`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`APPLE INC.,
`Petitioner,
`
`v.
`
`COREPHOTONICS, LTD.,
`Patent Owner.
`____________
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`____________
`
`
`PATENT OWNER’S SUR-REPLY TO PETITION FOR
`INTER PARTES REVIEW OF U.S. PATENT NO. 9,857,568
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`I.
`II.
`
`TABLE OF CONTENTS
`
`The specification’s teaching of how to calculate TTL/EFL
`
`Apple fails to rebut Corephotonics’ evidence on the term “image
`
`The Only Construction of “TTL” Consistent with the Specification Is
`
`Introduction .......................................................................................... 1
`One That Measures Distance to the Sensor .......................................... 1
`The specification says how the TTL/EFL ratio is measured ........... 1
`A.
`B.
`controls, even under the BRI standard ............................................ 3
`C.
`plane” .............................................................................................. 6
`D. Apple quotes Dr. Moore’s testimony misleadingly and out of
`context ............................................................................................. 7
`III. Ogino Does Not Render the TTL/EFL Ratio Limitation Obvious ....... 9
`A. Ogino does not disclose a TTL/EFL ratio less than 1.0 .................. 9
`B.
`at lines 5:65–6:2 with its configuration 6 ...................................... 13
`C.
`were combined .............................................................................. 15
`D. Dr. Sasián’s new Zemax analysis is untimely, and it also undercuts
`Apple’s arguments ........................................................................ 16
`E.
`TTL/EFL ratio limitation .............................................................. 20
`IV. One Skilled in the Art Would Not Have Been Motivated to Build
`Ogino Example 6 with an L11/L1e Ratio Satisfying the Claims ....... 22
`
`Apple fails to establish a motivation to combine Ogino’s teaching
`
`Ogino does not teach what the TTL would be if Ogino’s teachings
`
`Dr. Moore did not testify that Ogino example 6 discloses the
`
`i
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`A.
`B.
`
`V.
`
`
`
`The Handbook of Optics teaches away from the combination
`
`Apple proposes .............................................................................. 22
`Beich teaches away from the combination Apple proposes .......... 23
`Conclusion .......................................................................................... 24
`
`ii
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`Cases
`
`TABLE OF AUTHORITIES
`
`In re Gordon,
`733 F.2d 900 (Fed. Cir. 1984) ................................................................. 14
`
`In re Schwemberger,
`410 Fed. Appx. 298 (Fed. Cir. 2010) ........................................................ 3
`
`Kinetic Concepts, Inc. v. Smith & Nephew, Inc.,
`688 F.3d 1342 (Fed. Cir. 2012) ............................................................... 14
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) ................................................................. 4
`
`SightSound Techs., LLC v. Apple Inc.,
`809 F.3d 1307 (Fed. Cir. 2015) ................................................................. 4
`
`Vitronics Corp. v. Conceptronic, Inc.,
`90 F.3d 1576 (Fed.Cir.1996) ..................................................................... 4
`
`Rules
`
`
`
`37 C.F.R. § 42.100 .......................................................................................... 3
`
`
`
`iii
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`PATENT OWNER’S EXHIBIT LIST
`
`Number
`
`Description
`
`2001 U.S. Patent Publication No. ijkk/jilmnlo “Tang”
`
`2002 U.S. Patent Publication No. ijkk/jipmmkj “Tang”
`
`2003 U.S. Patent Publication No. ijkk/jioklpj “Chen”
`
`2004 Declaration of Marc A. Fenster in Support of Motion to Appear
`Pro Hac Vice on Behalf of Patent Owner Corephotonics, Ltd.
`
`2005 Declaration of Duncan Moore, Ph.D.
`
`2006 Curriculum Vitae of Duncan Moore, Ph.D.
