US 8346801

I'll search for this specific patent number as requested, then summarize based on the authoritative patent text provided.

Summary — US Patent No. 8,346,801 B2

Important scoping note: I searched for the literal number 8346801. Results referencing "the '680 patent" in PTAB/IPR documents (e.g., IPR2019‑00471, patent 7,734,680) and CAFC 2026 decisions (e.g., Receivership Estate of AudienceScience v. Google over 7,747,676 / 7,882,175 / 8,082,298) concern different patents entirely and must not be conflated with 8,346,801. A stray "8346801" hit in a foreign wine document is also unrelated. I found no CAFC 2026 docket, IPR, or litigation for 8,346,801.

Bibliographic data (confirmed consistent across Google Patents, FreePatentsOnline, Justia)

Abstract (as issued)

Embodiments relate to video retrieval by providing a first audio video file, determining a first identifier of a first piece of music in the first audio video file, looking up for the first identifier first meta data in a music database (in which the first meta data are associated with the first identifier), and providing second meta data of a second piece of music from the music database. The second piece of music is included in a second audio video file. A similarity measure is determined by comparing the first and second meta data, and the second audio video file (or an identifier thereof) is provided depending on the similarity measure. Further embodiments relate to a server, a user device, a system, and a computer program product.

Plain-language overview of the independent claims

Claim 1 — Method for video retrieval

1. Provide a first audio-video file (e.g., a video).
2. Determine a first identifier of a first piece of music in that file (e.g., an acoustic fingerprint) — the music may occur at any position in the file.
3. Using that identifier, look up first meta data in a music database, where the meta data are keyed to the identifier and are descriptive of the entire piece of music (not just the matched segment).
4. Provide second meta data for a second piece of music from the same database, where that second song is in a second audio-video file and its meta data likewise describe the entire piece; that song may also appear anywhere in the second file.
5. Using a microprocessor, compute a similarity measure by comparing the first and second meta data.
6. Provide the second video (or its identifier) when the similarity measure meets or exceeds a predetermined threshold.

Claim 15 — Server (apparatus counterpart)
A server with (a) a communication unit to receive a first audio-video file, (b) storage holding a music database, and (c) a data processor configured to perform the claim‑1 steps — determine the first identifier, look up the first and second meta data (each descriptive of an entire piece), compute the similarity measure, and output the second video/identifier when the threshold is met — with the communication unit further providing that video or identifier.

Claim 16 — Method (temporal-position variant)
Same steps as claim 1, but at least one of the first/second meta data describes a first and second temporal position of the first and second pieces of music within their respective files, and/or a temporal position and length of a found music segment within the first or second piece.

Claim 17 — Method (temporal-order variant)
Same steps as claim 1, but at least one of the first/second meta data describes a temporal order of pieces of music within the first and second audio-video files, respectively.

(Claims 2–14 depend from claim 1 and add: genre/mood/title/artist descriptors; BPM and other low‑level descriptors; collaborative‑filtering meta data; a confidence measure; user‑profile/collaborative filtering; temporal position; temporal order and length of pieces; web‑page/design comparison; fingerprint as identifier; independently‑derived "clean" meta data; and a similarity‑matrix database.)

Observation on claim scope

The granted claims include narrowing language not present in the originally filed description text — notably "wherein the first piece of music can appear at any position within the first audio video file," meta data "descriptive of an entire piece of music identified based on the first piece of music," and the threshold recitation "when the similarity measure meets or exceeds a predetermined threshold." The detailed description and the pre‑grant disclosure instead speak generally of providing the second file "depending on the similarity measure." This suggests amendment during prosecution (consistent with the EP‑side "Office Action issued Apr. 27, 2012" noted in the file). Uncertainty: I do not have the prosecution history / claim‑amendment record itself here, so I cannot authoritatively confirm why those terms were added.

Prior art of record (patent citations, 18 total; non‑patent citations, 6)

Notable references: US 7,356,830 (Philips); US 7,870,574 (Iceberg Industries); US 2009/0228423 (Hicken); US 6,993,532 (Microsoft); US 7,777,125 (Microsoft); US 7,885,657 (Shazam); US 8,073,854 (The Echo Nest); US 8,208,643 (Zhang). Non‑patent literature includes Jasinschi et al., "Automatic TV Program Genre Classification Based on Audio Patterns" (IEEE 2001) and Venugopal et al., "Audio Scene Analysis and Scene Change Detection in the MPEG Compressed Domain."

Notes on uncertainty / verification limits

• I did not independently query USPTO PatentCenter/PAIR or the CAFC docket system directly; my confirmation comes from Google Patents, FreePatentsOnline, and Justia pages surfaced by search, which are consistent with each other and with the full patent text supplied.
• No CAFC 2026 activity for US 8,346,801 was found. Any appearance of a "680 patent" in litigation/PTAB materials should be treated as an unrelated patent number.
• Dates for priority/expiration are the sources' characterizations, not legal conclusions.

I'll search multiple litigation databases for US patent 8346801.

Let me search patent litigation databases more specifically.

Litigation Search Results for U.S. Patent No. 8346801

Bottom line

I found no litigation involving U.S. Patent No. 8346801. Despite searching patent-litigation aggregators (Unified Patents litigation caselist, Docket Alarm, Justia federal district court dockets, RPX/insight dockets, and general web queries), no district court case, Federal Circuit appeal, ITC investigation, or PTAB proceeding naming this specific patent as an asserted or challenged patent appeared in the results.

Given the strict instruction to interpret identifiers literally and not to substitute similar numbers, I am reporting this as "no known litigation identified" rather than importing results for other patents (e.g., the Ameritox/Millennium '680 patent, the Antor '961 patent, or Korean/Ukrainian records bearing the number "8346801" in non-patent contexts, none of which concern this patent).

What the patent is (for context)

• Patent: US 8346801 B2 — "Context based video finder"
• Application: US 12/717,415; filed 2010-03-04
• Priority: 2009-03-24 (EP 09004184.9 / EP 2234024)
• Granted: 2013-01-01
• Original/current assignee: Sony Corporation
• Inventors: Wilhelm Hagg, Jana Eggink, Thomas Kemp, Franck Giron, Javier Alonso Garcia
• Legal status (per Google Patents): Expired – Fee Related; adjusted expiration 2030-08-26
• Family members: EP2234024B1 (not-in-force) and CN101847158A

Source: https://patents.google.com/patent/US8346801/en

Important distinction — citations are not litigation

The Google Patents page for US 8346801 lists both "Patent Citations" (18) and "Cited By" (10) entries. These are prior-art citations and forward citations, not lawsuits. For example:

• Backward citations (prior art): US 7,356,830 (Philips); US 7,870,574 (Iceberg Industries); US 7,885,657 (Shazam); US 8,073,854 (The Echo Nest); US 7,776,125 (Microsoft); among others.
• Forward citations / family citing: US 9,639,871; US 9,875,304; US 10,061,476; US 10,225,328; US 10,242,097; US 10,623,480; US 11,271,993; US 11,609,948 (all Aperture Investments) — these are patents that cite '801, not litigations.
• Non-patent citations (6): academic papers (e.g., Ahmad et al., "Audio-based Queries for Video Retrieval over Java Enabled Mobile Devices," Proc. SPIE-IS&T, 2006; Jasinschi et al., "Automatic TV Program Genre Classification Based on Audio Patterns," IEEE 2001), plus a European Office Action dated Apr. 27, 2012 in EP 09 004 184.9.

