Apple Computer Inc. v. Burst.com, Inc.

Filing 115

Reply Memorandum re 92 MOTION for Summary Judgment OF INVALIDITY BASED ON KRAMER AND AT&T PATENTS filed byApple Inc.. (Brown, Nicholas) (Filed on 6/21/2007)

Apple Computer Inc. v. Burst.com, Inc.

Doc. 115

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

MATTHEW D. POWERS (Bar No. 104795) [e-mail] GARLAND T. STEPHENS (admitted N.D.C.A., Texas Bar No. 24053910) [e-mail] NICHOLAS A. BROWN (Bar No. 198210) [e-mail] WEIL, GOTSHAL & MANGES LLP Silicon Valley Office 201 Redwood Shores Parkway Redwood Shores, CA 94065 Telephone: [phone] Facsimile: [phone] Attorneys for Plaintiff APPLE COMPUTER, INC. UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA

APPLE COMPUTER, INC., Plaintiff, v. BURST.COM, INC., Defendant.

No. C 06-0019 MHP APPLE COMPUTER, INC.'S REPLY IN SUPPORT OF ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY Date: July 19, 2007 Time: 2:30 pm Hon. Marilyn Hall Patel Complaint Filed: January 4, 2006 Trial Date: February 26, 2008

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case No. C 06-0019 MHP SV1:\273911\03\5vcn03!.DOC\15096.0006

Dockets.Justia.com

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 2 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 IV. III. I. II.

TABLE OF CONTENTS Page INTRODUCTION .............................................................................................................. 1 KRAMER ANTICIPATES THE AUDIO CLAIMS AT ISSUE........................................ 2 A. The "DPCM Encoding" In Kramer Is Compression, As Shown By The Text Of Kramer And The Admissions Of Burst's Experts ..................................... 3 B. Kramer Discloses Faster-Than-Real-Time Transmission Between Memory Cards ....................................................................................................................... 6 1. Kramer Shows That Its Card-To-Card Transfer Speed Is "At A Speed Much Faster Than That Required For Actual Sound Reproduction" ............................................................................................. 7 2. Burst's "Interleaving Of Sub-Band Data" Argument Should Be Rejected Because Its Contradicts The Clear Language Of Kramer............ 8 C. Kramer's Discloses A Memory Card Containing Random Access Storage (limitation only in `995 patent claims).................................................................... 9 D. Kramer Discloses The "Transceiver Apparatus" Limitations (limitation only in `995 patent claims).................................................................................... 10 E. Kramer Discloses The Limitations Of The Dependent Claims ............................ 12 KEPLEY ANTICIPATES THE AUDIO CLAIMS AT ISSUE ....................................... 12 A. B. C. D. E. Kepley Discloses Faster-Than-Real-Time Transmission...................................... 13 Kepley's Voice Mail Qualifies As "Audio/Video Source Information" .............. 14 Kepley's Disclosure Satisfies The "Compression Means" Element (limitation only in `995 patent claim 1) ................................................................ 15 Kepley Discloses The "Common Housing" Requirement (limitation only in `995 patent claims)................................................................................................ 16

Kepley Teaches Re-Transmitting Received Voice Mail (limitation only in claims `995-17 and `839-17)................................................................................. 17 CONCLUSION ................................................................................................................. 18

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

i

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

TABLE OF AUTHORITIES Page(s) CASES Bristol-Myers Squibb Co. v. Ben Venue Labs., Inc., 246 F.3d 1368 (Fed. Cir. 2001 .......................................................................................... 12 Default Proof Credit Card Sys. v. Home Depot U.S.A., Inc., 412 F.3d 1291 (Fed. Cir. 2005)................................................................................. 1, 3, 12 Jamesbury Corp. v. Litton Indus. Prods. Inc., 756 F2d 1556 (Fed. Cir. 1985)........................................................................................... 1 KSR Int'l Co. v. Teleflex Inc., __ U.S. __, 127 S.Ct. 1727 (April 30, 2007) ...................................................................... 1 MEHL/Biophile v. Milgraum, 192 F.3d 1362 (Fed. Cir. 1999)................................................................................. 1, 3, 12 Motorola, Inc. v. Interdigital Tech.Corp., 121 F.3d 1461, 1473 (Fed. Cir. 1997)............................................................................. 1, 3 Schering Corp. v. Geneva Pharms., Inc., 339 F.3d 1373 (Fed. Cir. 2003)......................................................................................... 11 Seachange Int'l, Inc. v. C-COR Inc., 413 F.3d 1361 (Fed. Cir. 2005)......................................................................................... 10 Upsher-Smith Labs., Inc. v. Pamlab, L.L.C., 412 F.3d 1319 (Fed. Cir. 2005)......................................................................................... 10

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

ii

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 4 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

I.