`
`2007 U.S. Patent No. q,nmr,qrk
`
`2008 U.S. Patent Publication No. ijkk/jilmnlo
`
`2009 U.S. Patent Publication No. ijkk/jipmmkj
`
`2010 U.S. Patent Publication No. ijkk/jioklpj
`
`2011 Transcript of the February kr, ijkm Deposition of
`Dr. José Sasián in the IPRijkq-jkklj and IPRijkq-jkklo
`
`2012 Transcript of the July i, ijkm Deposition of
`Dr. José Sasián in IPRijkm-jjjnj
`
`2013 Declaration of José Sasián in IPRijkq-jkklj and IPRijkq-jkklo
`
`2014 Excerpts from Robert E. Fischer, Biljana Tadic-Galeb,
`Paul R. Yoder, OPTICAL SYSTEM DESIGN, ind Ed. (ijjq)
`
`iv
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`Number
`
`Description
`
`2015 U.S. Patent m,opq,nkj
`
`2016 U.S. Patent q,ilq,pkr
`
`2017 Reserved
`
`2018 U.S. Patent Publication ijkk/jioklpk
`
`2019 Humpston et al., “Optical performance of bare image sensor die
`and sensors packaged at the wafer level and protected by a cover
`glass,” Proceedings of SPIE (ijjq)
`
`2020 Excerpts from Symmons & Schaub, FIELD GUIDE TO MOLDED
`OPTICS, SPIE (ijko)
`
`2021 William S. Beich, “Injection Molded Polymer Optics in the ikst-
`Century,” SPIE Proceedings, Vol. rqor (ijjr)
`
`2022 Luxin Nie, “Patent Review of Miniature Camera Lenses and A
`Brief Comparison of Two Relative Design Patterns” (ijkp)
`
`2023 Excerpts from José Sasián, INTRODUCTION TO ABERRATIONS IN
`OPTICAL IMAGING SYSTEMS (ijkn)
`
`2024 Kingslake & Johnson, LENS DESIGN FUNDAMENTALS, id ed.,
`Ch. l, (ijkj)
`
`2025 ZEMAX Optical Design Program, User’s Manual, July q, ijkk.
`
`2026 Transcript of the September r, ijkm Deposition of
`Dr. José Sasián in IPRijkm-jjjnj
`
`v
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`I.
`
`Introduction
`
`Ogino fails to disclose both the TTL/EFL ratio and Lkk/Lke ratio limita-
`
`tions of the challenged claims. Apple fails to demonstrate that one skilled in
`
`the art would have been motivated to combine the disparate teachings of
`
`Ogino—or to combine to those teachings with Beich—to obtain the inven-
`
`tions claimed in the ’roq patent. Moreover, Beich and the Handbook of Optics
`
`contain multiple rules and guidelines that teach away from the combinations
`
`proposed by Apple. For these reasons, and for the reasons set forth in Core-
`
`photonics’ earlier response, Apple’s grounds for unpatentability should be re-
`
`jected, and the patentability of the challenged claims should be affirmed.
`
`II. The Only Construction of “TTL” Consistent with the
`Specification Is One That Measures Distance to the Sensor
`
`A. The specification says how the TTL/EFL ratio is measured
`
`In its Reply, Apple states that “the image plane and the sensor are taught
`
`by the ’roq patent to be coincident.” (Paper ii at kn.) Apple’s expert says the
`
`same. (Ex. kjio, ¶ r.) Likewise, both Apple and its expert say that “‘image
`
`plane’ in light of the ’roq patent is just the plane ‘for image formation of an
`
`object.’” (Paper ii at kl; Ex. kjio, ¶ q.) On these two points, Apple and its
`
`expert are correct. Indeed, as these terms are used in the specification of
`
`the ’roq patent, the image plane is the surface of the sensor.
`
`x
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`(cid:128)e specification describes “an image sensor with an image plane on
`
`which an image of the object is formed.” (Ex. kjjk at k:oo–op (emphasis
`
`added).) (cid:128)e first embodiment is expressly described as containing a sensor at
`
`the location of the image plane. (Id. at n:lj–li.) (cid:128)e other two embodiments
`
`within the detailed description are variations on the first embodiment. (Id. at
`
`r:oo–o:l, p:in–iq.) (cid:128)ere is no suggestion in the ’roq patent that any of the
`
`embodiments lacks a sensor or that the TTL as part of the TTL/EFL ratio
`
`should be measured to any location other than that of the sensor.