None of these is a litigation record.

Caveats / limitations

1. Search coverage. The litigation databases queried (Unified Patents caselist, Docket Alarm, Justia, RPX-linked dockets) returned no hits. That is strong but not conclusive evidence of no litigation — none of these is an exhaustive PACER search.
2. By the numbers. This is a relatively narrow, single-assignee (Sony) patent that went Expired – Fee Related, which is consistent with it never having been asserted in an infringement campaign. It does not appear to have been transferred to an NPE.
3. No PTAB activity found. No IPR, PGR, or CBM petition against US 8346801 surfaced in the search results.
4. Recommendation for a definitive answer. To confirm the negative with certainty, a paid PACER / Lex Machina / Docket Navigator "patent number" query (specifically for U.S. Patent No. 8,346,801) and an ITC EDIS check should be run, since free aggregators can lag or omit low-profile cases.

If you require a verified negative suitable for a legal opinion or clearance analysis, I'd recommend confirming via Lex Machina or Docket Navigator's "Patents–Litigation" dataset keyed exactly to patent number 8,346,801. I cannot represent the negative as exhaustive based on the searches performed here.

I will not fabricate case numbers, parties, dates, or outcomes, and I have deliberately excluded records that merely contain the string "8346801" in unrelated contexts.

I'll verify the structured ODP result against public PTAB/CourtListener sources before finalizing.

PTAB / AIA Trial Landscape — US 8,346,801 B2 ("Context based video finder")

Verification note on sources: The authoritative structured block ("PTAB proceedings on file," USPTO ODP) returns zero AIA trial proceedings for US 8,346,801. I ran four independent web searches (patent number + IPR/PGR/CBM; petitioner-focused queries; family/EPO counterpart queries; Unified Patents queries) and surfaced no petition, institution decision, FWD, or CAFC appeal naming this patent. I did not invent any proceeding numbers. Because I could not query PTAB E2E or CourtListener's API directly with the tools available, I flag residual uncertainty as low but non-zero — a very recently filed, not-yet-indexed petition cannot be affirmatively excluded.

---

Proceedings overview

Total AIA trial proceedings on file: 0. No IPR, PGR, or CBM has ever been filed against US 8,346,801; there are consequently no claims invalidated, none sustained in an FWD, no settlements, and no institution denials — the categories are simply empty. Bottom line for a defendant: the patent is untested at the PTAB, not hardened by it. The absence of proceedings is not a sign of strength-through-survival; it is a sign that no accused infringer has ever found it worth $50K+ in filing fees to attack. For a defendant today that cuts both ways — you face no § 315(e)(2) estoppel from anyone, and the full universe of § 102/§ 103 art remains available, but you also cannot point to any government adjudication to leverage a stay or a § 101 squeeze.

---

Proceedings

None to report. There are no proceedings to rank by impact. I am expressly declining to populate the per-proceeding template (type / filed / panel / grounds / institution / FWD / settlement / appeal) because doing so would require fabricating a docket. As a practitioner, treat an empty section here as the finding itself.

---

Strategic summary

Claim-status map. All 17 claims (1–17) stand unadjudicated and untouched by any post-grant proceeding. There is no cluster of canceled claims, no substitute claims, and no certificate of correction or reexamination narrowing the scope. Claim 1 (method), claim 15 (server), and claims 16–17 (further method claims, added/amended during prosecution around the "can appear at any position" and "temporal position/temporal order" limitations) are all live on their face. The claims as issued already carry meaningful narrowing relative to the 2010 published application: the "wherein the first piece of music can appear at any position within the first audio video file" limitation and the "descriptive of an entire piece of music identified based on the first piece of music" limitation were evidently drafted to distinguish art that maps metadata to a whole video rather than to a music segment located anywhere in the timeline. That is defensible drafting — but it has never been tested by a Board panel.

Estoppel landscape. Because no FWD has ever issued, 35 U.S.C. § 315(e)(2) estoppel is a blank slate. No petitioner, real party in interest, or privy is barred from any § 102/§ 103 ground, and no party has generated IPR claim-construction positions that could be used as prosecution-history admissions against the patent owner in a district court. Practically: a defendant retains full freedom to run anticipation and obviousness on the music-fingerprint-and-similarity-metadata architecture (Shazam-era acoustic fingerprinting, Gracenote-style metadata matching, Echo Nest similarity vectors, the Microsoft "inferring similarities between media objects" line, and the Rovi/Comcast art cited in the family — e.g. US 7,356,830 and US 7,870,574 in the patent's own citation list). The only timing constraint worth diarizing is § 315(b): any IPR petition must be filed within one year of service of a complaint alleging infringement of this patent. There is no post-issuance bar otherwise — issuance was 2013-01-01 and the IPR window opened 2013-09-16.

Pattern signals. Three flags, none alarming: (1) No serial petitioner and no defensive aggregator. Unified Patents, RPX, and similar entities have no proceeding here, and Unified's public campaign pages (e.g., its SEP Video Codec Zone filings) do not reference this patent. Its low-level descriptor/content-matching subject matter falls outside the codec-focused zones Unified typically funds. (2) No parallel litigation surfaced. A targeted search for "8,346,801" in litigation contexts returned nothing; the patent appears never to have been asserted in district court or at the ITC, which explains the zero IPR count — IPRs almost always trail infringement suits. (3) Status is commercial disengagement, not invalidation. Google Patents records the patent as Expired – Fee Related with an adjusted expiration of 2030-08-26, i.e., maintenance fees were not paid and the term was cut short rather than litigated to death. The EP counterpart EP2234024B1 was granted 2012-10-03 (with an EPO Office Action dated 2012-04-27 of record) and is marked not_active / Not-in-force; the CN family member CN101847158A shows as active / Pending in the structured data — meaning whatever residual enforcement appetite exists in this family sits in China, not at the PTAB.