INTRODUCTION1 Apple's motion showed that Burst did not actually invent the so-called

"fundamental Burst function of faster-than-real-time transmission." The Kramer patent discloses sending compressed audio data between memory cards at "a speed much faster ... than that required for actual sound reproduction," so that those cards can be recorded in a "very short time" and used as "a replacement of conventional discs or cassettes" in a music distribution system. AT&T's Kepley patent discloses transmitting compressed voicemail between voicemail systems "faster than real time speech." Burst's opposition relies on the declaration of two experts, Dr. Hemami and Dr. Gersho, in an effort to show that despite these clear statements, neither Kramer nor Kepley actually discloses faster-than-real-time transmission. However, expert testimony that contradicts the plain language of a reference is insufficient to create a genuine issue of fact.2 Burst also makes a variety of other arguments that are contradicted by the disclosure of the prior art references, the testimony of its experts, or the disclosure of its patents. For example, Burst argues that the "DPCM [differential pulse code modulation] encoding referred to in Kramer does not constitute compression." However, in Burst's claim construction tutorial, Dr. Hemami stated in her slides that "DPCM coding" is an example of the "conventional algorithms" used for "audio compression in the Burst Patents." Dr. Hemami also told the Court that "differential pulse code modulation was and is very commonly used for audio coding. The
1

Apple's motion was filed on January 4, 2007, prior to the claim construction hearing, and was deliberately limited to Burst's audio-only claims and to anticipation to keep the issues more manageable. Since then, the Court has issued a claim construction order in this case, and the Supreme Court issued a significant opinion about the law of obviousness, KSR Int'l Co. v. Teleflex Inc., __ U.S. __, 127 S.Ct. 1727 (April 30, 2007). Accordingly, Apple plans to file shortly a further motion requesting summary judgment of invalidity of all 36 of Burst's asserted claims, based on both anticipation and obviousness, that takes into account both the claim construction order and the Supreme Court's decision in KSR. Apple would not object if both motions were consolidated for hearing. 2 For instance, in MEHL/Biophile v. Milgraum, 192 F.3d 1362, 1367 (Fed. Cir. 1999), the Federal Circuit granted summary judgment of invalidity notwithstanding the plaintiff's expert testimony, because that testimony contradicted the plain language of a prior art reference. See also Default Proof Credit Card Sys. v. Home Depot U.S.A., Inc., 412 F.3d 1291, 1299, 1300 n.2 (Fed. Cir. 2005) (expert testimony contradicting the express language of a patent is insufficient to prevent summary judgment for indefiniteness); Motorola, Inc. v. Interdigital Tech.Corp., 121 F.3d 1461, 1473 (Fed. Cir. 1997) (expert testimony that is inconsistent with the disclosure of the prior art is insufficient as a matter of law to support a jury verdict); Jamesbury Corp. v. Litton Indus. Prods. Inc., 756 F2d 1556, 1563 (Fed. Cir. 1985). 1
Case No. C 06-0019 MHP SV1:\273911\03\5vcn03!.DOC\15096.0006

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

specifications mentions a form of DPC[M], which is fibonacci delta compression."3 Similarly, Burst argues that the bubble memory disclosed in Kramer is not "random access storage," even though one of its own claims requires "random access storage means comprising a bubble memory," and Kramer describes bubble memory cards that "allow immediate recall of the data in any portion of the memory." In short, as shown below, Burst's opposition fails to raise any genuine issues of fact that could preclude summary judgment. II. KRAMER ANTICIPATES THE AUDIO CLAIMS AT ISSUE Kramer, which was filed in 1982, anticipates Burst's claims because it discloses sending compressed audio data between memory cards at "a speed much faster (at least 100 times) than that required for actual sound reproduction." Kramer at 4:24-26. This allows the

cards to be recorded in a "very short time" and used as "a replacement of conventional discs or cassettes" in a music distribution system. Kramer at 4:24-26, 4:6-8, 6:23-25. The memory cards described in Kramer each contained an "8 megabyte memory" which held "at least 3 1/2 minutes of music." Id. at 3:35-38. A single card could be played in a portable system with headphones, or multiple cards could "be played in any specified sequence" in system for "home, vehicle, or commercial applications." Id. at 6:13-17. For distribution of music, a "record shop can have a stock of `blank' cards and can encode one with the desired piece of music from a data store in the shop," which is "held on one of the card systems of the invention instructed to give a digital output." Id. at 6:25-30. Because recording onto a consumer's card could "be rapidly performed in a shop, there are considerable savings in costs of reproduction [and] distribution." Id. at 6:3637. Burst relies on four arguments to distinguish Kramer from its claimed inventions: (1) that the "DPCM encoding referred to in Kramer does not constitute compression"; (2) that the "transfer from card to card" in Kramer does "not transmit a work away from the card faster than real time"; (3) that "Kramer does not disclose random access storage," and (4) that Kramer does not meet the "common housing" requirement of the "transceiver" limitation.4 Each of these Kalay Decl., Exh. A [Burst Tutorial Slide 54]; Kalay Decl., Exh. B [Tutorial Tr.] at 57-58. 4 Opp. at 7:23, 11:8-10, 11:25, and 16:14-15.
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY 3

