Speech and Audio Coding Patents: Who Leads, Where Gaps Are 2026
- 44.1% concentration. The five leading assignees together hold 3,560 of the 8,069 records in scope — a field with a genuine crowded core, not just a long tail.
- Filing peaked in 2017. 213 records that year against 19 in the most recent (partial) year; the comparable complete-year trend shows filings down 25% from 2021 (134) to 2024 (100).
- G10L dominates, H04L trails far behind. 75.8% of records sit in speech/audio analysis and synthesis (G10L); the next largest class, digital transmission (H04L), covers only 17.3%.
Filing growth compares 2021 (134 records) with 2024 (100) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 8,069 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This dataset tracks 8,069 published patent records filed between 2015 and mid-2026 that combine core speech and audio codec concepts — speech codecs, neural audio codecs, packet loss concealment — with supporting techniques such as bitrate scalability, algorithmic delay control, linear predictive coding, jitter buffering, noise suppression and vector quantization. The scope spans classic telephony-era codec work through to more recent packet-loss and quality-estimation approaches, giving a view across both legacy telecom infrastructure and newer audio-quality tooling.
Filing activity is concentrated by both technology class and assignee. Most records sit inside speech and audio analysis/synthesis (G10L), with secondary presence in digital transmission, telephonic communication and coding/code conversion classes — a pattern consistent with codec technology that gets embedded into broader communication stacks rather than filed as a standalone category.
Let an AI agent run this analysis on your own technology
Pick a task. Every answer cites the patents behind it.
Filing trends and technology composition
Two views of the same 8,069-record scope: filing activity over time, and how records distribute across IPC subclasses.
Filings rose to a 2017 peak, then eased
Annual filings reached 213 in 2017, the high point across the tracked period. Using the last complete comparison available, 2021 (134) to 2024 (100) shows a 25% decline. Years after 2024 are still filling in as publications catch up with filing dates, so the drop-off visible at the very end of the chart understates true recent activity rather than confirming a further decline.
G10L carries three-quarters of the field
Speech & audio analysis/synthesis (G10L) appears in 75.8% of the 8,069 records — by far the largest single class. Digital information transmission (H04L, 17.3%), telephonic communication (H04M, 8.9%) and coding/code conversion (H03M, 7.9%) follow well behind, reflecting how codec inventions typically get filed alongside the transport or device context they run inside. Because records can carry multiple IPC codes, these shares add up to more than 100%.
Shares are the percentage of the 8,069 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Speech and Audio Coding with Eureka
This page is one run against one query. Ask Eureka your own question about speech and audio coding and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative and most-cited filings
Quality estimation model for packet loss concealment (US20240127848A1)
The filing describes training and using a quality estimation model that scores audio output from packet loss concealment models: degraded audio is run through one or more concealment models, the enhanced output is paired with quality labels, and a quality estimation model is trained to predict signal quality directly from the enhanced audio and those labels.Filed by Microsoft Technology Licensing, LLC; published 2024-04-18.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | WO1998057436A2 | Source coding enhancement using spectral-band replication | 1,163 |
| 2 | US20020061012A1 | Cable modem with voice processing capability | 637 |
| 3 | US6775267B1 | Method for billing IP broadband subscribers | 575 |
| 4 | US6373817B1 | Chase me system | 561 |
| 5 | US6680972B1 | Source coding enhancement using spectral-band replication | 540 |
| 6 | US6418147B1 | Multiple vocoder mobile satellite telephone system | 540 |
| 7 | US6081780A | TTS and prosody based authoring system | 519 |
| 8 | US6252952B1 | Personal user network (closed user network) PUN/CUN | 516 |
| 9 | US6246672B1 | Singlecast interactive radio system | 490 |
| 10 | US6765931B1 | Gateway with voice | 481 |
Citation counts reflect influence inside this searched corpus and skew toward older filings; they are not a measure of current commercial relevance.
Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the numbers mean for a filing decision
Three signals worth weighing before committing claim strategy in this space.
The core is genuinely crowded
Five assignees combined hold 3,560 of the 8,069 records in scope, and the top ten extend that to 57.9%. Filing directly into core G10L speech-codec claim territory means competing against dense prior art from established telecom and audio equipment holders rather than open ground.
Activity has eased from its 2017 peak
The peak year was 2017 at 213 filings. The most recent complete comparison, 2021 to 2024, shows a 25% decline. Recent-year totals shown in the chart are still incomplete because publication lags filing by roughly 18 months, so this should be read as a genuine multi-year easing rather than a cliff.