Do not confuse this patent with US 8,046,801. A trivially confusable number — 8,046,801 is Rovi Guides' "Interactive Television Program Guide with Remote Access," which was heavily litigated and went through Comcast v. Rovi Guides, IPR2017-01143 (and a consolidated set: IPR2017-00950, -00951, -00952, -01048, -01049, -01050, -01065, -01066), where the Board held claims 1–54 unpatentable under § 103(a). If a docket sheet, a demand letter, or an AI-generated summary tells you this patent has been invalidated at the PTAB, it is almost certainly describing 8,046,801, not 8,346,801. Verify the digit string character-by-character before relying on it.

---

Recommended next steps

• Because no proceedings exist, there is no FWD to link or quote. I will not manufacture one. The accurate statement is: US 8,346,801 has no PTAB record. For a defendant, that means there is no administrative disposition to cite in a motion to stay, no cancellation certificate to brandish, and no petitioner estoppel to exploit.
• Priority one — confirm the patent's operative status before doing anything else. Google Patents reports Expired – Fee Related, adjusted expiration 2030-08-26. If maintenance fees were in fact not paid and the lapse is unremedied, the patent may be unenforceable for the relevant damages period regardless of claim validity, and an infringement demand premised on it is worth challenging on that basis first. Pull the USPTO Patent Center maintenance-fee history and any § 41(c) petition to accept late payment to confirm — I could not access Patent Center with the tools available, and this is the single highest-leverage fact in the file.
• Diarize the § 315(b) one-year clock the moment you are served. No estoppel attaches from prior proceedings, so your IPR options are wide open, but they evaporate 12 months after service. Given the empty PTAB history, an institution-stage win here would be the first-ever construction of the "can appear at any position" and "descriptive of an entire piece of music" limitations.
• Attack surface to mine for an IPR or § 282 defense: the 2009-priority acoustic-fingerprint-and-metadata-similarity space was crowded. The patent's own IDS lists US 7,356,830, US 7,870,574, US 7,685,132, US 7,777,125, US 8,073,854 (Echo Nest — "Determining the similarity of music using cultural and acoustic information"), and US 8,208,643. Third-party art (Shazam-lineage fingerprinting, Gracenote metadata, last.fm/audioscrobbler collaborative-filtering statistics) is a plausible § 103 wellspring that has never been run against these claims.
• Standing caveat on this memo. My proceeding count is grounded in the ODP structured block plus four searches, not a direct PTAB E2E query. Before filing anything that asserts "no IPRs exist," re-run PTAB E2E and CourtListener against both 8346801 and 8,346,801, and search the assignee string "Sony" alongside the patent title. If a petition is pending and unindexed, it will surface there first.

I'll research the assignment/ownership record for US 8,346,801 before reconstructing the chain.

Let me search specifically for assignment records tied to application 12/717,415 and for any post-issuance transfers.

Ownership & Assignment Reconstruction — US 8,346,801 B2 ("Context based video finder")

Verification note up front. I could not directly query assignmentcenter.uspto.gov / assignment.uspto.gov in this session — the Assignment Center is not reachable via the tools available here, and targeted web queries for a patent-number assignment extract returned no indexed record for this patent's reel/frame. What follows is therefore built from (a) the authoritative Google Patents legal-events block supplied in the patent text (which does contain an assignment event), (b) corroborating bibliographic sources (FreePatentsOnline, Justia), and (c) inventor/assignee cross-referencing against sibling Sony patents. Reel/frame numbers and the correspondent of record are flagged as NOT RETRIEVED rather than guessed. Where I could not verify, I say so.

---

Inventors

All five named inventors, with residence data cross-referenced from the same inventor group's sibling Sony filings:

Basis for the employer attribution: the identical five-person group (Hagg, Eggink, Kemp, Giron, Alonso Garcia) appears as inventors on US 9,491,256 B2 ("Method and device for personalizing a multimedia application," assignee Sony Corporation, Tokyo) and, with Taufan Zimmer, on the Sony Stuttgart mood-model family cited in IEEE patent-citation records (US 7,868,239 B2, assignee Sony Corporation; assignee on related filings listed as Sony Deutschland GmbH). The Kemp/Hagg/Eggink trio is explicitly captioned "c/o Stuttgart Technology Center" in the Sony Deutschland mood-detection family. This is strong circumstantial support that all five were Sony researchers in Stuttgart, not independent inventors.

Unusual-pattern check — no anomaly found. I found no evidence that any inventor departed Sony within 12 months of the 2010-03-04 filing, and no evidence of a portfolio fire-sale following inventor attrition:

• The same five inventors continued filing together for Sony for years afterward — US 9,491,256 issued 2016-11-08, and the Kemp/Hagg/Eggink group appears on a Sony Deutschland content-recommendation filing issued 2016-06-21 (US 9,372,860, applicant Sony Deutschland GmbH).
• This is the inverse of the "inventors leave, then the portfolio is dumped" pattern. Continuity of the inventor group and of the assignee across a 6–7 year window is a strong indicator of a stable corporate R&D program.

Confidence: high on the employer (Sony Stuttgart); the specific corporate entity on the assignment (Sony Corporation vs. Sony Deutschland GmbH) is taken from the recorded assignment, below, not from residence.

---

Original assignee

---

Assignment timeline

The Assignment Center record, as surfaced through the patent's legal-events block, contains exactly one assignment. There are no post-issuance assignments. Reel/frame and correspondent were not retrievable with the tools available and are not fabricated below.

• Executed: not confirmed in this session (the pre-grant assignment was necessarily executed on or before the recording date; the US filing was 2010-03-04) / recorded 2010-06-04 — Reel NOT RETRIEVED / Frame NOT RETRIEVED
  • Conveyance: Assignment of Assignors' Interest (inventor-to-company assignment; recorded as "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: Franck Giron; Javier Alonso Garcia; Jana Eggink; Wilhelm Hagg; Thomas Kemp (all five named inventors, jointly)
  • Assignee: Sony Corporation, Tokyo, Japan
  • Correspondent: NOT RETRIEVED. I could not obtain the attorney/agent or firm of record for this recording. One unverified lead surfaced in the search: Sony's long-standing US trademark/serials attorney Daniel M. Gurfinkel (120 South LaSalle Street, Suite 1400, Chicago IL) appears as "Attorney of Record" in Sony's TTAB/TSDR trademark files. This is a different legal matter and a different property type — I am explicitly NOT asserting it as the correspondent for this patent's assignment. It is a lead to check, not a finding.
  • Context: Original employment/prosecution assignment. This is the routine instrument by which the five inventors transferred their rights to their employer as a condition of employment. It is not a sale, not a reorg, and not an NPE transfer.
  • Recurrence flag: N/A — only one event exists in the chain, so there is no second link on which a correspondent could recur. No repeat-correspondent pattern can form with a single record.

