Hubbry Logo
Codec 2Codec 2Main
Open search
Codec 2
Community hub
Codec 2
logo
7 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Codec 2
Codec 2
Codec 2
View on Wikipedia
from Wikipedia
Codec 2
DeveloperDavid Grant Rowe
Initial releaseAugust 25, 2010 (2010-08-25)
Stable release
1.2.0 / June 24, 2023; 2 years ago (2023-06-24)
Repositorygithub.com/drowe67/codec2
Written inC99
PlatformCross-platform
TypeAudio codec
LicenseGNU LGPL, v2.1
Websitewww.rowetel.com?page_id=452

Codec 2 is a low-bitrate speech audio codec (speech coding) that is patent free and open source.[1] Codec 2 compresses speech using sinusoidal coding, a method specialized for human speech. Bit rates of 3200 to 450 bit/s have been successfully created. Codec 2 was designed to be used for amateur radio and other high compression voice applications.

Overview

[edit]

The codec was developed by David Grant Rowe, with support and cooperation of other researchers (e.g., Jean-Marc Valin from Opus).[2]

Codec 2 consists of 3200, 2400, 1600, 1400, 1300, 1200, 700 and 450 bit/s codec modes. It outperforms most other low-bitrate speech codecs. For example, it uses half the bandwidth of Advanced Multi-Band Excitation to encode speech with similar quality.[citation needed] The speech codec uses 16-bit PCM sampled audio, and outputs packed digital bytes. When sent packed digital bytes, it outputs PCM sampled audio. The audio sample rate is fixed at 8 kHz.

The reference implementation is open source and is freely available in a GitHub repository.[3] The source code is released under the terms of version 2.1 of the GNU Lesser General Public License (LGPL).[4] It is programmed in C and current source code requires floating-point arithmetic, although the algorithm itself does not require this. The reference software package also includes a frequency-division multiplex digital voice software modem and a graphical user interface based on WxWidgets. The software is developed on Linux and a port for Microsoft Windows created with Cygwin is offered in addition to an Apple MacOS version.

The codec has been presented in various conferences and has received the 2012 ARRL Technical Innovation Award,[5] and the Linux Australia Conference's Best Presentation Award.[6]

Technology

[edit]

Internally, parametric audio coding algorithms operate on 10 ms PCM frames using a model of the human voice. Each of these audio segments is declared voiced (vowel) or unvoiced (consonant).

Codec 2 uses sinusoidal coding to model speech, which is closely related to that of multi-band excitation codecs. Sinusoidal coding is based on regularities (periodicity) in the pattern of overtone frequencies and layers harmonic sinusoids. Spoken audio is recreated by modelling speech as a sum of harmonically related sine waves with independent amplitudes called Line spectral pairs, or LSP, on top of a determined fundamental frequency of the speaker's voice (pitch). The (quantised) pitch and the amplitude (energy) of the harmonics are encoded, and with the LSP's are exchanged across a channel in a digital format. The LSP coefficients represent the Linear Predictive Coding (LPC) model in the frequency domain, and lend themselves to a robust and efficient quantisation of the LPC parameters.[7]

The digital bytes are in a bit-field format that have been packed together into bytes. These bit fields are also optionally gray coded before being grouped together. The gray coding may be useful if sending raw, but normally an application will just burst the bit fields out. The bit fields make up the various parameters that are stored or exchanged (pitch, energy, voicing Booleans, LSP's, etc.).

For example, Mode 3200, has 20 ms of audio converted to 64 bits. So 64 bits will be output every 20 ms (50 times a second), for a minimum data rate of 3200 bit/s. These 64 bits are sent as 8 bytes to the application, which has to unwrap the bit fields, or send the bytes over a data channel.

Another example is Mode 1300, which is sent 40 ms of audio, and outputs 52 bits every 40 ms (25 times a second), for a minimum rate of 1300 bit/s. These 52 bits are sent as 7 bytes to the application or data channel.

Adoption

[edit]

Codec 2 is currently used in several radios and Software Defined Radio Systems:

Codec 2 has also been integrated into FreeSWITCH and there's a patch available for support in Asterisk.