`
`Rather, the ’roq patent specification is clear that TTL is measured to the
`
`location of the sensor. (cid:128)e specification says that “the total track length on an
`
`optical axis between the object-side surface of the first lens element and the
`
`electronic sensor is marked ‘TTL.’” (Ex. kjjk at i:i–l.) (cid:128)en in the very next
`
`sentence, the specification states that “[i]n all embodiments, TTL is smaller
`
`than the EFL, i.e. the TTL/EFL ratio is smaller than k.j.” (Id. at l–o.) (cid:128)is
`
`second sentence is where the “ratio TTL/EFL of less than k.j” condition in the
`
`challenged claims is first introduced in the specification.
`
`(cid:128)ese two sentences at the start of the second column—the first explain-
`
`ing how TTL is measured and the second introducing the requirement that the
`
`TTL/EFL ratio be less than k.j—cannot be divorced from one another. One
`
`(cid:127)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`skilled in the art reading the specification would understand the second sen-
`
`tence to be introducing a requirement for the TTL/EFL ratio, where that ratio
`
`is calculated using the definitions of EFL and TTL introduced in the first sen-
`
`tence. Nothing in the specification suggests any other way of computing the
`
`TTL/EFL ratio, for the purposes of comparing it to k.j.
`
`B.
`
`The specification’s teaching of how to calculate TTL/EFL
`controls, even under the BRI standard
`
`Even under the broadest reasonable interpretation (BRI) standard for
`
`claim construction, the construction must be consistent with the specification.
`
`np C.F.R. § li.kjj(b) (ijko) (claim in an IPR given the “broadest reasonable
`
`construction in light of the specification of the patent in which it appears”
`
`(emphasis added)). Under the BRI standard, when a patentee acts as his own
`
`lexicographer and defines a term in the specification, that definition controls
`
`and not the ordinary meaning. In re Schwemberger, lkj Fed. Appx. imq, njn
`
`(Fed. Cir. ijkj) (construing claims under BRI standard).
`
`In interpreting the TTL/EFL ratio limitation of the ’roq patent claims,
`
`that ratio must be computed in a way that is consistent with how it is computed
`
`in the ’roq specification. (cid:128)at specification explicitly states that TTL is meas-
`
`ured to the sensor. To take a prior art reference like Ogino that contains a
`
`sensor (Ex. kjjr at r:lj-lk (“an imaging device kjj (refer to Fig. k), such as
`
`(cid:131)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`a CCD, which outputs imaging signals”), Fig. o (showing the location of im-
`
`aging device kjj in example o)), and to calculate TTL/EFL by measuring TTL
`
`to someplace other than the location of this sensor, is inconsistent with the
`
`specification and not a proper application of the BRI standard.
`
`Apple argues that the statement in the ’roq specification of how TTL is
`
`measured can be disregarded because it is not an “express definition.” (Paper
`
`ii at q.) In Phillips, the Federal Circuit recognized that the specification can
`
`define terms by implication, even without an “express definition.” Phillips v.
`
`AWH Corp., lkr F.nd knjn, knik (Fed. Cir. ijjr) (en banc) (the specification
`
`“acts as a dictionary when it expressly defines terms used in the claims or
`
`when it defines terms by implication” (emphasis added), quoting Vitronics
`
`Corp. v. Conceptronic, Inc., mj F.nd krpo, krqi (Fed.Cir.kmmo)). “Express def-
`
`inition” is not required when applying the BRI standard either. SightSound
`
`Techs., LLC v. Apple Inc., qjm F.nd knjp, knkp (Fed. Cir. ijkr) (stating in an
`
`appeal of a PTAB decision under BRI that “a claim term may be clearly rede-
`
`fined without an explicit statement of redefinition,” quoting Phillips, lkr F.nd
`
`at knkr).
