or smoothed.
`
`
`
`Petitioner does not contend that a “tempo value” is a “time domain
`
`differential,” much less a “rendering time domain differential.” Petitioner argues
`
`only that Baumgartner “amounts to a ‘smoothing’ a ‘frame’ differential between
`
`
`
`19
`
`
`
`the rendering of two media streams.” Pet., at 67. But the claim language requires
`
`calculation and use of a “rendering time differential.” If anything, Baumgartner
`
`teaches away from smoothing a rendering time differential by teaching that frames,
`
`rather than time differentials, are required to synchronize content.
`
`4.
`
`Combining “smoothing” with Janevski would not result in
`an operable method, much less the claimed methods
`There is another flaw in all of Petitioner’s obviousness arguments, separate
`
`
`
`from the fact that none of the prior art discloses the smoothing a rendering time
`
`differential element. What the Petition identifies as the rendering time differentials
`
`are not in fact time differentials, and it is unclear how these data points could be
`
`“smoothed.” Accordingly, even assuming that the prior art teaches the smoothing
`
`that is cl