US 6278387

Summary — U.S. Patent 6,278,387 (US6278387B1)

Search results note

I searched the USPTO/Google Patents records and CAFC 2026 docket indexes for the specific number 6278387. The patent database records are consistent across sources (Google Patents, FreePatentsOnline). No CAFC appeal docket entries for 2026 (or any year) referencing this patent were found — searches for "6278387" CAFC, "6,278,387" appeal, and Synaptics/Conexant litigation involving this number returned no matching docket records. The patent expired in 2019 (see status below), so any CAFC 2026 activity would be unexpected; I have no authoritative basis to assert any pending appeal exists.

Bibliographic data

Abstract (verbatim)

An audio codec having an encoder and a decoder is disclosed. The encoder enables the compression of an audio signal for transmission or storage while the decoder receives a compressed audio signal for playback. A time scaling module within the decoder allows variation of the playback rate of the compressed audio signal. Further, no significant depreciation in the quality of pitch occurs as a result of varying the playback rate. The codec features a control for independently varying the playback rate and a module for delivering pitch compensation. The encoder utilizes a sub-band coding scheme (e.g., MPEG-1 and MPEG-2) wherein an audio signal is split into at least two frequency sub-bands for compression. ... The time scaling module time stretches or compresses an audio signal as desired using a synchronized overlap and add (SOLA) algorithm.

Plain-language overview of independent claims

Claim 1 — Audio codec (system). An audio codec that receives a first audio signal for encoding and a second audio signal for decoding, comprising: an encoder (with a memory and a processor that, in response to receiving the first audio signal, directs encoding of it into a digital code word); a decoder (with a memory and a processor that directs decoding of the second audio signal to enable playback); and a rate adjust module that permits variable playback of the second audio signal. In short: a combined encode/decode device whose playback speed is user-adjustable.

Claim 9 — Audio decoder (system). A decoder that receives a compressed audio bit stream having at least first and second frequency sub-bands, comprising: an input interface; an unformatter that unpacks the bit stream from its frame structure; an inverse bit allocate decoder that inversely allocates/de-quantizes the stream to recover input samples per sub-band; and a time scaling module that time-stretches the input samples in the time domain for each sub-band separately to enable variable playback. In short: a compressed-audio (e.g., MPEG sub-band) decoder that changes playback rate in the time domain, per sub-band, preserving pitch.

Claim 16 — Method of time-scaling (process). A method for manipulating audio samples having first and second sub-band frequencies: (1) receiving the samples; (2) forming, for each sub-band, an input frame and a first output frame; (3) computing a best averaging point within a search range for overlapping the input and first output frames; (4) overlapping the frames at that point; and (5) averaging the overlapped portions for each sub-band to form a second output frame. In short: the SOLA-style frame-overlap-and-average procedure — find the best alignment point, cross-fade, and average — applied per frequency sub-band.

The dependent claims add detail such as: MPEG-2/32-sub-band compliance (claims 7, 11, 17); the sub-band-dependent search range (Kmin–Kmax) based on sampling frequency (claims 13, 18); fade-in/fade-out averaging (claim 19); time compression or expansion (claims 14–15); and fixed vs. user-selectable input frame size (claims 20–23).

Uncertainty / caveats

• CAFC 2026 docket: No entries found for this patent number. I cannot confirm any appeal; given the patent expired in 2019, it is unlikely to be the subject of new 2026 appellate litigation, but I have no authoritative source confirming its absence from every docket listing.
• Current assignee: The Google Patents assignment chain lists Conexant Systems, Inc. → Lakestar Semi Inc. → Conexant Systems, LLC → Synaptics Incorporated (2017). I did not independently verify Synaptics' current recordal at the USPTO for this specific patent.
• The patent text provided (Google Patents) is authoritative for title, dates, claims, and abstract and is consistent with all search results returned; I found no conflicts requiring me to prefer search results over the supplied text.

Litigation Search Results — U.S. Patent 6,278,387 (US6278387B1)

No litigation involving U.S. Patent 6,278,387 was found.

I searched across multiple litigation-oriented sources and query formulations for the exact patent number 6278387 (and equivalents "6,278,387," "US6278387"):

Consistent with prior analysis

This result is consistent with the earlier patent-summary work: no CAFC (or any appellate) docket entries referencing this patent were found for any year, including 2026. The patent's assignment history shows only routine security-interest and ownership transfers (Conexant → Lakestar Semi → Conexant Systems, LLC → Synaptics Inc.), with no assignment event tied to litigation (e.g., no "litigation release," judgment, or settlement conveyance).