No further recorded assignments exist. No transfer to any IP holding company, licensing entity, NPE, defensive aggregator, or acquirer is recorded for the life of this patent. No security agreement, merger, change-of-name, license, or release is recorded.

---

Timeline diagram

timeline
    title Ownership of US 8346801
    2009 : EP priority application filed
    2010 : US application filed by inventors
         : Inventors assign to Sony Corporation
    2013 : US patent granted
         : Recorded owner unchanged since 2010

---

NPE / troll-pattern signals

Additional ownership-relevant observation (not an NPE signal, but material to value): Google Patents records the patent as Expired – Fee Related with an adjusted expiration of 2030-08-26. If that status is accurate, the owner of record (Sony Corporation) remains the owner, but the patent is unenforceable for the lapsed interval unless revived under 35 U.S.C. § 41(c). I could not verify the maintenance-fee history or any revival petition in Patent Center with the tools available — this is an asserted status, not a verified one, and should be confirmed before being relied on. The EP counterpart EP2234024B1 is also marked not‑in‑force; the CN family member CN101847158A shows as active/pending. None of that implicates a US assignment chain.

---

Verdict

Insufficient data — but note the finding is substantive, not merely a data gap.

The Assignment Center record for US 8,346,801 consists of exactly one recorded assignment, the 2010-06-04 inventor-to-Sony Corporation conveyance, with no post-issuance transfer of any kind. That is the literal trigger for the "only the original assignment" prong of this verdict category. Substantively, it means none of the eight NPE signals is present: no shell-entity transfer, no known asserter, no cascading LLC chain, no pre-litigation assignment, no bankruptcy sale, no privateering, and no defensive-aggregator endpoint. The patent has remained with a solvent operating company since filing, was never asserted, and lapsed for fee non-payment — the profile of commercial disengagement by an operating owner, not of a troll chain.

Two things a reviewer should verify before relying on this: (1) pull the actual reel/frame and correspondent from the Assignment Center for application 12/717,415 / patent 8,346,801, since I could not reach the Assignment Center to capture them and have deliberately not invented them; and (2) confirm the "Expired – Fee Related" status and any § 41(c) revival petition in USPTO Patent Center, which is the single highest-leverage ownership/enforceability fact in the file.

Verify at: USPTO Assignment Center / Patent Assignment Search — search by patent number 8346801 or application 12/717,415.

I'll research the prior art citations for US 8346801 to provide a grounded analysis. Let me start with several searches in parallel.

Let me gather details on the remaining key references.

Prior-Art Analysis — US 8,346,801 B2 ("Context Based Video Finder")

1. Identification of the patent under analysis

Source: https://patents.google.com/patent/US8346801/en ; https://www.freepatentsonline.com/[8346801](/patent/8346801).html

Independent claims at issue. Claim 1 (method), claim 15 (server, mirrors claim 1), claim 16 (method adding temporal-position/length metadata), and claim 17 (method adding temporal-order metadata). Claim 1 requires, in substance: (a) providing a first audio-video (AV) file; (b) determining a first identifier of a first piece of music in it, where the music "can appear at any position"; (c) looking up first meta data in a music database keyed to that identifier, the meta data describing an entire piece of music; (d) providing second meta data of a second piece of music included in a second AV file; (e) determining, using a microprocessor, a similarity measure by comparing the first and second meta data; and (f) providing the second AV file or its identifier when the similarity measure meets or exceeds a predetermined threshold.

Legal framework applied. Anticipation under 35 U.S.C. § 102 requires that a single reference disclose every limitation, arranged as in the claim (MPEP 2131). Where a reference discloses most but not all limitations, or where the claim reads on a combination of references, the correct ground is § 103 (obviousness), not § 102. The analysis below therefore distinguishes genuine potential § 102 anticipation from § 103 art, and flags my confidence level. Note also that the examiner citations for this patent are marked with an asterisk in the Google Patents citation table (12 of the 18); the other six were third-party submissions.

---

2. Patent citations (18) — full citations, dates, descriptions

The 18 references below are listed exactly as cited on the face/record of US 8,346,801. Dates are taken from the citation table and corroborated by search where possible.

2.1 US 7,356,830 B1 — Method and apparatus for linking a video segment to another segment or information source

• Assignee / inventor: Koninklijke Philips Electronics N.V.; Nevenka Dimitrova
• Priority date: 1999-07-08/09 · Granted: 2008-04-08 · Expired 2019-07-08
• URLs: https://patents.google.com/patent/US7356830 ; https://patentimages.storage.googleapis.com/c4/e8/5c/c51e1608ffff6f/US7356830.pdf
• Description: Determines associations between an object/entity/feature of a video segment and a second information source containing the same feature, then retrieves and displays material from the second source in conjunction with, or in place of, the original segment. Associations are stored in a database (or in the transport stream). Features may be video (face/scene/event/object) or audio — including "a music signature extraction," speaker identification, transcript — or textual. Associations can be computed by visual, scene, event, textual, spoken-word, voice or audio similarity using feature vectors and standard distance measures (L1, L2, chi-square); "the invention can extract feature vectors and compute similarity using standard distance measures."
• § 102 relevance: Strongest single prior-art teaching of retrieving a second video based on audio (music) content similarity, and of associating a first video segment with other video segments via a stored database. Potentially anticipatory for the concept underlying claims 10 and 11 (retrieval of a web page / comparing a page's design is not taught, so claim 11 is not anticipated). For claim 1 it is close but does not clearly teach looking up meta data describing an entire piece of music from a music database keyed by a music identifier. Best characterized as primary § 103 art against claims 1, 10, 15.

2.2 US 7,870,574 B2 — Method and apparatus for automatically recognizing input audio and/or video streams

• Assignee: Iceberg Industries, LLC
• Date: 1999-09-21 · Granted: 2011-01-11
• URL: cited at https://patents.google.com/patent/US8346801/en
• Description: Automatic recognition of input audio and/or video streams (broadcast monitoring / content recognition). I could not retrieve the full text in this session; the description is inferred from the title and assignee's known media-monitoring portfolio and is flagged as medium confidence.
• § 102 relevance: General background art for recognizing audio/video streams (fingerprinting/identification). Relevant to the identifier-determining step of claim 1 and to claim 12 (fingerprint), but unlikely alone to anticipate claim 1.