`
`Apple’s argument that the specification’s instruction on how to calculate
`
`the TTL/EFL ratio should be disregarded in favor of the “plain and ordinary”
`
`(cid:132)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`meaning of TTL (Paper ii at m) fails for another reason, namely that Apple
`
`has not established that its preferred construction is the “plain and ordinary
`
`meaning.” (cid:128)e only evidence of how TTL is used outside of the ’roq patent
`
`cited in Apple’s petition was the Chen patent. (Paper i at kj–kk.) Apple’s reply
`
`refers to the plain and ordinary meaning but cites to no evidence that such a
`
`plain and ordinary meaning exists outside of the ’roq patent, let alone that this
`
`meaning is Apple’s proposed construction. (Paper ii at m.) As Corephotonics
`
`demonstrated in its response, the term TTL is used in multiple inconsistent
`
`ways in the art, including in prior art cited by the patent and by Apple in re-
`
`lated IPRs. (Paper ik at ik–ii.) (cid:128)e record suggests there is no accepted “plain
`
`and ordinary meaning” of TTL that should be applied in place of the definition
`
`utilized in the ’roq patent specification. (cid:128)e construction most consistent with
`
`the intrinsic evidence should be applied. See, e.g., Ex. ijjp, at i:ij-in (the
`
`Tang patent, cited by the ’roq patent) (“a distance on the optical axis between
`
`the object-side surface of the first lens element and the electronic sensor is
`
`TTL”).
`
`(cid:128)e limitation “ratio of TTL/EFL of less than k.j” is not one that can be
`
`construed to permit many different kinds of TTL. Determining whether an
`
`(cid:134)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`embodiment falls within the claims requires calculating TTL/EFL and com-
`
`paring the result to a threshold value. Where the art recognizes multiple ways
`
`of calculating a quantity such as TTL and those multiple ways can reach dif-
`
`ferent results, the correct way of calculating that quantity in the context of the
`
`patent claims is the way taught in the patent specification. Accordingly, the
`
`correct way to calculate TTL/EFL to determine whether Ogino satisfies
`
`the ’roq patent claims is to measure to Ogino’s sensor, not to some “image
`
`plane” located somewhere other than the sensor.
`
`C. Apple fails to rebut Corephotonics’ evidence on the term
`“image plane”
`
`As Corephotonics demonstrated in its response, the term “image plane”
`
`has multiple meanings, corresponding to different locations in real-world sys-
`
`tems. (Paper ik at ii–im.) Apple’s answer in its reply is that it is possible in
`
`some systems for the different image planes to be in the same location. (Paper
`
`ii at kj; Ex. kjio, ¶ n.) But as Corephotonics demonstrated and Apple’s ex-
`
`pert acknowledged during deposition, it is also possible for the different image
`
`planes to be in different locations. (Ex. ijio at in:ki–il:m.) A construction
`
`that happens to agree with the correct construction in some circumstances but
`
`disagrees in others cannot be the correct construction. Computing TTL by
`
`measuring to an image plane that does not coincide with Ogino’s sensor, as
`
`(cid:136)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`Apple proposes, is not applying a proper construction of the ’roq patent
`
`claims.
`
`Apple’s reply cites plots of MTF versus focus shift in Figures kB and iB
`
`of the ’roq as support for its construction. (Paper ii at kn–kl.) (cid:128)ese plots do
`
`show that moving the image sensor will change the contrast of the image, in
`
`some cases reducing the contrast. (Ex. ijio at km:kn–ik:i.) But that confirms
`
`Corephotonics’ interpretation of the claims. By including these plots, the ’roq
`
`specification expressly contemplates that there is more than one possible lo-
`
`cation of the image sensor for a given configuration of lens elements. To cal-
`
`culate the ratio of TTL/EFL, it is necessary to know where the sensor is, and
`
`not perform a calculation that ignores the location of the sensor and is solely
`
`based upon the configuration of the lens elements.
`
`D. Apple quotes Dr. Moore’s testimony misleadingly and out of
`context
`
`(cid:128)e testimony that Apple cites from Dr. Moore does not suggest a differ-
`
`ent conclusion. Dr. Moore’s testimony included the following exchange:
`
`Q. So if I were to ask you what the broadest construction
`was that is reasonable, it would be — it would include, at least,
`the measurement of the length along the optical axis from the
`object side surface of the first element to the image plane?
`MR. RUBIN: Objection, form.
`
`(cid:137)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`A. To where the sensor is, yes.
`Q. But if we’re just looking for the broadest construction of
`TTL that is reasonable —
`MR. RUBIN: Objection, form.
`Q. — would Dr. Sasian's construction of TTL be included
`in that broadest reasonable construction?
`MR. RUBIN: Objection, form.