2

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 6 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

arguments is clearly contradicted by the plain language of the Kramer reference, Burst's own experts, or both, and thus none of them provides a basis for denying summary judgment.5 A. The "DPCM Encoding" In Kramer Is Compression, As Shown By The Text Of Kramer And The Admissions Of Burst's Experts Burst argues that the "DPCM encoding referred to in Kramer does not constitute compression."6 However, in Burst's claim construction tutorial, Dr. Hemami explained in her slides that the "audio compression in the Burst Patents" uses "conventional algorithms" such as "DPCM coding." Dr. Hemami also told the Court that "Fibonacci delta compression"--the only audio compression algorithm mentioned in Burst's patents--was a form of DPCM: "Differential pulse code modulation was and is very commonly used for audio coding. The specifications mentions a form of DPC[M], which is Fibonacci delta compression."7 Burst's opposition relies on a new expert,8 Dr. Gersho, to tell the Court in carefully couched words that Kramer does not "employ `compression' as the Court has construed that term." Gersho Decl., ¶ 11. To reach this conclusion, Dr. Gersho relies on the fact that DPCM, which records the differences between samples, can be implemented using either an analog circuit or a digital circuit to calculate the differences between samples. Gersho Decl. at ¶ 11; Bellamy at 131-132. If a digital differencing circuit is used, the information is digitized first, and then differences between samples are calculated. If an analog differencing circuit is used, the

difference between samples is calculated first, and then this difference is digitized. Id. The output of the two methods is the same. The reduction in the number of bits required to represent the final data is therefore exactly the same relative to an uncompressed representation.
5

See MEHL/Biophile v. Milgraum, 192 F.3d at 1367; Default Proof Credit Card, 412 F.3d at 1300 n.2; Motorola, 121 F.3d at 1473.
6

Opp. at 7:23; see Opp. at 1:22-23 (leading with "Kramer just doesn't teach compression"). Kalay Decl., Exh. A [Burst Tutorial Slide 54]; Kalay Decl., Exh. B [Tutorial Tr.] at 57-58; see also id. at 37 ("Now, compression is sometimes referred to as encoding or coding. ... In this case you will see me throughout the tutorial using them, encoding and coding, I'm using them as synonyms for compression.").
7 8

Dr. Hemami submitted a declaration in opposition to Apple's motion that states that she has "read and studied" Kramer. Hemami Decl. ¶ 9. However, she offers no opinion suggesting that Kramer differs in any way from Burst's claims. Burst has refused to let Apple take a brief deposition of Dr. Hemami to inquire about this, or about the opinions she does express. 3
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Nevertheless, Dr. Gersho argues that creating DPCM data using an analog difference calculation is not "compression" under the Court's construction because the analog signal was not represented in bits before being compressed, and therefore the method did not "reduce" the number of bits in the signal. Gersho Decl. ¶ 11. And since, according to Dr. Gersho, this analog differencing is what "one with ordinary skill in the art would understand" Kramer to disclose, Kramer does not disclose compression under the Court's construction. This argument is flawed on several levels. To begin with, it requires reading the Court's claim construction order to imply that while encoding audio data into DPCM format using a digital difference calculation is compression, encoding the same audio data into the same DPCM format using an analog difference calculation is not, because for an analog difference calculation there are never any "bits" whose number is reduced. The language of the Court's construction of "time compressed representation" -- "a version of audio/video source information having a reduced number of bits" -- implies only that the number of bits must be reduced relative to another representation, not that bits must be created and then discarded. There can be no genuine dispute that DPCM is compression under the Court's construction. For example, one of the textbooks cited by Dr. Gersho in his declaration states that DPCM is "designed specifically to take advantage of the sample-to-sample redundancies in a typical speech waveform. Since the range of sample differences is less than the range of

individual samples, fewer bits are needed to encode difference samples."9 Another textbook cited by Dr. Gersho opens its "Differential PCM" chapter by explaining that it describes "differential coding or predictive coding systems where waveform redundancy is utilized in time-domain operations to realize straightforward reductions in bit rate."10 Moreover, even under Burst's strained view of the Court's construction, Burst's argument requires an untenably restrictive reading of Kramer. Kramer never states one way or the other whether the DPCM it describes uses analog or digital differencing. Burst relies entirely

9

Kalay Decl., Ex. C [Bellamy's Digital Telephony] 130 (underlining added); see also id. at 131 (a "DPCM system use[s] 2/3 bits per sample less than a PCM system with the same quality"). 10 Kalay Decl., Exh. D [Jayant & Noll, Digital Coding of Waveforms] at 252 4
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 8 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

on Dr. Gersho's declaration that "one with ordinary skill in the art would understand the reference to `encod[ing] .. into digital form' using `DPCM' at Kramer 3:10 to refer to converting an analog signal directly into a digital signal having a DPCM format." But this argument is undermined by the Bellamy textbook cited by Dr. Gersho, which states that "digital processing . . . is generally the most effective means of implementing a DPCM algorithm."11 Kramer's silence on whether to use analog or digital differencing to implement DPCM cannot reasonably be read to exclude what is "generally the most effective means" of implementing DPCM, notwithstanding Dr. Gersho's unexplained statement to the contrary. Finally, the fact that Kramer discloses compression is evident from the language of the reference itself, and not just from its reference to DPCM. Kramer states that "an 8 megabyte memory 22 should allow recording of at least 3 1/2 minutes of music." 3:35-37. That statement alone shows that the audio is compressed: according to Dr. Hemami, "a digitized uncompressed wideband audio signal" uses approximately 18.5 megabytes to store 3 1/2 minutes of music.12 Moreover, Kramer states that it computes the differences between audio samples, and between audio channels, so as be "more efficient" and produce a "substantial saving in information": A pulse code modulation coder quantizes sampled sound amplitudes; the differential technique is more efficient and utilizes the redundancies present in the sound . . . For recording in stereo, the encoding preferably is done by encoding one channel in terms only of its difference from the other channel, usually this difference will be small enough to allow encoding with substantial saving in information over that needed for separate encoding of two channels." Kramer at 3:14-35. In short, there can be no genuine dispute that Kramer discloses the claimed compression.

