audio

audio copied to clipboard

### 🐛 Describe the bug I attempted to use highpass_biquad to calculate a loss while inside of an autocast block: ``` self.floatFormat = torch.bfloat16 if torch.cuda.is_bf16_supported() else torch.float32 ... with...

pokepress

### 🐛 Describe the bug Windows Nightly builds are broken: [hud.pytorch.org/hud/pytorch/audio/nightly/1?per_page=50&name_filter=windows](https://hud.pytorch.org/hud/pytorch/audio/nightly/1?per_page=50&name_filter=windows) Workflow: [pytorch/audio/actions/runs/13030436154/job/36348301796#step:12:3346](https://github.com/pytorch/audio/actions/runs/13030436154/job/36348301796#step:12:3346) Error: ``` 2025-01-29T11:58:41.0321201Z C:/actions-runner/_work/_temp/conda_environment_13030436154/Lib/site-packages/torch/include\torch/csrc/dynamo/compiled_autograd.h(962): error C2872: 'std': ambiguous symbol ``` Failure is related to this PR: https://github.com/pytorch/pytorch/pull/144707...

atalman

fix: fixing batching in MMS_FA

3

Fixing batching in MMS_FA bundle

ex3ndr

Avoid torch.jit.isinstance to support torch.compile(MFCC(), fullgraph=True)

3

The following code fails on main but works in this PR. ```py from torchaudio.transforms import MFCC import torch mfcc = torch.compile(MFCC(), fullgraph=True) ``` Unfortunately, the compiled model can still not...

twoertwein

### 🐛 Describe the bug Related [issue is here](https://github.com/conda/conda/issues/14497). Package on conda-forge (`linux-64/torchaudio-2.5.1-cpu_py312h6db6e6d_0.conda`) is missing ffmpeg dynamic libs, so it is not possible to use it. Please fix this package....

andrew-aladjev

I applied SQUIM to assess speech quality as a way to correct the direction-of-arrival of a location-based speech enhancement system. [More info here](https://www.sciencedirect.com/science/article/pii/S1051200424005840). I'm feeding the last 3-second window of...

balkce

Add instruction to implement audio augmentation using tourchaudio.sox…

2

…_effects PLEASE NOTE THAT THE TORCHAUDIO REPOSITORY IS NO LONGER ACTIVELY MONITORED. You may not get a response. For open discussions, visit https://discuss.pytorch.org/.

RosieLe12

Add support for Pseudo Quadrature Mirror Filter

1

### 🚀 The feature Add ability to apply PQMF to singal and inverse transform to reconstruct signal. ### Motivation, pitch PQMF is used for example in Multi-band MelGAN and DurIAN....

uis246

### 🐛 Describe the bug I am trying to convert Nvidia NeMo's FilterbankFeaturesTA class to ONNX. Here is my code - ``` from nemo.collections.asr.parts.preprocessing.features import ( FilterbankFeatures, FilterbankFeaturesTA, make_seq_mask_like, )...

kabyanil

### 🚀 The feature Hi just interested roughly when we will be supporting fffmpeg 7 It supports av1 hardware encoding/decoding on 40 series or L4 GPUs ### Motivation, pitch just...

lee101