RuntimeError: The frame has 0 channels, expected 1. If you are hitting this, it may be because you are using a buggy FFmpeg version. FFmpeg4 is known to fail here in some valid scenarios. Try to upgrade FFmpeg?

[obadx]

🐛 Describe the bug

I having an error trying to read single channel wave file. I tried to upgrade ffmpeg version but the same error remains.

Reproducing the error using Google Colab Notebook:

! apt install ffmpeg
! pip install torchcodec

! ffmpeg -version

! wget https://raw.githubusercontent.com/obadx/quran-muaalem/main/assets/test.wav

from torchcodec.decoders import AudioDecoder


decoder = AudioDecoder("./test.wav", sample_rate=16000, num_channels=1)
print(decoder.get_all_samples().data[0].shape)

RuntimeError: The frame has 0 channels, expected 1. If you are hitting this, it may be because you are using a buggy FFmpeg version. FFmpeg4 is known to fail here in some valid scenarios. Try to upgrade FFmpeg?

! ffmpeg -i test.wav output.mp3

from torchcodec.decoders import AudioDecoder


decoder = AudioDecoder("./output.mp3", sample_rate=16000, num_channels=1)
print(decoder.get_all_samples().data[0].shape)
# torch.Size([140320])

Versions

Collecting environment information...
PyTorch version: 2.8.0+cu126
Is debug build: False
CUDA used to build PyTorch: 12.6
ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.4 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04.2) 11.4.0
Clang version: 14.0.0-1ubuntu1.1
CMake version: version 3.31.6
Libc version: glibc-2.35

Python version: 3.12.11 (main, Jun  4 2025, 08:56:18) [GCC 11.4.0] (64-bit runtime)
Python platform: Linux-6.1.123+-x86_64-with-glibc2.35
Is CUDA available: False
CUDA runtime version: 12.5.82
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_adv.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_cnn.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_precompiled.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_runtime_compiled.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_graph.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_heuristic.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_ops.so.9.2.1
Is XPU available: False
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture:                         x86_64
CPU op-mode(s):                       32-bit, 64-bit
Address sizes:                        46 bits physical, 48 bits virtual
Byte Order:                           Little Endian
CPU(s):                               2
On-line CPU(s) list:                  0,1
Vendor ID:                            GenuineIntel
Model name:                           Intel(R) Xeon(R) CPU @ 2.20GHz
CPU family:                           6
Model:                                79
Thread(s) per core:                   2
Core(s) per socket:                   1
Socket(s):                            1
Stepping:                             0
BogoMIPS:                             4399.99
Flags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single ssbd ibrs ibpb stibp fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt arat md_clear arch_capabilities
Hypervisor vendor:                    KVM
Virtualization type:                  full
L1d cache:                            32 KiB (1 instance)
L1i cache:                            32 KiB (1 instance)
L2 cache:                             256 KiB (1 instance)
L3 cache:                             55 MiB (1 instance)
NUMA node(s):                         1
NUMA node0 CPU(s):                    0,1
Vulnerability Gather data sampling:   Not affected
Vulnerability Itlb multihit:          Not affected
Vulnerability L1tf:                   Mitigation; PTE Inversion
Vulnerability Mds:                    Vulnerable; SMT Host state unknown
Vulnerability Meltdown:               Vulnerable
Vulnerability Mmio stale data:        Vulnerable
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed:               Vulnerable
Vulnerability Spec rstack overflow:   Not affected
Vulnerability Spec store bypass:      Vulnerable
Vulnerability Spectre v1:             Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers
Vulnerability Spectre v2:             Vulnerable; IBPB: disabled; STIBP: disabled; PBRSB-eIBRS: Not affected; BHI: Vulnerable
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Vulnerable

Versions of relevant libraries:
[pip3] intel-cmplr-lib-ur==2025.2.1
[pip3] intel-openmp==2025.2.1
[pip3] mkl==2025.2.0
[pip3] numpy==2.0.2
[pip3] nvidia-cublas-cu12==12.6.4.1
[pip3] nvidia-cuda-cupti-cu12==12.6.80
[pip3] nvidia-cuda-nvrtc-cu12==12.6.77
[pip3] nvidia-cuda-runtime-cu12==12.6.77
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cufft-cu12==11.3.0.4
[pip3] nvidia-curand-cu12==10.3.7.77
[pip3] nvidia-cusolver-cu12==11.7.1.2
[pip3] nvidia-cusparse-cu12==12.5.4.2
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-nccl-cu12==2.27.3
[pip3] nvidia-nvjitlink-cu12==12.6.85
[pip3] nvidia-nvtx-cu12==12.6.77
[pip3] nvtx==0.2.13
[pip3] optree==0.17.0
[pip3] pynvjitlink-cu12==0.7.0
[pip3] tbb==2022.2.0
[pip3] tcmlib==1.4.0
[pip3] torch==2.8.0+cu126
[pip3] torchao==0.10.0
[pip3] torchaudio==2.8.0+cu126
[pip3] torchcodec==0.6.0
[pip3] torchdata==0.11.0
[pip3] torchsummary==1.5.1
[pip3] torchtune==0.6.1
[pip3] torchvision==0.23.0+cu126
[pip3] triton==3.4.0
[pip3] umf==0.11.0
[conda] Could not collect

[Dan-Flores]

Hi @obadx, thanks for sharing the reproducing example.

It looks like running apt install ffmpeg in Google Collab installs ffmpeg 4.4 as the latest available version, which leads to the error message you are seeing.

To install a newer version of ffmpeg, alternate approaches include using a PPA (Personal Package Archive) or installing a static ffmpeg build manually.

[obadx]

Thanks for trying to help me @Dan-Flores . I tried that too but fails. can you install the appropriate version so it can work ?

[Dan-Flores]

Google Collab controls the environment and the versions of FFmpeg that are available. If you manage to get a working solution using alternative approaches, please share it here!

[obadx]

@Dan-Flores In addition to Colab, I tried FFmpeg version 7 on a different machine, but the issue remains. The issue is not related to this file only; I created a Gradio app for testing, and every recorded audio fails with the same error. I like this library because it is fast, but unfortunately, the error is persistent and I cannot solve it. May be the error cause is diffrent but the Exception does not handle it correctly.

It will be nice if you include ffmpeg binaries in the package it self to avoid such conflicts same as pytorch has done with NVIDIA runtimes

[zucchini-nlp]

Facing the same issue with ffmpeg 4 and the documentation states that version from 4 to 7 are all supported. @NicolasHug any plans to fix it?

We made torchcodec a default audio decoder in transformers in yesterday's release, but I can revert it again until wav files are decoded without issues

[NicolasHug]

I tried FFmpeg version 7 on a different machine, but the issue remains

@obadx , I am unable to reproduce what you are observing. I am able to successfully decode this file on FFmpeg 7 locally. Can you try again and confirm the FFmpeg version that you are using?

@zucchini-nlp we do support FFmpeg 4 to 7 :)
We can try to see if we can work-around the issue, but I don't believe this is a bug in TorchCodec: this is a problem with FFmpeg 4, specifically. Is there a change you could recommend the transformers users to use ffmpeg > 4 ?

[zucchini-nlp]

@NicolasHug we definitely can check if ffmpeg version is higher than 4 before decoding with torchcodec, and otherwise fallback to librosa. Just wanted to check-in and see it has a fix in torch side :)

[NicolasHug]

Thanks @zucchini-nlp , let us try to see if we can work around the issue within TorchCodec in some way. The check is here to prevent any incorrect memory access, but perhaps there is a way to maintain safety without errorring, which I definitely agree would provide a better UX. We'll keep you posted on the progress :)