There was an FM-to-Codec2 digital voice repeater in earth orbit on amateur radio CubeSat LilacSat-1 (call sign ON02CN, QB50 constellation), which was launched and subsequently deployed from the International Space Station in 2017.[13]

History

[edit]

The prominent free software advocate and radio amateur Bruce Perens lobbied for the creation of a free speech codec for operation at less than 5 kbit/s. Since Perens did not have the background himself, he approached Jean-Marc Valin in 2008, who introduced him to lead developer David Grant Rowe, who has worked with Valin on Speex on several occasions. Rowe himself was also a radio amateur (amateur radio call sign VK5DGR) and had experience in creating and using voice codecs and other signal processing algorithms for speech signals. He obtained a PhD in speech coding in the 1990s and was involved in the development of one of the first satellite telephony systems (Mobilesat).

He agreed to the task and announced his decision to work on a format on August 21, 2009. He built on the research and findings from his doctoral thesis.[14][15] The underlying sinusoidal modelling goes back to developments by Robert J. McAulay and Thomas F. Quatieri (MIT Lincoln labs) from the mid-1980s.

In August 2010, David Rowe published version 0.1 alpha.[16] Version 0.2 was released towards the end of 2011, introducing a mode with 1,400 bits/s and significant improvements in quantization.

In January 2012, at linux.conf.au, Jean-Marc Valin helped improve the quantization of line spectral pairs, which Rowe is less familiar with.[17] After several changes to the available bit rate modes in winter and spring 2011/2012, 2,400, 1,400 and 1,200 bit/s modes were available after May of that year.

Codec 2 700C, a new mode with a bit rate of 700 bit/s, was finished in early 2017.[18]

In July 2018 an experimental 450 bit/s mode was demonstrated, which was developed as part of a master thesis at the University of Erlangen-Nuremberg. By clever training of the vector quantization the data rate could be further reduced based on the principle of the 700C mode.[19]

References

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Codec 2

View on Grokipedia
from Grokipedia
Codec 2 is an open-source speech codec designed for communications-quality speech at low bit rates ranging from 700 to 3200 bits per second, primarily targeting bandwidth-constrained applications such as HF and VHF digital radio. Developed by Rowe (VK5DGR) and released under the GNU Lesser General Public License (LGPL), it employs sinusoidal coding techniques to compress speech while maintaining intelligibility and naturalness, outperforming proprietary codecs like MELP at very low bit rates such as 700 bit/s in informal subjective listening tests. The codec's architecture supports real-time encoding and decoding on resource-limited devices, making it suitable for and emergency communications. Originating from Rowe's 1997 PhD thesis on speech coding, development of Codec 2 began in 2009 with an initial focus on 2400 bit/s modes, evolving to include lower rates like 700 and 1300 bit/s through iterative improvements and community contributions. Supported by grants from the Amateur Radio Digital Communications (ARDC), including a $420,000 grant in 2023 to enhance FreeDV integration with commercial radios, the project is hosted on GitHub, where ongoing enhancements incorporate advanced features such as neural network post-filtering for enhanced quality. Its patent-free status and modular design have facilitated integration with modems like OFDM and coherent PSK, enabling robust performance over noisy channels. Codec 2 powers the FreeDV digital voice protocol, which has seen widespread adoption in the community since its 2012 launch, with global nets in regions including , the , and the . Applications extend to software like the FreeDV GUI for Windows, , and macOS, as well as hardware interfaces such as ezDV, supporting activities like monthly worldwide FreeDV days. In controlled evaluations, FreeDV modes using Codec 2 demonstrate speech quality comparable to or better than analog SSB on varied signal-to-noise ratios, underscoring its role in promoting open-source alternatives to proprietary digital voice systems.

Introduction

Overview

Codec 2 is an open-source speech codec designed for low-bitrate digital voice communications, targeting bit rates from 450 to 3200 bit/s to achieve communications-quality speech in bandwidth-constrained environments. It primarily serves applications in , enabling efficient voice transmission over narrow bandwidths in HF and VHF digital modes such as FreeDV. The codec accepts input as 8 kHz sampled 16-bit linear PCM audio and processes it in frames of 10 ms or 20 ms duration, depending on the selected mode. Licensed under the , Codec 2 was developed by David Grant Rowe to provide a patent-free alternative to low-bitrate codecs. Codec 2 has received recognition for its , including the 2012 ARRL Technical Innovation Award for advancing digital voice technology in and the Linux Australia Conference's Best Presentation Award for Rowe's 2012 talk at .