`A. I think so.
`
`(Ex. kjir at om:q–ik (emphasis added).)
`
`Apple cites only the second question and answer from this exchange. But
`
`in the first question, Dr. Moore was asked whether the “broadest construction
`
`[of TTL] that is reasonable” would include measurement “to the image plane.”
`
`Dr. Moore agreed, but only with the clarification that the image plane was
`
`“where the sensor is.” In the second question, Dr. Moore was asked again
`
`about the “broadest construction of TTL that is reasonable,” with no reference
`
`to the correct legal standard (“in light of the specification”), and whether Dr.
`
`Sasian’s construction (the construction Dr. Moore was asked about in the prior
`
`question) would be included in that broadest construction. He answered, “I
`
`think so.” But that is entirely consistent with his answer to the first question:
`
`a measurement to the image plane is within the broadest reasonable construc-
`
`tion of TTL, so long as that image plane is located where the sensor is.
`
`(cid:140)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`Dr. Moore’s testimony concerning film cameras is also entirely con-
`
`sistent with his testimony above. He testified that a person of ordinary skill
`
`would understand that TTL in a film camera is measured to the film, i.e. to the
`
`sensor of the film camera. (Ex. kjir at oq:m-ir.) Further, in an answer that
`
`Apple replaced with ellipses in its reply (Paper ii at p), Dr. Moore testified
`
`that TTL would not be measured to the paraxial image plane of the camera, if
`
`that image plane did not coincide with the location of the film. (Ex. kjir at
`
`oq:kq–ik.)
`
`III. Ogino Does Not Render the TTL/EFL Ratio Limitation Obvious
`
`A. Ogino does not disclose a TTL/EFL ratio less than 1.0
`
`Apple’s reply argues that Ogino Table kk “represents” an embodiment
`
`with (k) no cover glass and (i) a TTL/EFL ratio of less than k.j. (Paper ii at
`
`kr.) Neither assertion is true. Table kk (and Table ki) of Ogino corresponds to
`
`the same “configuration o” embodiment as Figure o. (Ex. kjjr at kr:km–ii
`
`(“Tables n to ki show specific lens data as Examples i to o, corresponding to
`
`the configuration of the imaging lenses shown in FIGS. i to o.”).) Figure o
`
`shows a cover glass “CG,” with thickness Dki and surfaces with radii of cur-
`
`vature of Rki and Rkn:
`
`(cid:142)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`
`
`(Ex. kjjr, Fig. o.)
`
`Table kk describes a configuration that contains this same cover glass CG,
`
`with Dki equal to j.njj and Rki and Rkn equal to infinity (corresponding to
`
`planar surfaces) (Ex. ijkk at qi:kq–ii):
`
`
`
`x(cid:143)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`(Ex. kjjr, Table kk.) Apple’s expert confirmed at deposition that “lines ki and
`
`kn of Table kk are describing the same object as is labeled ‘CG’ in Figure o.”
`
`(Ex. ijkk at qi:in–qn:k.) Table kk does not “represent” a configuration with
`
`the cover glass removed, any more than it “represents” configurations with
`
`one or more of the lenses removed or with any number of other changes made.
`
`When Apple’s expert entered the parameters from Table kk into the Ze-
`
`max software and calculated the TTL (called “TOTR” in Zemax), he found
`
`that the TTL is (cid:132).(cid:132)(cid:140)(cid:142), greater than the EFL of (cid:132).(cid:132)(cid:127)(cid:140):
`
`
`
`xx
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`(Ex. kjjn, § IX.)
`
`
`
`Q. I see. So the value towards the bottom right of the
`screen labeled “TOTR” and equal to 4.489 is the TTL that Ze-
`max calculated; is that right?
`A. Yes.
`
`(Ex. ijkk at kjn:kq-ii.)
`
`Apple claims that the “TL” of (cid:132).(cid:131)(cid:140)(cid:137) mm listed in Table kk is a total track
`
`length. As noted above, Apple’s own expert demonstrated this is false, calcu-
`
`lating the TTL of Table kk to be (cid:132).(cid:132)(cid:140)(cid:142) mm using Zemax. And Ogino itself
`
`confirms that its TL value is not the TTL of a lens configuration that lacks the
`
`cover glass. Ogino states that TL is a “total lens length” and that “in the total
`
`lens length TL, the back focal length portion uses an air-converted value.” (Ex.