2.3 US 2009/0228423 A1 — Music Recommendation System and Method

• Inventor: Hicken, Wendell T.
• Priority date: 2000-04-21 · Published: 2009-09-10
• URL: https://patents.google.com/patent/US20090228423A1/en (listing in https://patents.google.com/patent/US8346801/en)
• Description: Music recommendation system/method (this publication is in the family of the long-pending "Music Recommendation System and Method" line). Describes recommending music to a user based on characteristics/attributes and user preference. Full text not retrieved in this session — medium confidence.
• § 102 relevance: General § 103 art for the similarity-based recommendation concept and, in combination, for claims 6 (user profile / collaborative filtering) and 1.

2.4 US 2007/0038612 A1 — System and method for indexing, searching, identifying, and editing multimedia files

• Inventor: Sanghoon Sull
• Priority date: 2000-07-24 · Published: 2007-02-15
• URL: https://patents.google.com/patent/US20070038612A1/en
• Description: Indexing/searching/identifying/editing multimedia files (audio-based identification of multimedia content; content-based multimedia indexing). Full text not retrieved — medium confidence.
• § 102 relevance: § 103 art addressing identification and searching of multimedia (including video) files by content. Relevant to claim 1's retrieval-by-content step and claim 12.

2.5 US 6,993,532 B1 — Auto playlist generator

• Assignee: Microsoft Corporation
• Priority date: 2001-05-30 · Granted: 2006-01-31
• URL: https://patents.google.com/patent/[US6993532B1](/patent/US6993532B1)/en
• Description: Automatically generates a playlist of media items, e.g., based on a seed/similarity parameters and metadata attributes of media items. Full text not retrieved — medium-high confidence on the general teaching.
• § 102 relevance: § 103 art for the metadata-similarity → item-selection step. Notably relevant to claim 14 (similarity database / similarity matrix) and claim 1's comparison step, though in the music-domain rather than video-retrieval domain.

2.6 US 2003/0105743 A1 — Use of database queries for manipulation of media content

• Assignee: Sonicblue, Inc.
• Priority date: 2001-11-16 · Published: 2003-06-05
• URL: https://patents.google.com/patent/US20030105743A1/en
• Description: Uses database queries over media metadata to organize/manipulate media content. Full text not retrieved — medium confidence.
• § 102 relevance: § 103 art for querying a music/media database by metadata (claim 1's look-up step).

2.7 US 7,756,915 B2 — Automatic digital music library builder

• Assignee: Catch Media (Ben-Yaacov et al.)
• Priority date: 2003-01-02 (US 10/336,443) · Granted: 2010-07-13
• URLs: https://patents.google.com/patent/[US7756915B2](/patent/US7756915B2)/en ; https://FreePatentsOnline.com/[7756915](/patent/7756915).html
• Description: Builds a digital music library from broadcasted A/V (e.g., cable/satellite music channels). An audio parser marks the start/end of a song within the digital audio; an OCR-based meta-data extractor reads song metadata from the video frame; a database manager associates the identified meta-data with the marked song. Filtering by metadata (genre, title, artist, album, length) is supported; metadata verified against an internal CD database such as Gracenote.
• § 102 relevance: Strong teaching of (i) detecting song boundaries / temporal position within an A/V stream and (ii) associating metadata with a song located in a video. Potentially relevant to claims 7, 16, 17 (temporal position/order/length), and as § 103 art for claim 1's metadata-association concept. It does not teach similarity-based retrieval of a second video, so it does not anticipate claim 1.

2.8 US 7,949,050 B2 — Method and system for semantically segmenting scenes of a video sequence

• Assignee: British Telecommunications Public Limited Company
• Priority date: 2004-03-23 · Granted: 2011-05-24
• URL: https://patents.google.com/patent/[US7949050B2](/patent/US7949050B2)/en
• Description: Semantically segments scenes of a video sequence (video shot/scene boundary detection and semantic labeling). Full text not retrieved — medium confidence.
• § 102 relevance: Background/§ 103 art regarding video segmentation and content analysis. Peripheral to the claims (no music-metadata similarity retrieval). Likely cited to show video-segmentation/analysis environment.

2.9 US 2006/0080356 A1 — System and method for inferring similarities between media objects

• Assignee: Microsoft Corporation
• Priority date: 2004-10-13 · Published: 2006-04-13
• URL: https://patents.google.com/patent/US20060080356A1/en
• Description: Infers similarity between media objects (media-object feature/metadata comparison to derive similarity). Full text not retrieved — medium-high confidence on the general teaching.
• § 102 relevance: § 103 art for the core "determine a similarity measure by comparing meta data" step of claim 1.

2.10 US 2006/0092281 A1 — System and method for automatically customizing a buffered media stream

• Assignee: Microsoft Corporation
• Priority date: 2004-11-02 · Published: 2006-05-04
• URL: https://patents.google.com/patent/US20060092281A1/en
• Description: Automatic customization of a buffered media stream (insertion/substitution of content based on characteristics). Full text not retrieved — medium confidence.
• § 102 relevance: § 103 art relating to media-stream customization; peripheral to the retrieval-by-music-metadata claim structure.

2.11 US 7,777,125 B2 — Constructing a table of music similarity vectors from a music similarity graph

• Assignee: Microsoft Corporation
• Priority date: 2004-11-19 · Granted: 2010-08-17
• URL: https://patents.google.com/patent/[US7777125B2](/patent/US7777125B2)/en
• Description: Constructs a table of music similarity vectors from a music similarity graph — i.e., a precomputed similarity data structure over music items. Full text not retrieved — medium-high confidence.
• § 102 relevance: Notably relevant to claim 14 ("similarity database comprising a similarity matrix of pieces of music") and as § 103 art for the similarity-determination step of claim 1.

2.12 US 7,685,132 B2 — Automatic meta-data sharing of existing media through social networking

• Assignee: Mog, Inc.
• Priority date: 2006-03-15 · Granted: 2010-03-23
• URL: https://patents.google.com/patent/[US7685132B2](/patent/US7685132B2)/en
• Description: Automatic sharing of metadata for existing media (identifying media and associating/sharing metadata, including via social/collaborative services). Full text not retrieved — medium confidence.
• § 102 relevance: § 103 art for metadata association and for claim 4 / claim 6 collaborative-filtering aspects.

2.13 US 2008/0059453 A1 — System and method for enhancing the result of a query

• Inventor: Raphael Laderman
• Priority date: 2006-08-29 · Published: 2008-03-06
• URL: https://patents.google.com/patent/US20080059453A1/en
• Description: Enhances query results (ranking/filtering of search results based on additional signals). Full text not retrieved — medium confidence.
• § 102 relevance: Peripheral § 103 art for result ranking/threshold filtering (claim 1's threshold step).