Development Background

Codec 2 was initiated in 2010 by David Grant Rowe, an Australian electrical engineer specializing in and . Rowe earned his PhD in 1997 from the for a on techniques for harmonic sinusoidal coding of speech signals, which laid the groundwork for efficient low-bitrate representation of voiced speech using sinusoidal oscillators with parametric phase modeling. His experience in digital signal processing includes developing speech codecs and modems for open-source projects, such as FreeDV, which integrates Codec 2 with channel modulation techniques for high-frequency (HF) radio transmission. The primary motivation for Codec 2 stemmed from the limitations of existing open-source speech codecs, such as , which were designed for higher bit rates (typically above 2 kbit/s) and struggled to deliver intelligible speech at ultra-low rates suitable for bandwidth-constrained applications over HF channels. Rowe was particularly inspired by , a prominent advocate for and , who in 2008 called for the development of patent-free alternatives to proprietary military-grade codecs like MELP, emphasizing the need for accessible, low-complexity solutions for hobbyist and emergency communications. This push aligned with broader efforts to democratize digital voice technology, avoiding the licensing barriers that restricted adoption in non-commercial settings. Codec 2's foundational research draws heavily from 1980s advancements in sinusoidal speech modeling, pioneered by researchers including Robert J. McAulay and Thomas F. Quatieri, who introduced methods for decomposing speech into harmonic sinusoids to enable low-bitrate coding while preserving perceptual quality. (References to McAulay and Quatieri's 1986 work on sinusoidal transform coding.) Early development benefited from collaboration and support by Jean-Marc Valin, creator of the Speex codec, who provided insights on open-source implementation and integration challenges during initial discussions prompted by Perens. The project's initial goals centered on achieving communications-quality speech—defined as highly intelligible with acceptable distortion—at bit rates below 700 bit/s, while minimizing computational demands to run on resource-limited embedded systems like microcontrollers in radio transceivers.

Technical Specifications

Encoding and Decoding Process

Codec 2 operates on input speech sampled at 8 kHz in PCM format, processing it in frames of 20 ms (160 samples) for higher modes (3200 and 2400 bit/s) or 40 ms (320 samples) for lower bit rate modes, with internal using shorter windows such as 10 ms (80 samples) for LPC parameter estimation to capture quasi-stationary characteristics of the signal. This allows for efficient parameter estimation while minimizing delay. The process begins with voiced/unvoiced detection for each frame, which classifies the speech segment as periodic (voiced) or aperiodic (unvoiced) to guide subsequent modeling. This classification relies on two primary features: the signal's short-term energy, which is higher in voiced frames due to glottal pulses, and the , which is lower for voiced speech owing to its periodic nature compared to the noise-like unvoiced segments. These metrics enable a simple yet effective decision threshold to distinguish frame types without complex computation. For voiced frames, parameter extraction employs a sinusoidal model, representing the speech waveform as a sum of harmonically related sine waves: s(n)=∑m=1MAmcos⁡(ω0mn+θm)s(n) = \sum_{m=1}^{M} A_m \cos(\omega_0 m n + \theta_m), where ω0\omega_0 is the (pitch), AmA_m are the harmonic amplitudes, θm\theta_m the phases, and MM the number of harmonics within the 4 kHz bandwidth. The pitch ω0\omega_0 (typically 50-400 Hz) is estimated using an analysis-by-synthesis approach that minimizes spectral distortion, often via a non-linear for robustness. Harmonic amplitudes are derived from the (DFT) of the windowed frame, averaged over frequency bins around each harmonic to yield root-mean-square (RMS) magnitudes, with the spectral envelope modeled using line spectral pairs (LSPs). The spectral envelope, modeling vocal tract resonances, is captured using line spectral pairs (LSPs), which are roots of polynomials derived from (LPC) coefficients; these provide stable and efficient quantization of the 10th-order filter typically used. Encoding quantizes these extracted parameters into compact fixed-length bit fields, allocating bits to pitch, LSPs, harmonic amplitudes (or ), and voicing flags without employing to maintain low complexity and fixed delay. is applied to LSPs and sometimes amplitude vectors for perceptual optimality, as scalar methods may introduce spectral mismatches; for instance, in the 3200 bit/s mode, parameters are packed into 64 bits per 20 ms frame using multi-stage vector quantizers trained on speech data. Unvoiced frames simplify encoding by modeling excitation shaped by the LSP-derived , reducing bit allocation for harmonics. Decoding reconstructs the speech by synthesizing the sinusoidal components from the quantized parameters. For voiced frames, the speech is synthesized as a sum of harmonically related sine waves using the quantized pitch, amplitudes (derived from the LSP sampled at harmonics), and phases (modeled continuously across frames via quadratic or mixed excitation to avoid discontinuities, with overlap-add windowing of adjacent frames). For unvoiced frames, random is generated and shaped by the spectral derived from LSPs, using an LPC synthesis filter with coefficients aka_k obtained by converting LSPs via the relation A(z)=P(z)+Q(z)2A(z) = \frac{P(z) + Q(z)}{2}, where P(z)P(z) and Q(z)Q(z) are polynomials with roots at the conjugate pairs of LSP frequencies on the unit ; transitions between voiced and unvoiced are blended seamlessly.

