{"id":115540,"date":"2024-05-04T19:52:06","date_gmt":"2024-05-04T17:52:06","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/osnovatel-qemu-i-ffmpeg-opublikoval-zvukovoj-kodek-tsac"},"modified":"2024-05-04T19:52:06","modified_gmt":"2024-05-04T17:52:06","slug":"osnovatel-qemu-i-ffmpeg-opublikoval-zvukovoj-kodek-tsac","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/osnovatel-qemu-i-ffmpeg-opublikoval-zvukovoj-kodek-tsac","title":{"rendered":"The founder of QEMU and FFmpeg published the TSAC audio codec","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
French mathematician Fabrice Bellard, who founded projects such as QEMU, FFmpeg, BPG, QuickJS, TinyGL, and TinyCC, has published a sound encoding format called TSAC along with associated tools for compressing and decompressing audio files. The format is designed for data transmission at very low bitrates, for example, 5.5 kb\/s for mono and 7.5 kb\/s for stereo, while maintaining acceptable quality for music and speech. Using TSAC allows a musical composition lasting 3.5 minutes with a sampling rate of 44.1 kHz (stereo) to be packed into a file size of 192 KB, which will be almost indistinguishable from the original to the ear of an untrained listener. The project code is distributed under the MIT license. <\/p>\n
TSAC is based on the Descript audio codec, which has been enhanced to support stereo sound and translated to utilize a different machine learning model based on a neural network architecture called 'transformer.' This new implementation allows for increased compression by reconstructing lost details with respect to the model of human auditory perception. The model is approximately 200 MB in compressed form and is presented in a deterministic format, ensuring that the same result is produced regardless of the CPU\/GPU used and the number of threads involved in the computations. <\/p>\n
The encoder can operate using only the CPU for calculations (with AVX2 instructions supported for acceleration), but for achieving high performance, using a GPU is recommended. Currently, CUDA API can be used for acceleration with NVIDIA GPUs based on Ampere, ADA, and Hopper microarchitectures (RTX 3090, RTX 4090, RTX A6000, A100, and H100) that have at least 4 GB of video memory. FFmpeg is used for converting audio files before encoding.
Additionally, it's worth mentioning the updated utility ts_zip developed by Bellar, intended for efficient compression of text data, using a token prediction mechanism based on a machine learning system and the large language model RWKV 169M v4. When compressing a Wikipedia archive, the ts_zip utility achieved a data compression ratio of 7.3 times, while compressing the Linux kernel code 1.2 yielded a ratio of 7.8 times. In comparison, the compression levels using the xz utility were 4.7 and 5.5 times, respectively. The trade-off for this high compression efficiency is low compression speed and significant resource requirements (a minimum of 4 GB of RAM). On a system with an RTX 4090 GPU, the compression performance is approximately 1 MB\/s. \n Source: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"