22 23 24 25 26 27 28
11 12

B.

Kramer Discloses Faster-Than-Real-Time Transmission Between Memory Cards Burst makes two inconsistent arguments in an effort to show that Kramer's

Kalay Decl., Exh. C [Bellamy's Digital Telephony] at 133.

Dr. Hemami states in her declaration that "a digitized uncompressed wideband audio signal [such as from a CD] has a data rate of 705.6 Kilobits/second." Hemami Decl. ¶ 22. Multiplying 705.6 Kilobits/second by 210 seconds (3 1/2 minutes) dividing by 8 bits per byte shows that digitized uncompressed wideband audio uses approximately 18.5 megabytes in 3 1/2 minutes. 5
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

disclosure of a memory that outputs at "a speed much faster (at least 100 times) than that required for actual sound reproduction" is not actually a disclosure of faster-than-real-time transmission. Kramer at 4:24-26. First, Burst argues that Kramer describes only "transfer that occurs between internal components" as faster-than-real-time, and "never states the speed at which transfer from one memory card to another would occur." Opp. at 9:22-24. Second, Burst argues that, contrary to the plain language of the reference and to its first argument, the output of Kramer's memory internally is actually only at a "real-time" data rate because of supposed "interleaving of sub-band data." Opp. at 11:8-24. Both of these arguments are contradicted by the plain language of the Kramer patent. 1. Kramer Shows That Its Card-To-Card Transfer Speed Is "At A Speed Much Faster Than That Required For Actual Sound Reproduction"

Burst argues that Kramer "never states the speed at which transfer from one memory card to another would occur." In fact, Kramer states that "memory 22" is "clocked at the same speed . . . during recording and replay." Kramer at 4:1-4. Furthermore, Kramer expressly states that "memory 22" outputs "at least 100 times" faster-than-real-time during "replay": For replay of the recorded data . . . the controller 34 instructs the memory controller clock 36 to prepare the memory 22 for output of its stored data; the controller also will instruct the demultiplexer 24 (and decoders 26) how to deal with the data which they will receive, if these not adequately programmed into these components. This output will be at a speed much faster (at least 100 times) than that required for actual sound reproduction. Kramer at 4:9-26 (emphasis added). Therefore, recording "can take a very short time." Kramer at 4:6-8. Kramer explains that the faster-than-real-time output of memory 22 can be passed directly to the input of another memory card: In another mode of use, the system can be programmed so that the output is of the data in its digital form; the controller 34 then instruct [sic] the output register 30 to pass the data from the memory 22 directly to the output multiplexer 32, without passing through the [d]ecoders 26; such a digital output can be used, as mentioned above, as input into another memory, e.g. of another portable card of the invention. Id. at 5:5-12 (emphasis added). This shows, as Burst's expert Dr. Gersho admits, that "memory 22" of one memory card can be used to feed "direct digital output" into another memory card, and
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

6

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 10 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

that "there is no indication in the patent to support the idea that the card to card transfer rate is different" than "the rate for feeding data to the demultiplexer." Gersho Decl. ¶ 14. Given this admission (which in any event is clear from Kramer) and the express disclosure that the "memory 22" in the memory cards operates "at the same speed . . . during recording and replay," there can be no dispute that the card to card transfer speed is the same as the output speed of the memory, which Kramer describes being "at a speed much faster (at least 100 times) than that required for actual sound reproduction." 2. Burst's "Interleaving Of Sub-Band Data" Argument Should Be Rejected Because Its Contradicts The Clear Language Of Kramer

Despite Kramer's statement that the memory operates "at a speed much faster (at least 100 times) than that required for actual sound reproduction," Burst argues that Dr. Gersho's declaration shows that the memory actually only operates at a "real-time" data rate. Dr. Gersho's declaration on this subject has two parts, each of which is wrong. First, Dr. Gersho declares that because the set of components (consisting of "the multiplexer, decoders and analog signal mixer") that receives the output of memory 22 during audio playback themselves have "no memory," "one of ordinary skill would understand that the speed at which the data is delivered from the memory is the speed needed for real time reproduction of the sound." Gersho Decl. ¶ 13. This logic is flawed. The fact that "the

multiplexer, decoders and analog signal mixer" have no memory does not imply that the decoders must receive the sound data at a real time rate. Indeed, Kramer expressly discloses that this is not true, and explains why: each "decoder can read the data at the required slower reproduction rate by taking, e.g. only one out of every 100 bits of information presented to it at a time." Id. at 4:4552. This shows that each decoder receives data faster than real time, but uses the data at the "required slower reproduction rate" by ignoring 99 out of every 100 bits it receives. Thus, in order to play a song, the memory has to send its contents to each decoder 100 times, and each time, the decoder reads only one out of every 100 bits, so it can achieve the "required slower reproduction rate."13 Of course, when the Kramer system is used for card-to-card transmission
13