`
`kjjr at kl:lp–rn, kr:kq–ii.) (cid:128)is definition only makes sense for a lens con-
`
`figuration that has something other than air (i.e., the cover glass) within the
`
`back focal length portion. Otherwise, there would be no need to “air-convert[].”
`
`x(cid:127)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`In other words, the TL value listed in Table kk is a property of an embodiment
`
`that contains the cover glass, not one where the cover glass has been removed.
`
`B. Apple fails to establish a motivation to combine Ogino’s
`teaching at lines 5:65–6:2 with its configuration 6
`
`Because Ogino does not disclose an embodiment with a TTL/EFL ratio
`
`less than one, Apple’s petition proposes modifying Ogino’s configuration o
`
`based upon the statement at lines (cid:134):(cid:136)(cid:134)–(cid:136):(cid:127) in Ogino that a coating can be ap-
`
`plied to Lr “without using the optical member CG” and that the total length
`
`can be reduced. (Paper i at io–ip.) But Apple fails to establish that one skilled
`
`in the art would have been motivated to combine Ogino configuration o with
`
`Ogino (cid:134):(cid:136)(cid:134)–(cid:136):(cid:127) in this way.
`
`Ogino configuration o is the product of a different technique for reducing
`
`total length:
`
`As in the sixth embodiment shown in FIG. 6, the fifth lens L5
`may be configured to have a meniscus shape which is concave
`toward the image side, and the configurations of the first to fifth
`lenses Ll to L5 may be common to the configurations of the
`fourth embodiment except that the fifth lens L5 has a meniscus
`shape which is concave toward the image side. By making the
`fifth lens L5 have a meniscus shape which is concave toward
`the image side, it is possible to appropriately reduce the total
`length.
`
`x(cid:131)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`(Ex. kjjr at kn:ol–kl:r (emphasis added).) Apple provides no evidence that
`
`one skilled in the art reading Ogino would think that configuration o had any-
`
`thing other than an “appropriately” reduced total length or that it was in need
`
`of further reduction by the technique of Ogino (cid:134):(cid:136)(cid:134)–(cid:136):(cid:127). See Kinetic Concepts,
`
`Inc. v. Smith & Nephew, Inc., oqq F.nd knli, knom (Fed. Cir. ijki) (no moti-
`
`vation to combine two prior art references that both provide the same feature).
`
`(cid:128)e need for an articulated motivation to combine Ogino (cid:134):(cid:136)(cid:134)–(cid:136):(cid:127) with
`
`Ogino configuration o is made greater by the significant disadvantages to re-
`
`moving the cover glass. As Corephotonics showed in its response, the cover
`
`glass serves to reduce particle contamination, which is responsible for more
`
`than (cid:142)(cid:143)% of defects in camera modules. (Paper ik at lk; Ex. ijjr, ¶¶ kji–jl;
`
`Ex. ijkm, at n–l.) Apple does not dispute this evidence, and it does not provide
`
`evidence that removing the cover glass from the configuration o design would
`
`provide benefits sufficient to overcome its disadvantages in manufacturability
`
`and defect rate, and thereby provide a motivation to combine. See In re Gor-
`
`don, pnn F.id mjj, mji (Fed. Cir. kmql) (modification not obvious where it
`
`renders the apparatus inoperative for its intended purpose).
`
`x(cid:132)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`C. Ogino does not teach what the TTL would be if Ogino’s
`teachings were combined
`
`According to Apple’s reply, a POSITA reading the ’roq patent would
`
`understand that the image plane location is specified by a “lens prescription
`
`table,” and that TTL “can be determined by summing the widths of lens ele-
`
`ments and spacing between lens elements” from this table. (Paper ii at kk–
`
`kn.) But the only lens prescription table in Ogino for configuration o is that in
`
`Table kk, which specifies an image plane that is (cid:132).(cid:132)(cid:140)(cid:142) mm away from the ob-
`
`ject-side surface of the first lens element and a TTL greater than EFL, as ex-
`
`plained above. Ogino contains no lens prescription table for any configuration
`
`without a cover glass.