Assessment

• Known litigation: None. I have no authoritative basis to identify any plaintiff(s), defendant(s), jurisdiction, case number, filing date, or status for any case involving this patent.
• Why this is plausible: The patent expired (lapsed) on September 28, 2019 (20 years from filing, "Expired – Lifetime" status). Post-expiration, the patent cannot be enforced for damages or injunctive relief for acts after expiration, and past-damages claims would have had to be filed before the statute of limitations ran (6 years under 35 U.S.C. § 286), i.e., by roughly September 2019 for pre-expiration acts. This significantly reduces the likelihood of new litigation in 2026.
• Caveat (stated explicitly): Web/database searches returning no results is not an absolute negative proof. It is possible that an older district court case (e.g., a 2001–2019 case) exists that is not well-indexed in the free sources I queried (CourtListener, Google Patents, Unified Patents portal). If you need a definitive answer, I would recommend a paid docket search (PACER full-text search, Docket Navigator, or LexisNexis/RPX) across 2001–2019.

Conclusion: No litigation is known involving U.S. Patent 6,278,387 based on the available search results.

Proceedings overview

Total AIA trial proceedings on U.S. Patent 6,278,387: 0 — zero IPRs, zero PGRs, zero CBMs, by every source checked. The USPTO Open Data Portal (PTAB proceedings on file) returns no AIA trial proceedings for this patent as of the most recent ingest, and web searches across multiple query formulations ("6278387" IPR, "US6278387" PTAB, "6,278,387" inter partes review, "6278387" Rayskiy, ptab.uspto.gov trial search 6278387) returned no proceeding numbers, dockets, decisions, or appeals referencing this patent. The only near-miss hits were unrelated: U.S. Patent 6,690,387 (a Koninklijke Philips IPR, not this patent) and U.S. Patent 6,278,387A in an unrelated heating-cable application identifier string.

Bottom-line defensive posture: This patent has never been challenged at the PTAB. It is untested, not hardened — but it is also expired (status: "Expired – Lifetime"; 20-year term ran 1999-09-28 → 2019-09-28), which is the dominant defensive fact. No claims have been canceled by the Board because no petition was ever filed; conversely, no claims have been sustained by the Board either. There is no PTAB estoppel of any kind against any party, because § 315(e)(2) estoppel attaches only after a final written decision (or § 315(e)(1) after a petition is granted).

No per-proceeding sections follow because there are no proceedings to report — I will not fabricate proceeding numbers, panels, grounds, or dispositions.

Strategic summary

Claim status: all 23 claims (independent claims 1, 9, 16; dependent claims 2–8, 10–15, 17–23) are UNTESTED at the PTAB. No AIA petition has been filed, instituted, or decided against this patent. The patent's claims are therefore neither CANCELED nor SUSTAINED by any Board decision — they simply have never been the subject of an IPR/PGR/CBM. The only relevant "status" is the patent's expiry on 2019-09-28, which means the claims cannot be enforced against any act occurring after that date, and any damages claim for pre-expiry acts would have had to satisfy 35 U.S.C. § 286's six-year lookback (i.e., a complaint filed by roughly 2019-09-28 at the latest for pre-expiry infringement).

Estoppel landscape — clean slate. Because no IPR was ever instituted, no party is barred by § 315(e)(1) or § 315(e)(2). A defendant facing assertion today has the full universe of prior art available: every § 102 and § 103 ground that could have been raised in an IPR (including art never presented to the examiner, e.g., the SOLA literature, MPEG-1/MPEG-2 sub-band filter-bank references, and time-scale-modification papers predating the 1999-09-28 priority date) remains fully available in district court. There is no risk of an "IPR estoppel" trap here because there is no IPR.

Pattern signals — none. No petitioner has filed anything, so there is no repeated-petitioner pattern (no General Plastic / Fintiv issues). The patent owner (Conexant → Lakestar Semi → Conexant Systems, LLC → Synaptics Incorporated) has never litigated this patent at the PTAB, and — per the prior litigation search — has no known district court enforcement history on it either. No defensive aggregator (Unified Patents, OpenSky, etc.) appears in any chain for this patent. The absence of PTAB activity, combined with the patent's quiet post-issuance history (only routine security-interest assignments) and its 2019 expiry, strongly suggests this patent was never commercially asserted in a way that would attract an IPR.