As Kramer goes on to explain, in order to accommodate the fact that each decoder ignores 99 7
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

instead of playback, there is no need to slow the data down to a real-time rate, and thus no need to discard 99 out of every 100 bits. This provides Kramer with an important advantage: because the memory operates at a faster-than-real-time rate, recording "can take a very short time." Kramer at 4:6-8. As a result, customers in a record shop can load a new song onto a memory card in just a few seconds, instead of having to wait for several minutes. See Kramer 6:34-37. Dr. Gersho's second argument that Kramer's memory actually only operates at a "real-time" data rate is based on the assumption that "although Kramer does not explicitly state this, one of ordinary skill in the art would understand that [Kramer] illustrates the use of sub-band coding for the example of 100 sub-bands." Id. at ¶ 12. This assumption is wrong, because Kramer illustrates using 4 frequency sub-bands, not 100. Specifically, Kramer says that the "number of frequency bands may correspond to the number of instruments/voices in an ensemble, and should correspond to the number of decoders 26." Kramer at 3:21-30. Figure 1 of Kramer shows 4 decoders, not 100. Moreover, Dr. Gersho does not explain how 100 decoders would fit into Kramer's credit-card sized device. Based on the incorrect and unsubstantiated assumption "one of ordinary skill in the art would understand that [Kramer] illustrates the use of sub-band coding for the example of 100 sub-bands," Dr. Gersho concludes that the output of Kramer's memory is actually at 100 times the rate needed to decode only one of the postulated 100 sub-bands, and thus overall is a "realtime transmission" because 100 concurrent data streams are needed for the postulated 100 subband decoders. Id. at ¶ 13. Again, Kramer shows that this is not correct. Kramer expressly says that the memory output is "at a speed much faster (at least 100 times) than that required for actual sound reproduction." "Actual sound reproduction" requires reproduction of all of the sub-bands in a signal, not just one of them, so this statement shows that the data for all of the sub-bands is output "at a speed much faster" than real time. Accepting Dr. Gersho's "sub-bands" argument would require rewriting Kramer to say that its memory output at "a speed much faster than that out of every 100 bits it receives, the memory stores the audio data in "an interleaved fashion." Kramer at 4:45-52. This interleaving is not related to the supposed 100-fold sub-band encoding Dr. Gersho describes, it is related to the fact that Kramer's decoders only 1 out of every 100 bits it receives in order to achieve the "required slower reproduction rate."
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

8

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 12 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

required for individual sub-band sound reproduction." C. Kramer Discloses A Memory Card Containing Random Access Storage (limitation only in `995 patent claims) Burst argues that "Kramer does not disclose random access storage." Opp. at 11. Burst is wrong. Kramer states expressly that "the systems of the present invention" are portable cards with bubble memories that are "arranged so as to allow immediate recall of the data in any portion of the memory." Kramer at 1:23-24, 1:35-37. There can be no genuine dispute that these bubble memories are a type of random access memory. Burst's own patents say so expressly, requiring in Claim 2 of the `932 patent "random access storage means comprising a bubble memory." Burst even acknowledges that its patents "mention using bubble memory for random access storage." Opp. at 12 n.5. Nonetheless, Burst argues that "the memory in Kramer is specifically described as a serial access device," and is "specifically distinguished from memory configured to `allow immediate recall of the data in any portion of the memory." Opp. at 12:12-16. This argument

mischaracterizes Kramer. The passages Burst cites show that Kramer's memory is described as being "organized to appear to be a circular shift register," and as being "used ... for storage of data in digital form for retrieval serially." Kramer at 4:1-4, 2:25-29 (emphasis added). But the fact that Kramer describes using a bubble memory card to record data "in serial order" and having it appear to be a "circular shift register," does not suggest that the bubble memory card has ceased being a random access storage device. Hard discs do not suddenly stop being random access storage devices when they are used to record a stream of serial data, like audio and video information. Moreover, Kramer states expressly that "the systems of the present invention" are portable cards whose memory allows "immediate recall of the data in any portion of the memory." Kramer at 1:23-24, 1:35-37. The fact that these random access bubble memory cards are referred to as "the systems of the present invention" shows that Burst is simply wrong when it says that Kramer "disavows and teaches away from `random access storage.'" Describing the use of random access memory for storage of serial data is not "teaching away" from random access

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

9

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 13 of 21

1 2 3 4

storage.14 D. Kramer Discloses The "Transceiver Apparatus" Limitations (limitation only in `995 patent claims) As Burst correctly points out, the parties have agreed that the term "transceiver

5 apparatus" means "a combination of components, in a common housing." Claim 17 of the `995 6 patent requires a "transceiver apparatus" that receives compressed audio/video data faster-than7 real-time, stores it, and then transmits it faster-than-real-time. Claim 1 of the `995 patent requires 8 a "transceiver apparatus" that receives audio/video data, compresses it, stores the compressed 9 data, and then transmits the stored data faster-than-real-time. 10 For the reasons described above, Kramer's memory cards satisfy the "transceiver 11 apparatus" limitation of claim 17, because each memory card is designed to receive compressed 12 audio data faster-than-real-time, store it in random access memory, and then transmit that data to 13 another memory card faster-than-real-time. Burst does not argue to the contrary (other than 14 indirectly through the arguments addressed above). 15 Burst also argues the "common housing" requirement is not met for claim 1 of the 16 `995 patent because the "compression means" is not shown as being inside "the housing for the 17 replay unit or memory card," and because Kramer states that "the music signal is encoded 18 (outside the illustrated system")." Opp. at 6-7, Kramer at 3:9-10. It is true that neither the 19 memory card illustrated in Figure 1, nor the replay unit illustrated in Figure 2, contain a device 20 that performs compression. 21 However, by stating that "the music signal is encoded (outside the illustrated 22 system" Kramer makes clear that the "memory card" illustrated in Fig. 1 and the "replay unit" 23 illustrated in Fig. 2 are not the only devices involved. While it does not illustrate it in a figure, 24 Kramer does disclose an encoding system. Kramer at 3:42-45 ("With a bubble memory, it is 25 26 27 28 Burst's "teaching away" argument is legally wrong as well as factually wrong. The Federal Circuit has held that "teaching away is irrelevant to anticipation," and explained that "a reference is no less anticipatory if, after disclosing the invention, the reference than disparages it." Seachange Int'l, Inc. v. C-COR Inc., 413 F.3d 1361, 1380 (Fed. Cir. 2005); Upsher-Smith Labs., Inc. v. Pamlab, L.L.C., 412 F.3d 1319, 1323 (Fed. Cir. 2005).
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY 14

