Using the output to `stdout`

Sometimes you would like to process the results processed by ffmpeg directly in memory without outputting them to a file. In this case, you can utilize an output of FFmpeg.execute().

Note

To use the output to ffmpeg's stdout, you must use pipe:1 as the URL for the output file.

For example, you can extract audio from video in wave format and analyze the information of the audio directly as follows.

Synchronous APIAsynchronous API

import io
import wave

from ffmpeg import FFmpeg, Progress


def main():
    ffmpeg = (
        FFmpeg()
        .option("y")
        .input("input.mp4")
        .output(
            "pipe:1",
            {"codec:a": "pcm_s16le"},
            vn=None,
            f="wav",
        )
    )

    @ffmpeg.on("progress")
    def on_progress(progress: Progress):
        print(progress)

    wave_bytes = ffmpeg.execute()
    with wave.open(io.BytesIO(wave_bytes), "rb") as wave_file:
        print("Sample width in bytes:", wave_file.getsampwidth())
        print("Sampling frequency:", wave_file.getframerate())
        print("Number of frames:", wave_file.getnframes())


if __name__ == "__main__":
    main()

import asyncio
import io
import wave

from ffmpeg import Progress
from ffmpeg.asyncio import FFmpeg


async def main():
    ffmpeg = (
        FFmpeg()
        .option("y")
        .input("input.mp4")
        .output(
            "pipe:1",
            {"codec:a": "pcm_s16le"},
            vn=None,
            f="wav",
        )
    )

    @ffmpeg.on("progress")
    def on_progress(progress: Progress):
        print(progress)

    wave_bytes = await ffmpeg.execute()
    with wave.open(io.BytesIO(wave_bytes), "rb") as wave_file:
        print("Sample width in bytes:", wave_file.getsampwidth())
        print("Sampling frequency:", wave_file.getframerate())
        print("Number of frames:", wave_file.getnframes())


if __name__ == "__main__":
    asyncio.run(main())

Using the output to stdout

Using the output to `stdout`