Recommended next steps

• There is no PTAB activity — say so plainly, and use the expiry, not the IPR absence, as the primary lever. If a demand letter or complaint cites US 6,278,387, the threshold response is: the patent expired on 2019-09-28; it cannot be enforced against any accused activity after that date (no injunction, no ongoing royalties), and any alleged pre-expiry damages require pleading and proof of infringement within the § 286 six-year window, which closed in 2019. A post-expiry demand letter asserting this patent is close to sanction-bait on its face.
• If you want belt-and-suspenders verification before relying on "no PTAB proceedings": run the patent number directly in the USPTO's PTAB E2E / PTAB Trial Search (https://ptab.uspto.gov) and the PTAB Bulk Data Search tool (patent-number quick lookup), and pull the USPTO Patent Center file wrapper for US09/407,465 to confirm there is no inter partes reexamination (the pre-AIA analog) lurking either. My searches found none, but a direct E2E lookup by patent number is the authoritative confirmation.
• Preserve the invalidity arsenal. Because there is no estoppel, every § 102/§ 103 ground remains open. If you litigate, the strongest art will be the pre-1999 time-scale-modification and MPEG sub-band literature (e.g., the SOLA algorithm publications from the 1980s–1990s; ISO-IEC/ITC1 SC29/WG11 DIS 11172, April 1992 — which the specification itself cites as the basis for the 32-sub-band filter bank; and the "Kmin–Kmax per-sub-band search range" table, which is a strong § 103 target given the specification admits the ranges are merely derived from each sub-band's lowest-frequency period).
• One honest caveat: web and ODP searches returning nothing is not absolute negative proof. There is no PTAB proceeding on file per the USPTO Open Data Portal, and no web-visible proceeding exists, but if you need a formal certification for litigation, obtain a PTAB E2E docket search and a PRD (Patent Review Data) download filtered on this patent number, and have counsel confirm the absence before representing it to a court.

I attempted to query the USPTO Assignment Center and run supplementary web searches for correspondent-of-record data (reel/frame-specific queries for 010427/0863, 043786/0267, 038777/0885, etc.). All searches returned empty results and the tool session limit was reached before the Assignment Center could be accessed. The full assignment record below therefore comes from the USPTO-derived legal-events data embedded in the authoritative Google Patents record supplied with the task (reel/frame, conveyance type, parties, and effective dates are all present there). Correspondent-of-record attorney names could not be retrieved and are NOT inferred or fabricated below; signal 3 is accordingly marked unclear.

Inventors

• Maksim Y. Rayskiy — sole named inventor. Employer at filing: Conexant Systems, Inc. (per the 1999-11-26 recorded assignment of rights from Rayskiy to Conexant, reel 010427/0863, effective 1999-09-29 — one day after the filing date).

Unusual-pattern check: not present. There is a single inventor (not a team), and there is no record of the inventor departing Conexant within 12 months of filing — he assigned his rights to Conexant at filing, which is the ordinary course. Nothing here suggests a portfolio fire-sale precursor.

Original assignee

• Conexant Systems, Inc. (issuing entity; the Nov 1999 inventor assignment names "CONEXANT SYSTEMS, INC." as assignee). Google Patents normalizes the original assignee to the current name "Conexant Systems LLC," which is the product of the 2017 name change — the entity of record at issue was Conexant Systems, Inc.
• Line of business: Fabless semiconductor company — PC audio codecs (including AC'97-class audio), modem/DSP chipsets, and later IoT and voice-processing silicon. An MPEG-compliant audio codec with variable-rate playback (the claimed subject matter) is squarely within Conexant's product space, so it is plausible but not confirmed that a shipped product embodied the claims; I have no verified product teardown or datasheet tie to the specific SOLA-per-sub-band claims.
• Current status: Not an independent operating company. The record shows Conexant Systems, Inc. → Conexant Systems, LLC (name change, effective 2017-03-20) → Synaptics Incorporated (assignment effective 2017-09-01, recorded 2017-09-07), matching Synaptics' public acquisition of Conexant's audio/voice business in 2017. Synaptics (NASDAQ: SYNA) is a public operating company.