2.14 US 7,881,657 B2 — Method for high-throughput identification of distributed broadcast content

• Assignee: Shazam Entertainment, Ltd. (Wang, Wong, Symons)
• Priority date: 2006-10-03 (provisional 60/848,941) · Granted: 2011-02-01
• URLs: https://patents.google.com/patent/[US7881657B2](/patent/US7881657B2)/en ; https://www.freepatentsonline.com/y2013/0247082.html
• Description: A client sends a content-identification query (a sample or a fingerprint of broadcast content) to a server; the server computationally identifies the content, returns metadata (song title, artist, album), and stores the result in a metadata cache keyed to broadcast source with a validity window; subsequent queries for the same source within the window are answered from the cache without recomputation. Fingerprints ("landmarks," spectral-slice fingerprints, hashed features) are disclosed in detail; the reference notes video files may likewise be identified by fingerprints.
• § 102 relevance: Strong teaching of identifier-from-audio (fingerprint) → database look-up → return metadata. Directly relevant to claim 12 (identifier is a fingerprint) and the identifier/meta-data look-up of claim 1. It does not teach comparing first/second music metadata to score similar videos, so it is § 103 (not § 102) art against claim 1.

2.15 US 2008/0109404 A1 — Location dependent music search

• Assignee: Sony Ericsson Mobile Communications AB
• Priority date: 2006-11-03 · Published: 2008-05-08
• URL: https://patents.google.com/patent/US20080109404A1/en
• Description: Music search that depends on user location. Full text not retrieved — medium confidence.
• § 102 relevance: § 103 art for metadata-driven music search; demonstrates the assignee's own field. Peripheral to claims.

2.16 US 8,073,854 B2 — Determining the similarity of music using cultural and acoustic information

• Assignee: The Echo Nest Corporation (Brian Whitman; Tristan Jehan)
• Priority date: 2007-04-10 (provisional 60/922,647) · Granted: 2011-12-06
• URLs: https://patents.google.com/patent/[US8073854B2](/patent/US8073854B2)/en ; https://patentimages.storage.googleapis.com/1d/8f/72/7c27515a59ed9a/US8073854.pdf
• Description: Determines similarity between music collections by comparing metadata (acoustic metadata — feature vectors, acoustic fingerprints, tempo/rhythm/beats and spectral melody/pitch/harmony/timbre — and cultural metadata — description vectors of listener-derived terms, genres, moods). Multiple comparison algorithms are combined; comparison scores may be normalized, compared to a predetermined threshold and converted to binary (Yes/No) form; a similarity server returns similar tracks with similarity scores; user-specific basis vectors reflect personal taste.
• § 102 relevance: This is the closest prior-art teaching of the comparison-of-music-metadata-to-derive-a-similarity-measure-with-a-threshold heart of claim 1, and of the acoustic-fingerprint-as-identifier aspect (claim 12). However, it operates on music tracks, not on audio video files, and does not provide "the second audio video file or an identifier thereof." It is therefore best treated as the leading § 103 reference against claims 1, 3, 6, 12, 14, and 15 (combined with a video-retrieval reference such as US 7,356,830), rather than as a clean § 102 anticipation of claim 1. The full disclosure is available and verified.

2.17 US 8,208,643 B2 — Generating music thumbnails and identifying related song structure

• Inventor: Tong Zhang
• Priority date: 2007-06-29 · Granted: 2012-06-26
• URL: https://patents.google.com/patent/[US8208643B2](/patent/US8208643B2)/en
• Description: Generates music thumbnails and identifies related song structure (segment/thumbnail detection and structural similarity). Full text not retrieved in this session — medium confidence.
• § 102 relevance: § 103 art for music-segment/structure analysis; potentially relevant to claims 7/16/17 (segment position/length within a piece of music) as a secondary reference.

2.18 US 2011/0035373 A1 — Robust video retrieval utilizing audio and video data

• Assignee: Pixel Forensics, Inc.
• Priority date / filing: 2009-08-10 · Published: 2011-02-10
• URL: https://patents.google.com/patent/US20110035373A1/en (listing in https://patents.google.com/patent/US8346801/en)
• Description: Robust video retrieval that combines audio and video data features to match/retrieve videos. Full text not retrieved in this session — medium confidence, but the title alone is directly on point.
• § 102 relevance — important date caveat: This reference's effective date (2009-08-10) is later than the 2009-03-24 EP priority date but earlier than the 2010-03-04 US filing date of US 8,346,801. It could only be § 102(e)-type art if US 8,346,801 were not entitled to the EP priority date for the relevant subject matter (pre-AIA § 102(e) dates run from the reference's US filing date, and a § 119 foreign priority does not itself create a § 102(e) date, but the applicant's own foreign priority can defeat the reference). This is a genuine priority/date question rather than a clear anticipation, so I flag it as uncertain and note it is the reference whose timing most warrants scrutiny.

---

3. Non-patent citations (6)

The non-patent literature is dominated by audio-content-analysis / audio-based video-retrieval papers. The Ahmad et al. paper (#1) is the single most relevant because it explicitly addresses audio-based queries for video retrieval, i.e., the very premise of claim 1.

---

4. Claim-by-claim mapping to the most relevant prior art

Below, "PA" = potential anticipation (§ 102, single reference); otherwise the entry is § 103 art. Confidence is stated because full text for several references could not be retrieved in this session.

---

5. Assessment: most relevant prior art

Ranked by relevance to the novelty of claims 1/15:

1. US 8,073,854 B2 (The Echo Nest) — discloses the metadata-comparison → similarity measure → threshold → provide similar content pipeline, including acoustic fingerprints and cultural/descriptive metadata. Missing only the "audio video file" framing. Primary § 103 reference.
2. US 7,356,830 B1 (Philips / Dimitrova) — discloses retrieving a second video segment based on content similarity, expressly including music-signature audio features and stored associations. Primary § 103 reference; closest to the video-retrieval architecture.
3. US 7,881,657 B2 (Shazam) — fingerprint → server-side identification → metadata return, with caching and validity windows. Strong art for claim 12 and for the identifier step of claim 1.
4. US 7,756,915 B2 (Catch Media) — song boundary / temporal-position / length detection and metadata association within broadcast A/V. Key art for claims 7, 9, 16, 17.
5. US 7,777,125 B2 (Microsoft) — precomputed music similarity-vector table. Key art for claim 14.
6. Ahmad et al. (SPIE, 2006) — audio-based queries for video retrieval. Most on-point non-patent literature for the claim 1 premise.