Supported Modes and Bit Rates

Codec 2 operates in several fixed-rate modes tailored to varying bandwidth requirements, ranging from 450 bit/s to 3200 bit/s, each defined by a specific number of bits per frame and frame duration to maintain constant output suitable for channel-constrained applications like HF radio. Higher-rate modes typically employ 20 ms frames, while lower-rate modes extend to 40 ms frames to optimize bit efficiency and reduce synchronization overhead. This structure ensures robust through predictable bit-field packing, where parameters are quantized and arranged in a mode-specific order without variable-length coding. The following table summarizes the supported modes, their bit rates, bits per frame, and frame durations:
ModeBit Rate (bit/s)Bits per FrameFrame Duration (ms)
320032006420
240024004820
160016006440
140014005640
130013005240
120012004840
7007002840
4504501840
Bit allocations within each mode prioritize essential speech parameters such as pitch (fundamental frequency), , voicing decisions, and envelope representation via line spectral pairs (LSPs) or (VQ). For example, in the 3200 mode, 7 bits are allocated to pitch, 5 to , 2 to voicing, and the remaining 50 bits to spectral amplitudes using 10 LSPs for high-fidelity representation. In contrast, the 700 mode assigns 6 bits to pitch (including unvoiced and indicators), 4 bits to , and 18 bits to a two-stage VQ for details using fewer effective components, emphasizing robustness in noisy channels. The 450 mode further compresses this to 6 bits for pitch, 3 for , and 9 bits for single-stage VQ of magnitudes from a 512-entry , suitable for extreme bandwidth limits. These allocations reflect trade-offs between and : higher modes dedicate more bits to detailed LSP quantization (e.g., 10 LSPs in 3200 versus effectively 5 in 450 via VQ), preserving nuances, while lower modes consolidate encoding to favor resilience and minimal overhead, often at the cost of naturalness. The fixed bit-field format across modes facilitates efficient packing and decoding, with no additional headers in the core to maintain low latency.

Features and Capabilities

Speech Quality and Performance

Codec 2 delivers communications-grade speech quality suitable for low-bandwidth applications, with Mean Opinion Scores (MOS) typically ranging from 3.5 at 2400 bit/s to around 2.5-3.0 at 1300 bit/s, indicating fair to good perceptual quality for voice transmission. At higher modes like 3200 bit/s, the MOS approaches 4.0, approaching toll-quality levels while maintaining low latency essential for real-time radio use. Intelligibility remains high across modes, enabling effective communication even at ultra-low rates such as 700 bit/s, where subjective evaluations confirm reliable in clean conditions. Common artifacts in Codec 2 output include harmonic distortion during voiced speech segments and added in unvoiced portions, particularly noticeable at below 1000 bit/s. Phase mismatches in the sinusoidal modeling can produce a characteristic "buzzy" , which becomes more prominent in lower-rate modes but does not severely impair overall comprehension. These artifacts stem from the codec's parametric approach, which prioritizes bandwidth efficiency over perfect . In comparisons, Codec 2 outperforms MELP in naturalness and reduced robotic artifacts at under 1000 bit/s, as demonstrated in listening tests where it achieved higher subjective preference scores at 600-700 bit/s. Relative to Opus, Codec 2 is less efficient for general-purpose audio encoding due to Opus's superior compression at rates above 6 kbit/s, but it excels in speech-only scenarios with ultra-low and minimal delay for HF radio. Against , Codec 2 provides better quality at extreme low rates like 700 bit/s while incurring higher , making it preferable for bandwidth-constrained environments despite Speex's broader versatility at moderate rates. Codec 2 demonstrates robustness in noisy environments when integrated with forward error correction (FEC) mechanisms, such as those in the FreeDV system, which mitigate packet loss and channel impairments common in HF radio. Ongoing enhancements, including a 2023-2025 ARDC-funded project as of April 2024, target segmental signal-to-noise ratio (SNR) improvements at 700 bit/s through refined spectral quantization and excitation modeling, aiming to boost performance in adverse conditions without increasing bit rates.