`
`Simply removing the cover glass from configuration o does not alter the
`
`TTL, unless other changes to the lens prescription are made, such as by mov-
`
`ing the sensor, or changing the shape or location of the first lens. Ogino does
`
`suggest “applying a coating to the fifth lens Lr or the like without using the
`
`optical member CG” as one of several possible ways to “reduce the total
`
`length.” (Ex. kjjr at r:oo–o:i, o:rr–rm, p:nr–lq, p:ro–rq, q:q–kl, m:r–m,
`
`m:nj–nr, m:oj–on, kj:nk–nr, kj:rn–ro, kk:kk–ko, kk:li–lm, kn:no–nm, kl:n–r.)
`
`But Ogino does not specify what other changes, if any, would be made at the
`
`same time as the cover glass were removed. In particular, it does not teach
`
`x(cid:134)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`that the sensor should be moved to the distance provided for “TL” in one of
`
`Ogino’s tables, or to any other specific distance that would provide a TTL less
`
`than EFL.
`
`D. Dr. Sasián’s new Zemax analysis is untimely, and it also
`undercuts Apple’s arguments
`
`To obtain a lens assembly with TTL/EFL ratio less than k.j requires mod-
`
`ifying one the embodiments of Ogino, and Apple’s petition fails to demon-
`
`strate why or how a person skilled in the art would make the specific modifi-
`
`cation that it proposes. (cid:128)e only lens design that Apple’s expert analyzed in
`
`any detail in his original declaration was configuration o, with the cover glass
`
`and with TTL greater than EFL. (Ex. kjjn, § IX; Ex. ijio at nr:i–r.) In his
`
`reply declaration, Apple’s expert provided for the first time an analysis in Ze-
`
`max of a configuration without a cover glass and with a TTL that is purport-
`
`edly less than the EFL. (Ex. kjio, § VI.)
`
`(cid:128)is analysis from Dr. Sasián is untimely, provided at a stage where it
`
`was too late for Corephotonics to provide responsive opinions from its own
`
`expert. (cid:128)ere is no reason that Apple could not have provided this analysis of
`
`this configuration—which it contends is “represented” by Ogino Table kk and
`
`which it relies upon to satisfy the TTL/EFL ratio less than k.j limitation—
`
`x(cid:136)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`together with its original petition. Accordingly, this untimely analysis should
`
`be disregarded.
`
`(cid:128)is untimely Zemax analysis also does not demonstrate what Apple im-
`
`plies it does. Apple suggests that lens design software automatically “provides
`
`the location of the image plane where the image quality has been optimized”
`
`(Paper ii at ki) and that the Zemax modeling unambiguously determines the
`
`TTL if the cover glass is removed from Ogino configuration o (id. at ko). (cid:128)e
`
`deposition testimony of Apple’s expert demonstrates otherwise.
`
`Dr. Sasián testified that the location of the Zemax “image surface” (la-
`
`beled “IMA” in Dr. Sasián’s printouts) is determined by the thicknesses of the
`
`layers that proceed it and that those thicknesses can be entered by hand. (Ex.
`
`ijio at lk:kl–ll:kp.) He testified that the Zemax image surface can be
`
`changed arbitrarily without changing the lens properties and that it does not
`
`necessarily match the optimal image location:
`
`Q. So without changing the shapes or the optical properties
`of any of the lens elements, a person using Zemax could change
`the location of the image surface for this lens system by enter-
`ing a different value for the thickness in the row for surface 12,
`correct?
`
`x(cid:137)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`A. Yes. And in that regard, the labeling of “Image” by Ze-
`max might not necessarily coincide with the location of the im-
`age.
`
`(Ex. ijio at lr:p–kl.) Indeed Dr. Sasián testified that the image surface iden-
`
`tified in Zemax is not necessarily the “image plane”:
`
`Q. So is it your testimony that, just because the lens pre-
`scription in Zemax labels the surface as the image surface, that
`that’s not necessarily the location of the image plane?
`A. For a given lens structure, yes.
`
`(Ex. ijio at lp:kj–kl.)