6. Important caveats

• Anticipation vs. obviousness. No single cited reference that I was able to verify appears to disclose every limitation of claim 1, particularly the combination of (i) a music identifier within an audio-video file, (ii) look-up of entire-piece meta data from a music database, and (iii) provision of a second audio-video file upon threshold satisfaction. The realistic challenge posture is § 103, using the references above in combination; genuine § 102 anticipation is most plausible only for narrow dependent claims (e.g., claim 12 by US 7,881,657, or claims 7/9 by US 7,756,915) and for claim 10.
• Unverified references. I retrieved full text for US 7,356,830, US 7,756,915, US 7,881,657, US 8,073,854, US 7,777,125, and the Ahmad citation context. For US 7,870,574, US 2009/0228423, US 2007/0038612, US 6,993,532, US 2003/0105743, US 7,949,050, US 2006/0080356, US 2006/0092281, US 7,685,132, US 2008/0059453, US 2008/0109404, US 8,208,643, and US 2011/0035373, I relied on the citation table plus title/assignee-level knowledge; my characterizations for those are marked medium confidence and should be confirmed against full text before being relied upon.
• Date/priority nuance. US 2011/0035373 (filed 2009-08-10) post-dates the 2009-03-24 EP priority date but pre-dates the 2010-03-04 US filing date; whether it is prior art depends on the priority-date entitlement question discussed in § 2.18.
• Procedural note. The § 102/§ 103 designations I give are analytical, not legal conclusions and not a substitute for a validity opinion; the examiner's labels in the patent file should be consulted for the grounds actually applied. The Google Patents citation table marks 12 of the 18 patent citations with an asterisk (indicating examiner citation); six were third-party submissions.

Primary sources cited: https://patents.google.com/patent/US8346801/en · https://www.freepatentsonline.com/8346801.html · https://patents.google.com/patent/US7356830 · https://patents.google.com/patent/US8073854B2/en · https://patents.google.com/patent/US7881657B2/en · https://patents.google.com/patent/US7756915B2/en · https://patents.google.com/patent/US7777125B2/en · https://patents.google.com/patent/US7685132B2/en · https://patents.google.com/patent/US7949050B2/en · https://patents.google.com/patent/US6993532B1/en · https://patents.google.com/patent/US8208643B2/en · https://patents.google.com/patent/US20110035373A1/en

I'll analyze the obviousness question using the prior art cited on the patent page. Let me first verify the substance of the key references so my citations are accurate.

Now let me verify the remaining key references — the video-retrieval and music-similarity references.

Let me verify the remaining two references I'll rely on heavily for the secondary claims.

Obviousness Analysis — US 8,346,801 B2 ("Context Based Video Finder")

1. Framework and legal standard

Governing law. The application was filed 2010‑03‑04 with a foreign priority date of 2009‑03‑24 (EP09004184). Both dates precede 16 March 2013, so pre‑AIA 35 U.S.C. § 103(a) applies. The controlling standard is KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), under the Graham factors: scope/content of the prior art, differences between the prior art and the claims, level of ordinary skill, and secondary considerations.

Level of ordinary skill in the art (POSITA). A person with a bachelor's degree in electrical engineering, computer science, or equivalent, plus 2–4 years of experience in content‑based multimedia retrieval, audio signal processing, and information retrieval/recommendation systems. All references below are from the same or analogous fields (content identification, music similarity, video retrieval).

Note on dates (checked literally against the citation table): Seventeen of the eighteen U.S. patent citations and all five non‑patent citations pre‑date the 2009‑03‑24 priority date and are available as § 102/§ 103 art. One cited reference does not qualify: US 2011/0035373 A1 (Pixel Forensics) was filed 2009‑08‑10 and published 2011‑02‑10 — both after the priority date — so it is not prior art and cannot be used in a § 103 combination, notwithstanding that the examiner listed it. I flag it here and exclude it from the combinations below.

---

2. Claim 1 — element mapping to the primary combination

Primary combination: Shazam (US 7,881,657 B2) + The Echo Nest (US 8,073,854 B2) + Ahmad et al. (SPIE 2006).

• US 7,881,657 B2 (Shazam Entertainment, priority 2006‑10‑03) — fingerprint acquisition from an audio sample, identification of the track against a database, and return of track metadata (title, artist, album), plus relative‑time‑offset and track length. https://patents.google.com/patent/[US7881657B2](/patent/US7881657B2)/en
• US 8,073,854 B2 (The Echo Nest, priority 2007‑04‑10) — "Determining the similarity of music using cultural and acoustic information"; compares metadata (acoustic/cultural/explicit) of a target against a reference, combines comparison scores, and can compare the result to a predetermined threshold to yield a binary Yes/No. https://patents.google.com/patent/[US8073854B2](/patent/US8073854B2)/en
• Ahmad, Cheikh, Kiranyaz & Gabbouj, "Audio‑based queries for video retrieval over Java enabled mobile devices," Proc. SPIE 6074, Multimedia on Mobile Devices II (17 Jan 2006) — a client‑server framework that retrieves video clips using audio content as the query, computes a (dis)similarity distance between the query and each database item, and ranks/returns matches. https://researchportal.tuni.fi/en/publications/audio-based-queries-for-video-retrieval-over-java-enabled-mobile-

Result: every element of claim 1 is disclosed or rendered obvious by the combination, including the two limitations that appear to have been the allowance hook — "descriptive of an entire piece of music identified based on the first piece of music" and "can appear at any position." Shazam's full‑track identification + RTO/length disclosure and Echo Nest's whole‑track metadata comparison meet these directly.

---

3. Why a POSITA would have combined these references

1. Same field / analogous art. All three are content‑based media retrieval systems (fingerprinting, music similarity, audio‑based video query). KSR permits combination of references "designed to address the same problem."

2. Express teaching, suggestion, or motivation in Shazam itself. US 7,881,657 states that although the examples concern audio streams, "the invention is not so limited, but rather may be applied to a wide variety of broadcast content, including video, television, internet streaming… video files may be identified using similar techniques for identifying audio files… and checking a database of videos to identify videos that have similar markings." This is a textbook TSM (and, under KSR, an explicit design incentive) to apply Shazam's fingerprint→metadata engine to video.

3. Known problem, known solution — the "clean metadata" rationale. The '801 specification itself concedes the problem: browsing large video collections is "time consuming and cumbersome," and metadata derived from a user's video is sparse/unreliable, so canonical ("clean") metadata from an independent music database is used. Echo Nest and Microsoft US 2006/0080356 both recognize exactly this reliability problem — Microsoft notes that "when relying on the metadata alone, if that data is either entered incorrectly… any similarity analysis based on that metadata will also be inaccurate," and expressly proposes combining audio fingerprinting with metadata. That is a direct motivation to bolt Shazam's fingerprinting onto Echo Nest's metadata‑similarity engine.