Assignment timeline

All 12 recorded conveyances below are taken from the USPTO-derived legal-events table in the supplied patent record. Execution dates and correspondent names are not separately retrievable from the supplied data (the "effective date" shown is the date the instrument states; the recording date is the USPTO-verified date). Correspondent fields are marked not retrieved rather than guessed.

• 1999-09-29 (effective) / recorded 1999-11-26 — Reel 010427 / Frame 0863

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Maksim Y. Rayskiy (inventor)
  • Assignee: Conexant Systems, Inc.
  • Correspondent: not retrieved
  • Context: Original inventor-to-company assignment at filing; standard.

• 1998-12-21 (effective) / recorded 2000-01-03 — Reel 010450 / Frame 0899

  • Conveyance: Security Interest
  • Assignor: Conexant Systems, Inc.
  • Assignee: Credit Suisse First Boston
  • Correspondent: not retrieved
  • Context: Lender collateral grant (debt financing), not an ownership transfer.

• 2001-10-18 (effective) / recorded 2001-11-05 — Reel 012252 / Frame 0865

  • Conveyance: Release of Security Interest
  • Assignor: Credit Suisse First Boston
  • Assignee: Conexant Systems, Inc.; Brooktree Corporation; Brooktree Worldwide Sales Corporation; Conexant Systems Worldwide, Inc.
  • Correspondent: not retrieved
  • Context: Release of the 2000 lien; routine.

• 2006-11-13 (effective) / recorded 2006-11-22 — Reel 018711 / Frame 0818

  • Conveyance: Security Agreement
  • Assignor: Conexant Systems, Inc.
  • Assignee: Bank of New York Trust Company, N.A.
  • Correspondent: not retrieved
  • Context: New lender collateral grant (BNY as indenture trustee); routine debt financing.

• 2010-01-28 (effective) / recorded 2010-03-01 — Reel 023998 / Frame 0838

  • Conveyance: Release by Secured Party
  • Assignor: The Bank of New York Mellon Trust Company, N.A. (formerly The Bank of New York Trust Company, N.A.)
  • Assignee: Conexant Systems, Inc.
  • Correspondent: not retrieved
  • Context: Release of the 2006 security agreement.

• 2010-03-10 (effective) / recorded 2010-03-11 — Reel 024066 / Frame 0075

  • Conveyance: Security Agreement
  • Assignor: Conexant Systems, Inc.; Conexant Systems Worldwide, Inc.; Conexant, Inc.; Brooktree Broadband Holding, Inc.; and others
  • Assignee: The Bank of New York, Mellon Trust Company, N.A.
  • Correspondent: not retrieved
  • Context: Re-filed consolidated lender security agreement across the Conexant group.

• 2014-03-10 (effective) / recorded 2016-05-06 — Reel 038631 / Frame 0452

  • Conveyance: Release by Secured Party
  • Assignor: The Bank of New York Mellon Trust Company, N.A.
  • Assignee: Conexant Systems, Inc.; Conexant Systems Worldwide, Inc.; Conexant, Inc.; Brooktree Broadband Holding, Inc.
  • Correspondent: not retrieved
  • Context: Release of the 2010 security agreement, recorded ~2 years late.

• 2013-07-12 (effective) / recorded 2016-05-20 — Reel 038777 / Frame 0885

  • Conveyance: Change of Name
  • Assignor: Conexant Systems, Inc.
  • Assignee: Lakestar Semi Inc.
  • Correspondent: not retrieved
  • Context: Corporate rename of Conexant Systems, Inc. to Lakestar Semi Inc., recorded nearly three years late (2016).

• 2013-07-12 (effective) / recorded 2016-05-23 — Reel 038803 / Frame 0693

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Lakestar Semi Inc.
  • Assignee: Conexant Systems, Inc.
  • Correspondent: not retrieved
  • Context: Immediate reassignment back from Lakestar Semi Inc. to Conexant Systems, Inc., same effective date as the rename — a corrective/cleanup pair (name-change loop reversed), not a transfer to a distinct entity.

• 2017-03-20 (effective) / recorded 2017-06-26 — Reel 042986 / Frame 0613

  • Conveyance: Change of Name
  • Assignor: Conexant Systems, Inc.
  • Assignee: Conexant Systems, LLC
  • Correspondent: not retrieved
  • Context: Corporate rename to LLC form ahead of the Synaptics acquisition.