10

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 14 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

necessary to provide, in the encoding and replay systems, means for generating a rotating magnetic field...") (emphasis added). This "encoding system" inherently has all the components necessary to compress, store, and transmit the compressed data faster-than-real-time, as required by claim 1 of the `995 patent.15 Specifically, the encoding system must have a "compression means" since its purpose is to encode audio data into a compressed (DPCM) format. It must have "transmission means" since its purpose is to load that data into the memory cards. This process of loading a memory card must occur faster-than-real-time. As discussed above, the memory card is described as operating at the same, faster-than-real-time rate for both recording and replay, and recording is described as happening in a very short time. And finally, the "encoding system" must have some place to store the data it generates as it performs the compression. Indeed, the passage quoted above describes the encoding system as using a bubble memory. Admittedly, Kramer does not say expressly whether this "encoding system" is contained within a single housing. However, Dr. Hemami states in her declaration that the word "system" suggests "a single stand-alone unit" with its own housing, and that one of ordinary skill in the art would so understand. Hemami Decl., ¶ 30. Under this reasoning, Kramer discloses the "encoding system" in a single housing, and hence anticipates the "transceiver apparatus" limitation. E. Kramer Discloses The Limitations Of The Dependent Claims16 Burst does not dispute that Kramer's memory cards are a "removable recording medium," which is the element added by claims 44 and 47 of the `839 patent and claim 44 of the `995 patent. Thus, these claims are anticipated for the reasons already described. Dependent claim 9 of the `839 and `995 patents adds the requirement that the source information be digital. As shown in Apple's motion, Kramer describes using a controller that is "programmed in the factory so as to correctly deal with digitized analogue data." Kramer at 3:5-7. Moreover, as Burst itself has argued, the DPCM encoding described by Kramer could
15

Something not expressly described in a reference is nonetheless disclosed if it is inherently (i.e. necessarily) present in order to accomplish what is described. See, e.g., Schering Corp. v. Geneva Pharms., Inc., 339 F.3d 1373, 1378-79 (Fed. Cir. 2003) (finding that a patent describing a drug inherently disclosed the compound formed by the drug in the human body).
16

Burst has withdrawn its infringement contentions regarding Claim 16 of the `995 and `839 patents, so Apple has agreed to withdraw without prejudice its motion as to those claims. 11
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

be used to process either digital or analog audio data. The Federal Circuit has held that "the disclosure of a small genus may anticipate the species of that genus even if the species are not themselves recited."17 Thus, under Burst's argument, Kramer's disclosure of DPCM compression is disclosure of both the analog and digital species of audio "source information." Finally, Apple withdraws without prejudice this motion as to claim 15 of the `839 patent and `995 patents. These claims will be addressed in the supplemental motion for summary judgment of invalidity described above that Apple will file shortly. III. KEPLEY ANTICIPATES THE AUDIO CLAIMS AT ISSUE The Kepley patent anticipates Burst's claims because it describes a voice mail system where compressed voicemails to be transmitted "faster than a realtime voice message transmission." It describes receiving voice data, compressing "the voice data from 64k bits/s down to 16k bits/s," storing the compressed message in "disk storage," and then transmitting the compressed voicemail between voicemail systems "faster than real time speech" using "digital high speed transmission facilities." Burst makes five arguments in an effort to distinguish Kepley from its claims: (1) that the "Kepley voice mail system is incapable of the fundamental Burst function of faster-than-real-time transmission"; (2) that voice mail is not the claimed "audio/video source information"; (3) that Kepley does not "employ[] the specific methods required by the Court's construction of `compression means'"; (4) that Kepley "fails to meet the `common housing' requirement of the `transceiver apparatus' limitation"; and (5) that Kepley fails to teach "that the recipient can re-transmit the received voice mail after listening to it."18 Each of these arguments is contradicted by the plain language of the Kepley reference or of Burst's own patent, or by Burst's own experts, and none of them provides a basis for denying summary judgment.19 A. Kepley Discloses Faster-Than-Real-Time Transmission The Kepley patent, which issued to AT&T, expressly states that the "use of digital high speed transmission facilities of speed greater than 9.6 Kbps enables the exchange of digitally Bristol-Myers Squibb Co. v. Ben Venue Labs., Inc., 246 F.3d 1368, 1380 (Fed. Cir. 2001). Opp. at 2:7-8, 2:11-12; 2:14-15, 16:14-15, and 23:18-19. 19 See MEHL/Biophile, 192 F.3d at 1367; Default Proof Credit Card, 412 F.3d at 1300 n.2.
18 APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY 17