`
`(cid:128)e image surface in Dr. Sasián’s Zemax model was not the result of a
`
`process to “optimize[] . . . image quality,” as Apple’s reply suggests. (Paper
`
`ii at ki.) Rather, he used one of twelve different thickness “solves” available
`
`in the Zemax software to determine the thickness of a particular air layer:
`
`Q. . . . So page (cid:132)(cid:137)(cid:131) is the beginning of Chapter x(cid:136) of this
`Zemax manual, titled "Solves." You see that?
`A. Yes.
`Q. And you would agree that this page and the next page
`lists certain solves that can be used for calculating curvature or
`thickness or other parameters in a lens design?
`MR. MCDOLE: Objection. Form.
`
`x(cid:140)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`THE WITNESS: These two pages provide a number of
`solves, yes.
`BY MR. RUBIN:
`Q. And the solve that you used to calculate the thickness
`of x.(cid:132)(cid:127)(cid:132) was the marginal ray height solve; is that right?
`MR. MCDOLE: Objection. Form.
`THE WITNESS: I used the one indicated with the code M,
`as in Mary.
`
`(Ex. ijio at rj:ik–rk:ki; Ex. ijir at lpn–lpl.) Even after choosing that one
`
`solve out of the available options, Dr. Sasián had a choice of values to use for
`
`the two parameters to that solve, and different choices for those parameters
`
`could have changed the total length calculated by Zemax. (Ex. ijio at ri:p–
`
`ij.)
`
`(cid:128)e result of the solve that Dr. Sasián used was to place the image surface
`
`at the “paraxial focus”:
`
`Q. And this says that the marginal ray height solve can be
`used to constrain the image surface to the paraxial focus; is that
`right?
`A. Yeah, it says, “which can be used to constrain the im-
`age surface to the paraxial focus,” yes.
`Q. And in order to constrain the image surface to the par-
`axial focus, you set the marginal ray height value to zero and
`the pupil zone to zero; is that right?
`
`x(cid:142)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`A. As I recall, yes.
`
`(Ex. ijio at ri:i–kj.)
`
`As Corephotonics demonstrated in its response to the petition—and Ap-
`
`ple does not dispute—this paraxial focus is not necessarily the location to ob-
`
`tain optimal image quality. (Paper ik at ip–iq.) (cid:128)ere is no teaching in Ogino
`
`to suggest that the sensor and image plane should be located at this paraxial
`
`focus if the cover glass is removed from configuration o, rather than at some
`
`other location.
`
`E. Dr. Moore did not testify that Ogino example 6 discloses the
`TTL/EFL ratio limitation
`
`Apple again mischaracterizes Dr. Moore’s testimony. For example, Ap-
`
`ple claims that Dr. Moore agrees that Ogino example o does not require a
`
`cover glass and that it meets the TTL/EFL limitation when the cover glass is
`
`removed. (Paper ii at kr.) But the testimony that Apple cited never mentions
`
`Ogino example o, let alone says it meets the TTL/EFL limitation when the
`
`cover glass is removed. (Ex. kjir, ro:ij–rp:kj.) While he agrees that the
`
`cover glass is optional in the inventions of the ’(cid:143)(cid:131)(cid:127) and ’(cid:137)x(cid:127) patents (two Core-
`
`photonics patents), he testifies that cover glasses serve an important purpose
`
`of preventing dust contamination.
`
`(cid:127)(cid:143)
`
`
`
`Case No. IPR2019-00030
`U.S. Patent No. 9,857,568
`
`Apple also mischaracterizes Dr. Moore’s testimony concerning Dr. Sas-
`
`ián’s calculations on page (cid:131)(cid:143) of Exhibit kjjn in IPRijkq-jkklj. In those cal-
`
`culations, Dr. Sasián calculated TTL after removing the cover glass and mov-
`
`ing the image plane (and imaging device kjj) to a location x.(cid:132)(cid:127)(cid:132) mm from the
`
`fifth lens element. IPRijkq-jkklj, Ex. kjjn at iq–nj. Dr. Moore agreed that
`
`Dr. Sasián had correctly calculated what TTL would be under those circum-
`
`stances. But Dr. Moore did not concede that Ogino actually described that
`
`combination of modificati