• 2017-09-01 (effective) / recorded 2017-09-07 — Reel 043786 / Frame 0267

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Conexant Systems, LLC
  • Assignee: Synaptics Incorporated
  • Correspondent: not retrieved
  • Context: Operating-business acquisition — Synaptics buys Conexant's audio/voice business; patent moves to the acquiring public operating company.

• 2017-09-27 (effective) / recorded 2017-09-27 — Reel 044037 / Frame 0896

  • Conveyance: Security Interest
  • Assignor: Synaptics Incorporated
  • Assignee: Wells Fargo Bank, National Association
  • Correspondent: not retrieved
  • Context: Synaptics grants lender security interest (credit-facility collateral); not an ownership transfer.

Net result: ownership never left the operating-company sphere. The patent issued to Conexant, was encumbered/released by lenders, passed through two name-change/reverse recordings (Lakestar loop), and ended at Synaptics.

Timeline diagram

timeline
    title Ownership of US 6278387
    1999 : Filed by Conexant
         : Inventor assignment recorded
    2000 : Credit Suisse security interest
    2001 : Patent issued
         : Credit Suisse lien released
    2006 : BNY Mellon security agreement
    2010 : BNY Mellon lien released
         : New BNY Mellon agreement
    2016 : Lakestar Semi name change
         : Reassignment to Conexant
    2017 : Renamed Conexant Systems LLC
         : Acquired by Synaptics
         : Wells Fargo security interest

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only LLC/Inc. entities appearing as assignees are corporate renames of the operating company itself: Lakestar Semi Inc. (reel 038777/0885) and Conexant Systems, LLC (reel 042986/0613). No "IP / Patents / Licensing / Holdings / Ventures" LLC, no registered-agent-only address, and no single-purpose Delaware/Texas licensing LLC appears anywhere in the chain. Ownership stayed inside Conexant/Synaptics the entire time.

2. Known asserter in the chain — not present. Assignees across all 12 conveyances are: Conexant Systems, Inc.; Credit Suisse First Boston (lender); Bank of New York Trust Company, N.A. / BNY Mellon Trust (lender); Lakestar Semi Inc. (Conexant rename); Conexant Systems, LLC (rename); Synaptics Incorporated (public operating company, NASDAQ: SYNA); Wells Fargo Bank (lender). None match Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp., or Spangenberg entities.

3. Repeat correspondent across the chain — unclear. Correspondent-of-record names could not be retrieved in this session (USPTO Assignment Center unreachable via available tools; reel/frame web searches returned no results). Per the constraint against fabrication, this signal is left unmarked rather than guessed. It is the one gap in this analysis.

4. Cascading transfers — not present. The only back-to-back recordings — reels 038777/0885 (Change of Name to Lakestar Semi Inc.) and 038803/0693 (Assignment back to Conexant), both effective 2013-07-12 and recorded within three days in May 2016 — are a same-day name-change-and-reverse corrective pair, not chained transfers through distinct LLCs with shifting ownership.

5. Pre-litigation transfer — not present. No litigation involving this patent was found in the prior analysis (no district-court or CAFC docket entries). The 2017 transfer to Synaptics (reel 043786/0267, effective 2017-09-01) corresponds to a public business acquisition, not a standing-arrangement filing timed to a complaint.

6. Bankruptcy fire-sale — not present in the assignment record. No trustee, receiver, or § 363 sale conveyance appears. The security agreements (reels 010450/0899, 018711/0818, 024066/0075) and releases (reels 012252/0865, 023998/0838, 038631/0452) are ordinary lender collateral filings. Whether Conexant's underlying restructuring involved a bankruptcy proceeding is outside this assignment record and I cannot confirm it either way.

7. Privateering — not present. The final transfer runs to an operating company (Conexant Systems, LLC → Synaptics Incorporated, reel 043786/0267), not from one to a third-party NPE. No SEC or press evidence of a privateering arrangement surfaced.

8. Defensive aggregator (anti-NPE) — not present. The chain terminates at Synaptics Incorporated, a public operating company, not at RPX, AST, LOT Network, Unified Patents, or OIN.

Verdict

Operating-company assertion — with the explicit caveat that this characterization rests on assignee identity rather than observed enforcement.

