当前位置：文江博客话题详情

从 wav 和 aiff 文件创建 mp3 预览

发布于 2024-08-27 12:23:42 字数 216 浏览 9 评论 0原文

我想创建一个程序来制作 aiff 或 wav 文件前 30 秒的 mp3。我还希望能够选择位置和长度，例如 2:12 到 2:42 之间的音频。有没有任何工具可以让我做到这一点？

脱壳就OK了。该应用程序将在 Linux 服务器上运行，因此它必须是一个可以在 Linux 上运行的工具。

我不介意分两步完成——即首先创建 aiff/wav 剪切图的工具，然后将其传递给 mp3 编码器。

原文

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

萌逼全场 2024-09-03 12:23:42

SoX 和 trim 谓词可以做到这一点。如果您的 sox 没有构建 MP3 支持，那么您必须在之后将输出通过管道传输到 lame，或者找到一个支持 MP3 的。

回复收藏 0 原文

叫嚣ゝ 2024-09-03 12:23:42

我想使用尽可能低级别的东西，所以我最终使用了 RubyAudio ，libsndfile 的包装器。

require "rubygems"
require "ruby-audio"

EXTRACT_BEGIN = 11.2
EXTRACT_LENGTH = 3.5

RubyAudio::Sound.open("/home/augustl/sandbox/test.aif") do |snd|
  info = snd.info
  ["channels", "format", "frames", "samplerate", "sections", "seekable"].each do |key|
    puts "#{key}: #{info.send(key)}"
  end

  # TODO: should we use a 1000 byte buffer? Does it matter? See RubyAudio::Sound rdocs.
  bytes_to_read = (info.samplerate * EXTRACT_LENGTH).to_i
  buffer = RubyAudio::Buffer.new("float", bytes_to_read, info.channels)

  snd.seek(info.samplerate * EXTRACT_BEGIN)
  snd.read(buffer, bytes_to_read)

  out = RubyAudio::Sound.open("/home/augustl/sandbox/out.aif", "w", info.clone)
  out.write(buffer)
end

I wanted to use something as low level as possible, so I ended up using RubyAudio, a wrapper for libsndfile.

require "rubygems"
require "ruby-audio"

EXTRACT_BEGIN = 11.2
EXTRACT_LENGTH = 3.5

RubyAudio::Sound.open("/home/augustl/sandbox/test.aif") do |snd|
  info = snd.info
  ["channels", "format", "frames", "samplerate", "sections", "seekable"].each do |key|
    puts "#{key}: #{info.send(key)}"
  end

  # TODO: should we use a 1000 byte buffer? Does it matter? See RubyAudio::Sound rdocs.
  bytes_to_read = (info.samplerate * EXTRACT_LENGTH).to_i
  buffer = RubyAudio::Buffer.new("float", bytes_to_read, info.channels)

  snd.seek(info.samplerate * EXTRACT_BEGIN)
  snd.read(buffer, bytes_to_read)

  out = RubyAudio::Sound.open("/home/augustl/sandbox/out.aif", "w", info.clone)
  out.write(buffer)
end

回复收藏 0 原文

四叶草在未来唯美盛开 2024-09-03 12:23:42

使用 LAME 作为 mp3 编码部分。使用 shntplit 分割文件。您需要将分割点放入提示文件中，但这很容易。

回复收藏 0 原文

笑，眼淚并存 2024-09-03 12:23:42

在包含 *.wav 文件的目录中运行此 Bash 单行代码。

for wavfile in *.wav; do \
  sox "${wavfile}" "preview-${wavfile}" trim 0 60 fade 3 57 3; \
  lame --preset standard "preview-${wavfile}" \
    "preview-`basename ${wavfile} .wav`".mp3; \
  rm "preview-${wavfile}"; \
done

第一个 60 秒。 3 秒淡入和 3 秒淡出。原始 wav 文件保持不变。预览文件带有“preview-”前缀。您可以通过更改“trim 0 60”来选择位置和长度以满足您的需求。
需要：sox、lame

如果您有一个包含 mp3 文件的目录并且需要创建预览，请运行以下命令：

for mp3file in *.mp3; do \
  mpg123 -w "${mp3file}.wav" "${mp3file}"; \
  sox "${mp3file}.wav" "preview-${mp3file}.wav" trim 0 60 fade 3 57 3; \
  rm "${mp3file}.wav"; \
  lame --preset standard "preview-${mp3file}.wav" "preview-${mp3file}"; \
  rm -v "preview-${mp3file}.wav"; \
done

需要：mpg123、sox、lame

Run this Bash one-liner in a directory with *.wav files.

for wavfile in *.wav; do \
  sox "${wavfile}" "preview-${wavfile}" trim 0 60 fade 3 57 3; \
  lame --preset standard "preview-${wavfile}" \
    "preview-`basename ${wavfile} .wav`".mp3; \
  rm "preview-${wavfile}"; \
done

First 60 seconds. 3 seconds fade-in and 3 seconds fade-out. Original wav files stay untouched. Preview files come with a "preview-" prefix. You'll be able to choose location and length by changing "trim 0 60" to fit Your needs.
Requires: sox, lame

If You have a directory with mp3 files and need to create previews, run this:

for mp3file in *.mp3; do \
  mpg123 -w "${mp3file}.wav" "${mp3file}"; \
  sox "${mp3file}.wav" "preview-${mp3file}.wav" trim 0 60 fade 3 57 3; \
  rm "${mp3file}.wav"; \
  lame --preset standard "preview-${mp3file}.wav" "preview-${mp3file}"; \
  rm -v "preview-${mp3file}.wav"; \
done

Requires: mpg123, sox, lame

回复收藏 0 原文

十秒萌定你 2024-09-03 12:23:42

我写了一个 python 库 pydub，这使得这变得微不足道，尽管它使用 ffmpeg 来进行转换以支持更多格式...

from pydub import AudioSegment

sound = AudioSegment.from_file("/input/file.aiff", format="aif")

# 2 min and 12 sec, them convert to milliseconds
start = (2*60 + 12) * 1000
end = start +  (30 * 1000)
snip = sound[start:end]

# add 3 second fade in and fade out
snip = snip.fadeIn(3000).fadeOut(3000)

# save as mp3
snip.export("/output/file.mp3", format="mp3")

I wrote a python library, pydub, that makes this trivial, though it uses ffmpeg in to do the conversions in order to support more formats…

from pydub import AudioSegment

sound = AudioSegment.from_file("/input/file.aiff", format="aif")

# 2 min and 12 sec, them convert to milliseconds
start = (2*60 + 12) * 1000
end = start +  (30 * 1000)
snip = sound[start:end]

# add 3 second fade in and fade out
snip = snip.fadeIn(3000).fadeOut(3000)

# save as mp3
snip.export("/output/file.mp3", format="mp3")

回复收藏 0 原文

~没有更多了~