Claims cluster around the codec core
Three in four records touch G10L speech/audio analysis and synthesis, while transmission-layer classes like H04L (17.3%) and H04M (8.9%) appear far less often. This suggests transport-integration angles are comparatively less saturated than core encode/decode claim language.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to speech and audio coding, with the prior art for and against each one.
Who is filing, and where activity is cooling
The leading names in this scope are legacy telecom equipment makers and codec specialists; recent-year momentum has slowed across nearly all of them.
A single dominant filer
The top-ranked assignee holds 1,254 records, well ahead of fifth place at 294 and tenth place at 204 — a steep drop-off after the leader that flattens into a longer tail across the rest of the ranking.
Research-institute partnerships persist
Co-assignee activity is modest — only 10 pairs recorded — but the strongest pairing, a research institute working alongside a major codec specialist, points to continued joint filing between academic labs and industry on core codec algorithms.
Cooling activity even among top filers
Recent-year filing counts are down sharply across nearly every major assignee tracked, with several logging zero filings in the latest year. This is consistent with the broader 2021–2024 decline rather than a company-specific pullback.
| Assignee | Recent year | YoY |
|---|---|---|
| Telefonaktiebolaget LM Ericsson (publ) | 3 | -40% |
| Fraunhofer-Gesellschaft zur Forderung der angewandten Forschung e.V. | 2 | -87% |
| Dolby International AB | 1 | -80% |
| Qualcomm Incorporated | 0 | -100% |
| Nokia Technologies Oy | 0 | — |
| Broadcom Corporation | 0 | — |
| Panasonic Corporation | 0 | — |
| VoiceAge Corporation | 0 | — |
Where to take this analysis
The dataset points to a crowded codec core and a cooling filing trend, but the detail worth acting on sits below the aggregate numbers.
Check freedom-to-operate against the leaders
With 44.1% of records held by five assignees, a claim chart against their core G10L filings is the first check before investing in codec-level claims.
Run a claim comparison in EurekaTrack the under-claimed branches
Packet-loss quality scoring and adaptive jitter-buffer control show thinner filing density than core encode/decode claims — worth a dedicated search before committing to a filing strategy.
Explore white space in EurekaCommon questions about speech and audio coding patents
One assignee leads clearly with 1,254 records in this dataset, well ahead of the rest of the ranked field — fifth place sits at 294 and tenth place at 204. The leading names are a mix of legacy telecom equipment makers and dedicated codec specialists, reflecting decades of continuous filing around core speech-coding algorithms. Because the ranking counts records rather than a curated top list, it reflects filing volume, not necessarily current market share or technical superiority.
Filing peaked in 2017 at 213 records and has eased since. The clearest comparable trend, using only complete filing years, is a 25% decline from 134 filings in 2021 to 100 in 2024. Years after 2024 currently show lower counts in the chart, but that reflects publication lag of roughly 18 months rather than an actual collapse in filing activity, so treat the very latest years as incomplete rather than conclusive.
The dominant class is G10L, speech and audio analysis/synthesis, appearing in 75.8% of the 8,069 records in scope. Secondary classes include H04L (digital information transmission, 17.3%), H04M (telephonic communication, 8.9%) and H03M (coding and code conversion, 7.9%), showing that codec inventions are usually filed alongside the transmission or device context they operate within rather than as isolated algorithm claims. A single record commonly carries several of these classes at once.
Filing density thins out in areas adjacent to the crowded codec core, such as neural packet-loss quality scoring, cross-codec bitrate renegotiation, adaptive jitter-buffer delay control and low-delay vector quantization for real-time streams. These sit next to heavily filed core encode/decode claim territory but have not attracted the same concentration of filings from the leading assignees. A freedom-to-operate search focused on these narrower branches is more likely to surface open claim space than a search centred on core G10L codec algorithms.
It is concentrated at the top: five assignees hold 44.1% of all 8,069 records in scope, and the top ten hold 57.9%. That leaves a long tail across the remaining ranked companies, each holding a comparatively small share. For a new entrant, this means the core claim space around established codec algorithms is likely to require careful freedom-to-operate analysis against a small number of large portfolios.
Research Speech and Audio Coding in depth with Eureka
Go past this page: query the whole speech and audio coding corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.