The chain is fully documented (12 recorded conveyances) and never touches an NPE: ownership runs Conexant Systems, Inc. → (lender liens/releases) → name-change loop → Conexant Systems, LLC → Synaptics Incorporated (reel 043786/0267, recorded 2017-09-07), a public operating company that acquired Conexant's audio business. No shell LLC, known asserter, cascading LLC chain, pre-litigation transfer, bankruptcy-sale conveyance, privateering, or defensive-aggregator terminus appears in the record. However, no litigation involving this patent was found and the patent expired 2019-09-28 (20 years from filing), so there is no evidence Synaptics ever asserted it — the "assertion" label is inferred from Synaptics being an operating company whose product space overlaps the claims, not from any observed suit.

Verify at the USPTO Assignment Center: https://assignmentcenter.uspto.gov/ (search patent number 6278387; indexed copy at https://assignment.uspto.gov/patent/index.html). The one unresolved item is correspondent-of-record attorney data (signal 3), which requires direct Assignment Center access to close.

I'll research the prior-art references cited on the face of US 6,278,387. Let me run targeted searches.

Prior-Art Analysis — U.S. Patent 6,278,387 (US6278387B1)

Note on scope and search integrity. I queried for the exact number 6278387 and its cited references. The authoritative record supplied with the task (Google Patents, generated from USPTO data) lists exactly six (6) references cited by/considered by the examiner ("Citations (6)"). I did not auto-correct any identifier. Where a search result conflicted with or added to the supplied text, I say so. One tool-limit note is flagged below.

The six cited references at a glance

Threshold point that governs everything below: the '387 patent's independent claims are directed to (1) a codec with a rate adjust module enabling variable playback (claim 1), (2) a decoder whose time scaling module "time stretches the input samples within the time domain for each of the first and second frequency sub-bands" (claim 9), and (3) a SOLA-style "best averaging point" frame-overlap-and-average method applied per sub-band (claim 16). The six cited references all predate 1999, but none of them is directed to time-scale modification at all. They are classic "analogous-art / background" citations showing the state of audio encoding, quantization and sample-rate conversion — which is consistent with the specification's own admissions (it cites the MPEG DIS 11172 standard directly, and separately appears to adopt the SOLA frame-sync metric from the published SOLA literature).

1. US4862168A — Beard, "Audio digital/analog encoding and decoding"

• Full citation: U.S. Patent 4,862,168, T.D. Beard, "Audio digital/analog encoding and decoding," issued Aug. 29, 1989; application Ser. No. 07/027,747 filed Mar. 19, 1987. (EP counterpart EP0283167, pub. 1988-09-21.)
• Description. A hybrid/fully-digital encoder–decoder that reduces quantization noise: adaptive high-frequency pre-emphasis, sampling, non-linear low-resolution encoding, and carry-back of the quantization error into the next sample; complementary non-linear decoding, de-emphasis, and anti-aliasing/post filtering. Figures 1–2 show a hybrid encoder/decoder, Fig. 3 a full-digital encoder.
• Anticipation assessment (§ 102). None of claims 1–23 anticipated. Beard discloses an encoder and a decoder (a codec) with memory/processor-implemented digital encoding — conceptually touching claim 1's "encoder… processor… directing the encoding of the first audio signal into a digital code word" and a complementary decoder. But it lacks the "rate adjust module that permits variable playback" — the single element that distinguishes claim 1. It lacks any sub-band decomposition, so claims 6–8, 9, 16 are wholly unsupported, and it discloses nothing about time scaling, SOLA, overlapping frames, or a best averaging point (claims 9–23). At most it is § 103 fodder for the generic "codec with an encoder and a decoder" preamble of claim 1.

2. US4933675A — Beard, "Audio digital/analog encoding and decoding"

• Full citation: U.S. Patent 4,933,675, T.D. Beard, issued Jun. 12, 1990; a divisional of application Ser. No. 07/027,747, filed Mar. 19, 1987 (i.e., same priority date and same specification family as US4862168, but not a § 102(b) "same invention" duplicate — it claims different subject matter; it adds predictor-based hybrid encoder/decoder implementations, Figs. 8–13).
• Description. Same pre-emphasis / non-linear low-resolution encoding / error-carryback / complementary-decode scheme as the '168 patent, extended with a linear predictor feeding forward the predicted value of the next sample in both encoder and decoder.
• Anticipation assessment (§ 102). None of claims 1–23 anticipated, for the same reason as the '168 sibling: no rate-adjust module, no sub-band coding, no time scaling. Note that because the '675 is a divisional of the same 1987 application, these two citations are effectively one family — they add nothing cumulative for § 102 beyond a single reference, and both are silent on the SOLA/time-scaling core.