4. Efficiency/cost incentive. Both Echo Nest and Microsoft tout metadata comparison as computationally cheaper than audio matching (Microsoft: "without the need to perform computationally expensive audio matching analyses"), and Ahmad expressly motivates audio‑based retrieval for large databases and mobile clients. A POSITA optimizing a large video library would predictably adopt metadata‑based similarity after fingerprint identification.

5. KSR predictability rationale. Combining (i) a known fingerprint‑based music identifier, (ii) a known metadata‑similarity engine, and (iii) a known audio‑query video retrieval front‑end amounts to "arranging old elements" to achieve a predictable result — identification of a song in a video, retrieval of comparable songs' metadata, and return of matching videos. No new and unexpected mechanism is required; the claim is a straightforward pipelining of known functions.

---

4. Alternative and cumulative combinations

Combination A (Microsoft‑centric), for claim 1: Shazam US 7,881,657 + US 2006/0080356 A1 (Microsoft, "inferring similarities between media objects") + US 7,777,125 (Microsoft, "constructing a table of music similarity vectors from a music similarity graph") + Ahmad.

• US 2006/0080356 explicitly uses audio fingerprints and/or metadata to identify media objects in a stream and infers similarity from their positions, for "media object filing, retrieval, classification, playlist construction." Motivation: it expressly addresses the metadata‑reliability shortfall of prior art and states it "allows otherwise new or unknown songs… to be included in the similarity analysis."
• US 7,777,125 supplies the similarity‑vector table storage (relevant to claim 14).

Combination B (video‑linking centric): Ahmad + US 7,356,830 (Philips) + Echo Nest + Shazam. Philips ("linking a video segment to another segment or information source") supplies the "provide the second audio video file or an identifier thereof" step in a video‑linking context.

Combination C (music‑recommendation centric): Shazam + US 2009/0228423 (Hicken, Music Recommendation System and Method) + Echo Nest. Hicken supplies recommendation/collaborative‑filtering metadata (claims 4 and 6).

I did not verify the internal text of US 7,356,830, US 7,777,125, US 2009/0228423, US 6,993,532, US 7,685,132, or US 8,208,643 by search; my descriptions of those are based on the titles/assignees as retrieved from the patent page's citation table plus general knowledge, and should be treated as lower‑confidence than the verified references (Shazam, Echo Nest, Microsoft 2006/0080356, Pixel Forensics, Ahmad).

---

5. Dependent claims 2–14

6. Claims 15, 16, 17

• Claim 15 (server: communication unit + storage + data processor). Same substance as claim 1, plus conventional server hardware. Shazam's "sample analyzer 108" (request server, recognition server, database 112), Echo Nest's "similarity server 140" with "metadata library 190," and Ahmad's client‑server (Tomcat servlets) each disclose the claimed server architecture. Obvious for the reasons in § 3.
• Claim 16 adds temporal position/length of the piece (and/or of a found music segment) — Shazam's RTO + length; US 2006/0080356's relative positions.
• Claim 17 adds temporal order of pieces — US 2006/0080356's ordered list.

---

7. Weaknesses, caveats, and honest uncertainty

1. Claim 11 is the strongest non‑obviousness candidate. Comparing the visual design of two web pages (logos, background color, fonts, patterns) to determine music similarity is disconnected from the audio‑fingerprint/metadata core. None of the cited references discloses it, and I cannot identify a clear motivation to combine web‑page aesthetics with music‑metadata similarity. A validity challenge should treat claim 11 as likely surviving on its own, while every other claim falls.

2. The "entire piece of music" / "any position" language was evidently the allowance hook. Examination history is not fully in the record here, but the claim language reads as an amendment distinguishing art that matched only short segments. Shazam's disclosure that a several‑second sample yields the full track identity, length, and relative time offset is the reference most directly calculated to defeat that distinction — this is the key point of attack.

3. Pixel Forensics US 2011/0035373 A1 should be removed from any § 103 combination. Filed 2009‑08‑10, after the 2009‑03‑24 priority date; it is not § 102/§ 103 art. (It remains useful only as background evidence of the state of the art of video matching by audio+video feature cross‑correlation, not as invalidating art.)

4. US 2006/0080356 A1 was abandoned (per the prosecution record, failure to respond to an office action) — irrelevant to its status as prior art; it is a printed publication as of 2006‑04‑13 and remains fully available under § 102(b).

5. Secondary considerations. Nothing in the file evidences unexpected results, long‑felt need, industry praise, or copying. The patent is "Expired – Fee Related," so no litigation/validity record is available to illuminate secondary considerations. Absent such evidence, the Graham secondary‑consideration factor is neutral.

6. Claim construction note. "Mo meta data… descriptive of an entire piece of music identified based on the first piece of music" should be construed as the specification's "clean metadata" — metadata about the whole song looked up from a canonical database using an identifier derived from a possibly short excerpt (spec, ¶ describing "clean" meta data). This construction is what makes Shazam's full‑track identification + Echo Nest's whole‑track metadata comparison dispositive.

---

8. Conclusion

• Claims 1–10 and 12–17 would likely have been obvious under pre‑AIA § 103(a) over Shazam US 7,881,657 + Echo Nest US 8,073,854 + Ahmad et al., optionally in view of Microsoft US 2006/0080356 and US 7,777,125 (for similarity storage/vector tables and temporal‑order content), US 7,356,830 (video‑to‑video linking), US 2009/0228423 / US 6,993,532 / US 7,685,132 (collaborative filtering and user profiles), Jasinschi / Pfeiffer / Venugopal / Rasheed (genre, BPM, and low‑level audio descriptors), and US 8,208,643 (piece length/structure).
• The motivation to combine is strong and multi‑sourced: identical field; Shazam's own express extension of audio‑fingerprinting to video; the recognized metadata‑reliability problem with an express fingerprint+metadata combination in Microsoft; and the computational‑efficiency and scalability incentives recited in Echo Nest, Microsoft, and Ahmad — all consistent with the KSR "predictable result/known elements" rationale.
• Claim 11 (web‑page‑design similarity) is the one claim for which I find no adequate prior‑art support in the cited references and which appears to lack a clear motivation to combine; it is the most defensible claim on validity.
• Caveat: US 2011/0035373 (Pixel Forensics) is not valid prior art (post‑priority filing) and must be excluded from any § 103 combination, despite its citation by the examiner.

Sources: https://patents.google.com/patent/US8346801/en (authoritative claim text and citation table); https://patents.google.com/patent/US7881657B2/en; https://patents.google.com/patent/US8073854B2/en; https://patents.google.com/patent/US20060080356A1; https://patents.google.com/patent/US20110035373; https://researchportal.tuni.fi/en/publications/audio-based-queries-for-video-retrieval-over-java-enabled-mobile-