Computational Requirements and Implementations

The of Codec 2, written in , relies on , particularly requiring a hardware (FPU) for real-time operation on microcontrollers due to the high dynamic range in (LPC) analysis that necessitates 32-bit floating-point precision. For encoding and decoding at 3200 bit/s, it demands approximately 10-20 MIPS on or DSP processors, enabling real-time performance on standard hardware. Optimizations for low-power embedded devices include ports targeting to minimize computational overhead, with an STM32 implementation supporting the 700 and 1600 bit/s modes at under 1 MIPS, suitable for resource-constrained environments like microcontrollers. These adaptations leverage ARM CMSIS libraries and compiler flags to achieve efficient execution, such as reducing frame encoding time to about 15 ms per 40 ms frame on STM32F4 series at 180 MHz with optimizations enabled. The core C library supports seamless integration into diverse platforms, facilitating real-time encoding and decoding on personal computers, smartphones via , and microcontrollers. For research purposes, simulation models compatible with and implemented in are provided, allowing algorithm experimentation without hardware dependencies. A key challenge remains the reliance on floating-point operations for LPC and other components, which has driven ongoing refactoring for enhanced portability across fixed-point and low-end architectures; a significant codebase overhaul occurred in July 2023 on the primary repository.

Applications and Adoption

Use in Amateur Radio

Codec 2 serves as the speech codec for legacy modes of FreeDV, an open-source digital voice mode designed for high-frequency (HF) and very high-frequency (VHF) amateur radio transmissions using existing analog radios; FreeDV's flagship mode as of 2025 is RADE, which employs the FARGAN ML . This integration enables low-bitrate voice communication within narrow bandwidths of approximately 1.1 to 1.6 kHz, fitting seamlessly into single sideband (SSB) allocations without requiring dedicated digital infrastructure. In practice, FreeDV employs Codec 2 modes such as 1600 bit/s for operations and 700 bit/s for duplex scenarios, where the compressed audio is modulated using techniques like coherent (COHPSK) modems combined with (FEC) to mitigate errors from channel fading common in HF propagation. Real-world deployments highlight Codec 2's utility in , particularly in contests and emergency communications. Operators utilize FreeDV during organized events like monthly FreeDV Activity Days and dedicated nets, such as the net on 7.182 MHz LSB, to conduct reliable voice contacts under variable conditions. A notable example is its implementation in the LilacSat-1 , launched in 2017 and operational until 2019, which featured an FM uplink to Codec 2 binary (BPSK) digital voice downlink operating at 145.985 MHz uplink and 436.510 MHz downlink, facilitating space-to-ground amateur voice transmissions. Codec 2 is also supported by various software tools, enhancing its accessibility for on-air activities. The primary benefits of Codec 2 in this context include enabling full-duplex voice operations within traditional SSB frequency slots and significantly reducing bandwidth requirements compared to analog (FM), which typically demands 10-15 kHz. This efficiency allows multiple digital voice channels to coexist in crowded spectrum segments, improving spectrum utilization for amateur operators while maintaining intelligible speech over noisy HF links. Codec 2 has also been adopted in the M17 project, an open-source digital voice protocol for VHF/UHF amateur radio that uses Codec 2 as its speech codec, providing a patent-free alternative to proprietary systems like DMR.