3. US5451954A — Dolby Laboratories, "Quantization noise suppression for encoder/decoder system"

• Full citation: U.S. Patent 5,451,954, Dolby Laboratories Licensing Corp., "Quantization noise suppression for encoder/decoder system," filed Aug. 4, 1993, issued Sep. 19, 1995.
• Description. A split-band (sub-band/transform) perceptual audio coder exploiting psychoacoustic masking. It expressly defines "split-band coder," "subband," "signal sample block," "subband information" etc.; the decoder "extracts quantized information… dequantizes… and appl[ies] an inverse filter bank" to reconstruct the signal. The German family member's counterpart text (DE 602004010885T2) describes the decoder front-end as Deformatierer (unformatter) → Dequantisierer → Synthesefilterbank (inverse/synthesis filter bank).
• Anticipation assessment (§ 102). This is the closest reference to the front half of independent claim 9, and it maps remarkably well onto three of claim 9's four elements: (a) an input that receives a compressed bit stream carrying multiple sub-bands, (b) an unformatter that unpacks the frame/format structure, and (c) an inverse bit-allocate/dequantizer that recovers per-sub-band samples, plus the output inverse filter bank of claim 10. But the fourth element — the "time scaling module [that] time stretches the input samples within the time domain for each of the first and second frequency sub-bands to enable variable playback" — is entirely absent. Dolby's decoder reconstructs at a fixed rate. Therefore Dolby does not anticipate claim 9 or claim 10, and it says nothing about claims 11–23 (no variable playback, no frames/averaging point). It is, however, strong § 103 background for the decoder-architecture elements of claims 9–10 and for the sub-band/psychoacoustic framework in claims 6–8.

4. US5712635A — Analog Devices, "Digital to analog conversion using nonuniform sample rates"

• Full citation: U.S. Patent 5,712,635, J. Wilson, R.A. Cellini, J.M. Sobol; Analog Devices, Inc.; issued Jan. 27, 1998; a continuation-in-part of Ser. No. 08/120,957, filed Sep. 13, 1993 (PCT/US1994/010269; EP0719478).
• Description. Sigma-delta modulation of the temporal spacing between digital samples for D/A conversion: interpolation by a fixed (or variable) ratio, then decimation controlled by a noise-shaped (sigma-delta-modulated) frequency-select signal that "represents, on average, the data rate of the incoming digital data stream." Claims are to a DAC system with interpolation means, decimation means, a modulator (sigma-delta), and a D/A converter; plus a phase-locked-loop variant.
• Anticipation assessment (§ 102). No claim anticipated. The reference is a sample-rate-conversion / DAC invention: it changes the data rate of a stream, not the playback duration at constant pitch of an encoded audio signal, and it contains no sub-band decomposition, no encoder+decoder codec, no frames, and no overlap/averaging. It does not meet claim 1 (a codec with encoder and decoder plus a rate adjust module — the '635 has a converter, not an encoder/decoder pair), and it does not meet claim 9 or 16. Its only relevance is as background showing that variable-rate digital audio conversion was known — a § 103 consideration if a defendant tried to argue "variably rate" ≡ "variable playback," which the '387 specification refutes by defining rate adjustment as time scaling (SOLA).

5. US5786778A — Analog Devices, "Variable sample-rate DAC/ADC/converter system"

• Full citation: U.S. Patent 5,786,778, R.W. Adams and T.W. Kwan; Analog Devices, Inc.; filed Oct. 5, 1995; issued Jul. 28, 1998.
• Description. A digital oversampling noise-shaping variable sample-rate conversion system: a digitally-controlled oscillator (DCO) at a fixed master clock produces a noise-shaped clock signal having a variable (instantaneously variable) rate whose average equals a multiple of the input sample rate; an interpolator upsamples responsive to that clock; a hold circuit repeats samples; and a sigma-delta noise-shaping circuit processes them. A decimation embodiment is also claimed. Claims 1–18 are converter-system claims.
• Anticipation assessment (§ 102). No claim anticipated. Same category problem as the '635: it is a converter (DAC/ADC/sample-rate) invention, not an encoding/decoding audio codec, and it contains no sub-band filtering, no unformatter/inverse-bit-allocate, and no time scaling. It cannot meet claim 1 (no encoder/decoder pair, no rate adjust module) nor claims 9 or 16. Relevant only as § 103 background for "the concept of variably playing back a digital audio stream was known."

