Definition of the Opus Audio Codec

By Jean-Marc Valin, Koen Vos, Timothy B. Terriberry. 2012

372 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Speech & Audio, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Speech & Audio

In Signal Processing

Citations in the corpus

Toward Domain-Invariant Speech Recognition via Large Scale Training (2018, background): “MP3 at 64 kbps, Opus [33, 21] at 24 kbps, and SBC [34, 35] with a bitpool size of 24, 16 blocks and 8 subbands are unseen during training; the rest are seen.”
vq-wav2vec: Self-Supervised Learning of Discrete Speech Representations (2019, methodology): “As baselines, we consider various lossy compression algorithms applied to the TIMIT audio data and train wav2letter models on the resulting audio: Codec2 3 as a low bitrate codec, Opus (Terriberry & Vos, 2012) as a medium bitrate codec and MP3 and Ogg Vorbis (Montgomery, 2004) as high bitrate codecs”
SoundStream: An End-to-End Neural Audio Codec (2021, methodology): “Traditional audio codecs – Opus [9] and EVS [10] are state-of-the-art audio codecs, which combine traditional coding tools, such as LPC, CELP and MDCT, to deliver high coding efﬁciency over different content types, bitrates and sampling rates, while ensuring low-latency for real-time audio…”
High Fidelity Neural Audio Compression (2022, methodology): “Lyra-v2 is a neural audio codec, while EVS and Opus are competitive standard codecs.”
FunCodec: A Fundamental, Reproducible and Integrable Open-Source Toolkit for Neural Speech Codec (2023, background): “Traditional speech codecs rely on a carefully designed pipeline that incorporates expert knowledge of psycho-acoustics and speech synthesis to achieve efficient coding [1,2].”
High-Fidelity Audio Compression with Improved RVQGAN (2023, neutral): “For EnCodec, Lyra, and Opus, we use publicly available open-source implementations provided by the authors.”
Speech Resynthesis from Discrete Disentangled Self-Supervised Representations (2021, methodology): “For comparison, we included other speech codecs in our evaluation: Opus [36] wideband at 9 kbps VBR, Codec2 [52] at 2.4 kbps and LPCNet [32] operating at 1.6 kbps.”
Enhancing into the Codec: Noise Robust Speech Coding with Vector-Quantized Autoencoders (2021, background): “The most popular codecs, like the Opus codec in wideband mode, can produce high-quality speech compression at around 9 kb/s [1].”
A Real-Time Wideband Neural Vocoder at 1.6 kb/s Using LPCNet (2019, neutral): “Considering that uncompressed LPCNet is able to achieve higher quality than 9 kb/s Opus, we believe it is worth exploring higher LPCNet bitrates in the 2-6 kb/s range.”
End-to-End Binaural Speech Synthesis (2022, background): “Audio codecs have long relied on traditional signal processing and in-domain knowledge of psychoacoustics in order to perform encoding of speech [6] or general audio signals [7, 8].”
Neural Feature Predictor and Discriminative Residual Coding for Low-Bitrate Speech Coding (2022, background): “It has been used in various communication and entertainment applications after standardization [1, 2] and open-sourcing [3].”
Source Coding of Audio Signals with a Generative Model (2020, methodology): “[4] Jan Skoglund and Jean-Marc Valin, “Improving Opus low bit rate quality with neural speech synthesis,” arXiv preprint arXiv:1905.04628, 2019.”
Disentangled Feature Learning for Real-Time Neural Speech Coding (2022, methodology): “We compare our Disen-TF-Codec for single speaker case with Opus [21] and Lyra [3], two codecs used for real-time communications.”
Perceived Audio Quality for Streaming Stereo Music (2014, background): “The first test used 10 samples which matched those used by Hoene et al.[7] in MUSHRA tests to evaluate the OPUS codec.”
High Quality Audio Coding with Mdctnet (2022, background): “A subjective assessment indicated that at 24 kb/s the deep codec offers an overall-mean performance equal to the Opus switched speech-transform codec [16] operating at twice the bitrate.”
Objective Assessment of Perceptual Audio Quality Using ViSQOLAudio (2017, background): “When audio is encoded, many popular encoders add a buffer of zero signal samples to the beginning of the degraded (encoded) signal during the encoding windowing process [16], [17].”