Integrations in Software and Hardware

Codec 2 has been integrated into several open-source software frameworks for voice over IP (VoIP) and software-defined radio (SDR) applications. In FreeSWITCH, the mod_codec2 module enables support for Codec 2 encoding and decoding, allowing its use in scalable telephony platforms for low-bitrate audio transmission. Quisk, an SDR application, incorporates Codec 2 through the FreeDV library, enabling real-time digital voice processing in amateur and experimental radio setups. On mobile platforms, Android applications such as codec2_talkie leverage Codec 2 for APRS-enabled digital voice transceivers, providing alternatives to proprietary VoIP tools like MagicJack in low-bandwidth scenarios. In hardware, Codec 2 is embedded in modules from Rowe Research, including the SM1000, which connects to SSB radios for standalone digital voice operation without requiring a PC. These modules support FreeDV modes and are designed for integration into embedded systems. Codec 2 also runs efficiently on single-board computers, where it has been ported for and voice applications, often paired with processors in custom transceivers. Beyond radio, Codec 2 finds use in low-bandwidth over constrained links, such as VoIP , where its compression enables multiple calls on narrow channels—for instance, supporting up to 32 calls at 2000 bit/s on a single 64 kbit/s line. In communications, it powered digital voice downlinks, as demonstrated in LilacSat-1's FM-to-Codec 2 , with ongoing research exploring further applications in space-based systems. The FreeDV provides programmatic access for stacking Codec 2 with custom (FEC) and implementations, allowing developers to build tailored protocols for packet data over radio. Released under the GNU Lesser General Public License (LGPL) version 2.1, Codec 2's licensing promotes widespread adoption in open-source projects by permitting dynamic linking without requiring full source disclosure of host applications, though it imposes restrictions on proprietary modifications that could limit commercial integrations.

History and Development

Key Milestones

Codec 2's development progressed through several key releases and innovations, marking its evolution as an open-source speech codec tailored for low-bandwidth applications. The project achieved its first public milestone on August 25, 2010, with the release of version 0.1 alpha, which introduced the initial 2400 bit/s mode capable of compressing speech for transmission over channels. By 2012, Codec 2 expanded its capabilities with the addition of the 3200 bit/s mode, enabling higher quality options for use. That year, the project also received significant recognition, including the ARRL Technical Innovation Award presented to its creator, David Rowe, for advancing digital voice technologies in . These developments coincided with integrations into major software, such as FreeDV, facilitating real-world testing and adoption. In , enhancements to the bit/s mode culminated in the 700C variant, which incorporated improved phase modeling to enhance speech naturalness and robustness under noisy conditions. The same year saw Codec 2's space deployment aboard LilacSat-1, a launched on May 25 from the , featuring an FM-to-Codec 2 for communications. A notable advancement in ultra-low bandwidth occurred in with the introduction of the experimental bit/s mode, designed to support digital voice transmission at signal-to-noise ratios as low as -4 dB, opening possibilities for earth-moon-earth (EME) contacts and extreme low-power operations. The project reached a milestone on July 24, 2023, with the release of version 1.2.0, which included optimizations for performance, bug fixes, and repository cleanup to streamline ongoing development.

Recent and Future Developments

In July 2023, the Codec 2 project underwent a major refactoring on its primary repository, where legacy code was migrated to a separate deprecated repository (codec2-dev) to streamline ongoing development and focus on newer implementations. This restructuring supported active work on new algorithms and FreeDV modes, including enhancements to low-bitrate encoding techniques. A significant initiative began in early 2023 with an ARDC grant funding the WP2000 , aimed at improving Codec 2 speech at approximately 700 bit/s and 1200-2400 bit/s through advanced spectral quantization, excitation modeling, and robustness. The two-year program, running through 2025, involves testing against speech samples and comparisons to SSB and commercial codecs to boost performance in noisy environments. By April 2024, the pivoted toward approaches, pausing traditional Codec 2 development in favor of the Radio (RADE), a neural network-based mode integrated into FreeDV for better low-bitrate efficiency. From 2024 to 2025, no major stable releases of Codec 2 occurred, but repository commits emphasized modem improvements, such as HF OFDM and FSK enhancements for the , alongside support for embedded platforms like via the SM1000 hardware. Research branches explored AI-assisted parameter estimation, exemplified by RADE's architecture, which reduces quantization needs at rates below 700 bit/s—offering a conceptual bridge to emerging neural speech coders while maintaining open-source accessibility for spectrum-constrained use. A preview release of FreeDV in October 2024 introduced the RADE mode, marking initial deployment of these ML-driven enhancements. Looking ahead, development plans prioritize greater speech naturalness at bit rates under 700 bit/s, leveraging ML to surpass traditional sinusoidal modeling limits, as outlined in the ongoing ARDC project pivot confirmed in mid-2025. Community input shapes these efforts through the 2024 FreeDV Feature Request Form, which solicits proposals for new modes and integrations to sustain Codec 2's role in low-bandwidth HF/VHF applications. This focus addresses sustainability in amateur radio by enabling robust digital voice in bandwidth-limited scenarios, with RADE providing a pathway for broader adoption in resource-constrained systems.