12

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 16 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

encoded and compressed voice mail messages faster than real time speech." Kepley at 13:31-37. In an effort to side-step this clear statement, Burst argues that AT&T's design would not work. Burst labels the disclosure "purely aspirational" and relies on expert testimony to argue that the "Kepley system is incapable of faster than real-time transmission." Opp. at 21-22. This is nonsense. Burst's argument boils down to the claim that AT&T simply got voice transmission rates wrong. Kepley says that "the use of digital high speed transmission facilities of speed greater than 9.6 Kbps" enables faster-than-real-time transmission. Burst argues that rather than 9.6 Kbps, a bandwidth of 18.8 to 19 Kbps would have been required to transmit "voice mail as described in Kepley in real time," that "faster-than-real-time transmission would require transmission bandwidths exceeding those," and that a specific network protocol described in Kepley for use with its voice mail system had a data rate of "only 9.6" Kbps. However, "digital high speed transmission facilities" with speeds even Dr. Hemami acknowledges would be fast enough (i.e. greater than 18.8 to 19 Kbps) were well known and available from AT&T and others long before Kepley was filed. As Burst admits, Kepley itself expressly describes a 19.2 Kbps data link. Opp. at 22. Further, a T1 line, available from AT&T since at least the 1960s, had a data rate of 1.544 megabits/s.20 Similarly, as stated in Apple's motion and not contested by Burst, the classic 1976 ethernet paper by Robert Metcalfe and David Boggs describes transmission at 3 megabits/s.21 AT&T's disclosure in Kepley of using "digital high speed transmission facilities" to send "compressed voice mail messages faster than real time speech" was not aspirational, impossible, or even difficult. B. Kepley's Voice Mail Qualifies As "Audio/Video Source Information" Burst argues that Kepley's voice mails are not "audio/video source information," because "audio/video source information" must be "something produced by the exercise of creative talent" and voice mail is not. This argument fails on both counts. "Audio/video source

20

Kalay Decl., [1966 Ronne paper]. 21 Brown Invalidity Decl., Exh. 7 [1976 Metcalfe & Boggs paper].
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

13

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

information" does not require a creative element,22 and even if it did, voice mail would qualify. Burst's argument that "audio/video source information" must be the result of "creative" effort relies on its allegation that "voice processing and wideband audio processing are vastly different" due to the "divergent characteristics of speech and wideband audio." Even assuming that is true, it does not imply that voice information is not audio information, nor does it imply that "audio/video information" is limited to the result of "creative" effort. And of course voice information, including voicemail, is frequently the result of creative effort. Burst's patents themselves describe the transmission of voice-only data, explaining that "a user can dictate an audio presentation and send it to a remote location (e.g., an office) at an accelerated rate without having to monopolize the transmission medium (e.g., the fiber optic cable) for an extended length of time." `932 Patent at 12:29-34. Thus, voice-only information is plainly described as being within the scope of Burst's patents. Moreover, since dictating "an audio presentation" is no different that dictating an extended voice mail message, voice mail is plainly within scope of the "audio/video source information" Burst chose to claim. Furthermore, even if "audio/video source information" required "creative" effort, as Burst insists, voice mail would qualify. Burst concedes in its brief that voice data in the form of "audiobooks or recited poetry" are "works" that contain the supposedly required "creative" element. Opp. at 19. Like other voice data, a voice mail can reflect "creative" effort. C. Kepley's Disclosure Satisfies The "Compression Means" Element (limitation only in `995 patent claim 1) Burst argues that Kepley does not anticipate because it does not disclose the "algorithm of comparing two or more samples and coding certain differences," which is required by the Court's construction of "compression means." Opp. at 20:8-10. This is gamesmanship of the sort that Apple sought to avoid by filing this motion as part of the claim construction process. Kepley discloses that audio data is compressed using "bandwidth compression (compress the voice data from 64k bits/s down to 16k bits/s)" and "silence compression (encode the length of Apple does not agree that the term "work" in the Court's claim construction order necessarily requires a creative element. Claim Construction Order at 24.
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY 22

14

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 18 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

long silences...)." Kepley 8:29-33. Since silence is the absence of sound, silence compression necessarily compares consecutive samples and codes the differences as the number of samples that do not contain changing signals. Burst seeks to avoid summary judgment on the grounds that there are many "widely known algorithms for compressing audio from 64 kb/s to 16 kb/s," and some of them do not involve "coding certain differences between samples." Accordingly, Burst reasons, Kepley's disclosure of "bandwidth compression" and "silence compression" is not specific enough to anticipate its claims. But the Burst patents' sole disclosure of audio compression is that "the audio data can be compressed with conventional algorithms, e.g., a Fibonacci delta compression algorithm." `995 patent at 5:33-35 (emphasis added). Thus, Burst's argument boils down to the claim that there are too many "widely known algorithms" within the scope of Kepley's disclosure for it to be specific enough to anticipate Burst's disclosure of compressing using "conventional algorithms." This is nonsense. Kepley's disclosure of using "bandwidth compression" and "silence compression" is not meaningfully different from Burst's disclosure of using "conventional algorithms" for compression. Thus, Kepley anticipates the audio "compression means" in Burst's claims. D. Kepley Discloses The "Common Housing" Requirement (limitation only in `995 patent claims) Burst's argues that Kepley's storage devices are housed separately from its compression and transmission device because the "interface" between the "voice storage processor" and the "data base processor" shows "that they are separately housed devices." Opp. at 17; Hemami Decl. ¶ 29 (the "interface ... suggests that separate housings are possible and even likely"). This argument contradicts the positions Burst took, and that the Court accepted, during the claim construction process. Specifically, Burst argued then that the Izeki reference did not disclose transmitting to an external device when it described sending data through an "interface" to a "reproduction device." Claim Construction Reply Br. at 54. Now Burst and Dr. Hemami are saying that the opposite is true.
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

