{"slug": "torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder", "title": "TorchCodec 0.14: HDR Video Decoding for CPU and CUDA, and Fast Wav Decoder", "summary": "Meta's PyTorch team released TorchCodec 0.14, adding a fast WavDecoder that bypasses FFmpeg for direct WAV file reading and HDR video decoding support for CPU and CUDA with full float32 precision. The update also improves audio seeking speed and removes the NVIDIA NPP dependency for simpler CUDA decoding setup.", "body_md": "[TorchCodec 0.14](https://meta-pytorch.org/torchcodec/0.14/) is out! It is compatible with `torch >= 2.11`\n\n. It comes with two major additions: a fast audio [ WavDecoder](https://meta-pytorch.org/torchcodec/0.14/generated/torchcodec.decoders.WavDecoder.html#torchcodec.decoders.WavDecoder), and support for\n\n[HDR video decoding](https://meta-pytorch.org/torchcodec/0.14/generated_examples/decoding/hdr_decoding.html)!\n\n## Fast wav decoder\n\nTorchCodec now has a dedicated [ WavDecoder](https://meta-pytorch.org/torchcodec/0.14/generated/torchcodec.decoders.WavDecoder.html#torchcodec.decoders.WavDecoder) for decoding WAV files. It bypasses FFmpeg entirely and reads WAV data directly, resulting in significantly faster decoding. It supports multiple sample formats (int16, int32, float32, etc.), and can decode from files, bytes, or file-like objects.\n\n``` python\nfrom torchcodec.decoders import WavDecoder\n\ndecoder = WavDecoder(\"audio.wav\")\nsamples = decoder.get_all_samples()  # AudioSamples with data and sample_rate\n```\n\n## HDR Video Decoding\n\n`VideoDecoder`\n\n[now supports HDR](https://meta-pytorch.org/torchcodec/0.14/generated_examples/decoding/hdr_decoding.html) (High Dynamic Range) video decoding without losing precision. When `output_dtype=torch.float32`\n\nis specified, the decoder outputs RGB float32 frames in `[0, 1]`\n\n, preserving the full HDR color range. This is supported for both CPU and CUDA!\n\n``` python\nimport torch\nfrom torchcodec.decoders import VideoDecoder\n\ndecoder = VideoDecoder(\"hdr_video.mp4\", output_dtype=torch.float32)\nframe = decoder[0]  # Full HDR precision in float32\n```\n\n## Other Improvements\n\n**Improved audio seeking**:seeking is now much faster (`AudioDecoder`\n\n[#1449](https://github.com/meta-pytorch/torchcodec/pull/1449))**Dropped NPP dependency**:`TorchCodec`\n\nno longer depends on NVIDIA's NPP library, which will simplify installing and using TorchCodec for CUDA decoding.", "url": "https://wpnews.pro/news/torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder", "canonical_source": "https://github.com/meta-pytorch/torchcodec/releases/tag/v0.14.0", "published_at": "2026-06-10 14:08:36+00:00", "updated_at": "2026-06-14 22:40:51.126057+00:00", "lang": "en", "topics": ["developer-tools", "ai-infrastructure"], "entities": ["Meta", "PyTorch", "TorchCodec", "FFmpeg", "NVIDIA", "NPP"], "alternates": {"html": "https://wpnews.pro/news/torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder", "markdown": "https://wpnews.pro/news/torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder.md", "text": "https://wpnews.pro/news/torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder.txt", "jsonld": "https://wpnews.pro/news/torchcodec-0-14-hdr-video-decoding-for-cpu-and-cuda-and-fast-wav-decoder.jsonld"}}