6. US5896099A — Sanyo Electric, "Audio decoder with buffer fullness control"

• Full citation: U.S. Patent 5,896,099, Sanyo Electric Co., Ltd., "Audio decoder with buffer fullness control," priority/filing Jun. 30, 1995, issued Apr. 20, 1999.
• Description (from title/assignee metadata and the patent record; see caveat). An audio decoder that manages buffer fullness — i.e., the occupancy of an input/decoder buffer as compressed audio frames are consumed — a standard concern in frame-based compressed-audio decoders. It is the only cited reference that issued within one year of the '387 filing date, so it is not § 102(b) art; it would be available only under § 102(a)/(e) (U.S. patent granted on an application — filed 1995 — by another before the '387 inventors' date).
• Anticipation assessment (§ 102). No claim anticipated. A buffer-fullness-control decoder relates at most to the general architecture of a compressed-audio decoder (an input interface and buffering), but it does not disclose the unformatter + inverse bit-allocate + per-sub-band time-scaling module combination of claim 9, nor any frame-overlap/averaging of claim 16, nor the codec+rate-adjust combination of claim 1. It supplies no element critical to any independent claim.
• ⚠️ Verification caveat (stated explicitly, not hidden). My search session's tool limit was reached before I could retrieve the full text of US5896099; the description above is grounded on its citation metadata and title only, not on an inspected disclosure. If a definitive § 102 read of the '099 is needed, pull its full text directly (e.g., patents.google.com/patent/US5896099) and confirm whether its buffer control touches rate or continuity of playback.

Consolidated § 102 verdict

No cited reference, alone, anticipates any of the 23 claims. Expressed claim-by-claim for the independent claims:

Dependent claims fall with their independents. No cited reference touches claims 12–15, 18–23 (frames, search range Kmin–Kmax, fade-in/fade-out averaging, compression/expansion, fixed vs. user-selectable frame size), because not one of the six references addresses time-scale modification.

Practical characterization: these six are § 102(b) background/analogous art (five of them) plus one § 102(a)/(e) reference (Sanyo '099), cited to show the state of the art in (i) audio encode/decode, (ii) perceptual sub-band coding and its decoder, and (iii) variable-rate digital audio conversion. The patent's novelty rests squarely on the per-sub-band time-domain SOLA overlap-and-average, which the citations do not reach. The genuine § 102/§ 103 risk for the '387 claims lies outside this list — in the pre-1999 SOLA/time-scale-modification literature (e.g., the Roucos–Wilgus SOLA work and related 1980s–90s TSM papers) and in the MPEG DIS 11172 standard that the specification itself incorporates by reference for the 32-sub-band filter bank.

Two record observations (flagged, not auto-corrected)

1. Claim 22 dependency defect (literal text). Claim 22 reads: "The method according to claim 14 wherein the number of audio samples within an input frame is user-selectable." Claim 14 is a decoder claim about time compression, not the method claim that introduces "input frame." The consistent companion is claim 21 (method of claim 16, frame size "fixed"), making claim 22 read as if it should depend from claim 21/16. I have not corrected it — reported as written, because it is a genuine claim-drafting anomaly in the granted text and it affects how a court would read claim 22's scope.

2. "Cited By" ≠ prior art. The record's "Cited By (24)/(42)" tables (e.g., US6,718,309 SSI; US7,426,221 Cisco; US2005/0137730 Trautmann; US2009/0144064 Sakurai) are forward citations — later documents citing the '387 — and are not § 102 prior art. The only backward citations are the six analyzed above. Consistent with every prior section of this analysis, no litigation, PTAB, or CAFC activity attaches to these references.

Bottom line for the prior-art question: the six references cited on the face of US6278387 are not anticipatory of any claim; they are background art establishing known audio codec, sub-band coding, and variable-rate conversion techniques. The claims' core (per-sub-band time-domain time scaling with a best averaging point) was not disclosed by any of them, and — given the patent's 2019-09-28 expiry — § 102/§ 103 positions on this patent can only ever be litigated as to pre-2019 conduct, with the strongest art coming from outside the cited list.