15

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
23

Burst's current argument is also contradicted by the "dotted lines" argument it made previously. In Izeki, dotted lines surround the "reproduction device" that Burst insisted was an internal device. Burst argued that "as Dr. Hemami explained, the dotted lines to not indicate external, or `physically separate' components, but rather functional groupings."23 Now, in the Kepley reference, there are dotted lines surrounding the "data base processor" components, yet Burst asks the Court to conclude that these components are physically separate. Burst cannot have it both ways. Burst's current position is also internally inconsistent. Burst argues, relying on Dr. Hemami, that the use "of the word `system' to describe the data base processor components" shows that they are separately housed. Opp. at 15; Hemami Decl. ¶ 29. But as shown to the right, Kepley also refers to the combination of the "voice storage processor" components and the "data base processor" components as part of a "voice mail service system." E.g. Kepley at Fig. 1, 4:33-34; 4:52-58. Under Burst's own

logic, this shows that the two groups of components share a housing. Furthermore, even if the Court were to find that Kepley's "data base processor" were housed separately from the "voice storage processor," Kepley would still anticipate claim 17 of the `995 patent, which does not require "compression means," only receiving, storing, and transmitting. This is because the voicemails sent by Kepley's voice mail system are transmitted and received by the "data base processor" portion of the voice mail system. Kepley at Fig. 2, 5:66-6:5 (stating that voice mail transfer occurs over "communication lines 104," which emerge from the "data base processor"). As explained above, the messages are compressed, and are sent faster than real-time. Moreover, as Burst acknowledges, the "data base processor" is what stores the voicemails. Thus, if the "data base processor" is separately housed, it satisfies the

"transceiver" limitation of claim 17 because it contains hardware for receiving compressed voicemail, storing it, and transmitting onwards faster than real-time.

C.f. Kalay Decl., Exh. F [Izeki] at Fig. 1; Brown Invalidity Decl Exh. 3 [Kepley] at Fig. 2. 16
Case No. C 06-0019 MHP

APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 20 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

E.

Kepley Teaches Re-Transmitting Received Voice Mail (limitation only in claims `995-17 and `839-17) Burst's argues that Kepley "wholly fails to address the concept of re-

transmission," relying on the declaration of Dr. Hemami.24 Burst is wrong. Kramer clearly describes forwarding voice mail messages. Kepley at 12:47-13:37. A forwarded voice mail has plainly been "re-transmitted": first it was sent to the original recipient, then it was re-transmitted to the recipient of the forwarded voice mail. Thus, there can be no genuine dispute that Kepley discloses re-transmitting a previously stored, compressed voice mail, as required by claim 17 of the `995 and `839 patents. F. Kepley Discloses The Limitations Of Dependent Claim 9 of the `839 and `995 Patents Dependent claim 9 of the `839 and `995 patents adds the requirement that the source information be digital. As shown in Apple's motion, Kepley shows that the voice data, before compression, is "64 k bits/s," and is thus digital. Burst argues that this digital information does not qualify as the "digital source information" required by the claim because it was analog voice information before it was converted into digital form. Under this logic, however, a CD would not provide "digital source information" because the recorded sound was in analog form before it was recorded and digitized. Accordingly, Burst's argument should be rejected. IV. CONCLUSION For the reasons stated herein, the Court should grant Apple's motion for summary judgment that claims 1, 9, 17, and 44 of the `995 patent, and claims 1, 9, 17, 44, and 46 of the `839 patent are invalid as anticipated.

Dr. Hemami never says that Kepley "wholly fails to address the concept of re-transmission." Indeed, Dr. Hemami acknowledges the opposite, stating that the "subsection titled `Voice Mail Forwarding' ... discloses digital transmission of the voice mail message." Hemami Decl. ¶ 36. All Dr. Hemami says is that "there is no discussion about implementation or execution of retransmission in Kepley," because, "for example, header construction as taught in Figure 4 is unclear for a forwarded message." Hemami Decl. ¶ 55 (emphasis added). This is irrelevant to whether Kepley discloses retransmission, and is wrong in any event: there are more than 50 lines of description in Kepley's `Voice Mail Forwarding' section, including a detailed recitation of the steps needed to accomplish the "message transmission sequence" used for forwarding a voice mail. Kepley at 12:47-13:37.
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

24

17

Case No. C 06-0019 MHP

Case 3:06-cv-00019-MHP

Document 115

Filed 06/21/2007

Page 21 of 21

1 2

Dated: June 21, 2007

WEIL, GOTSHAL & MANGES LLP By: /s/ Nicholas A. Brown Attorneys for Plaintiff Apple Computer, Inc.

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
APPLE'S REPLY ISO ITS MOTION FOR SUMMARY JUDGMENT OF INVALIDITY

18

Case No. C 06-0019 MHP