References

  1. https://www.rowetel.com/wordpress/?page_id=452
  2. https://files.tapr.org/meetings/DCC_2011/DCC2011-Codec2-VK5DGR.pdf
  3. https://github.com/drowe67/codec2
  4. https://www.icqpodcast.com/news/2023/3/15/ardc-grant-funds-freedv-project
  5. https://freedv.org/
  6. https://www.drpeterjamieson.com/html_papers/csci_2021.pdf
  7. https://files.tapr.org/meetings/DCC_2011/DCC2011-Codec2-Slides-VK5DGR.pdf
  8. http://www.arrl.org/news/arrl-board-of-directors-names-award-recipients-at-2012-second-meeting
  9. https://www.rowetel.com/downloads/1997_rowe_phd_thesis.pdf
  10. https://tapr.org/wp-content/uploads/DCC2010-Codec2-presentation-K6BP-VK5DGR.pdf
  11. https://www.rowetel.com/wordpress/?p=128
  12. https://www.rowetel.com/downloads/codec2/thomas/open_souce_speech_codec_450bps.pdf
  13. https://www.rowetel.com/wordpress/?p=5373
  14. https://www.rowetel.com/?page_id=452
  15. https://www.rowetel.com/wordpress/?p=5520
  16. http://opus-codec.org/comparison/
  17. https://www.rowetel.com/wordpress/?p=5966
  18. https://freedv.org/ardc-grant-project-plan/
  19. https://freedv.org/davids-freedv-update-april-2024/
  20. https://www.researchgate.net/publication/363748210_Transplantation_of_Codec2_Speech_Compression_Algorithm_Based_on_STM32_Processor
  21. https://stackoverflow.com/questions/77515896/audio-codec2-on-stm32
  22. https://github.com/openwebrx/pkg-codec2
  23. https://freedv.org/radio-autoencoder/
  24. https://vk3tbs.home.blog/2025/08/23/freedv/
  25. https://amsat-uk.org/2017/05/19/lilacsat-1-cubesat-iss/
  26. https://www.nanosats.eu/sat/lilacsat-1.html
  27. https://m17project.org/
  28. https://developer.signalwire.com/freeswitch/FreeSWITCH-Explained/Modules/mod_codec2_6587354/
  29. https://groups.io/g/n2adr-sdr/topic/quisk_version_4_2_29_january/103666841
  30. https://github.com/sh123/codec2_talkie
  31. https://www.rowetel.com/?p=8084
  32. http://blog.marxy.org/2012/12/codec2-and-modem-on-raspberry-pi.html
  33. https://www.youtube.com/watch?v=KsywWf8dQgU
  34. https://destevez.net/2016/10/lilacsat-1-codec-2-downlink/
  35. https://www.rowetel.com/?p=655
  36. http://www.arrl.org/news/surfin-developing-new-digital-voice-software
  37. https://www.rowetel.com/?p=2605
  38. https://www.rowetel.com/?p=5373
  39. https://www.rowetel.com/?p=6212
  40. https://github.com/drowe67/codec2/releases/tag/1.2.0
  41. https://freedv.org/2024/10/
  42. https://freedv.org/2025/07/
  43. https://freedv.org/wp-content/uploads/2024/03/FreeDV-027-Feature-Request-Form-V1.1.pdf
Add your contribution
Related Hubs
User Avatar
No comments yet.