Compression Settings for Audio Files

Here’s a cookbook for compressing audio files for distribution via the web. This assumes we want to create the smallest files with the highest audio quality.

Whether you use Apple Compressor, Adobe Media Encoder or any of a wide variety of audio compression applications, these rules all hold true.

BASIC RULES

You don’t really need to compress audio files. They are small when compared to video. However, the larger the file size, the longer it takes to download and web audiences today have a very short threshold for waiting.

Given that, here are some rules that can help you determine compression settings.

• The audio codec for MPEG-3 files is MP3.
• The audio codec for MPEG-4 files is AAC.
• Stereo files require double the bit rate (data rate) of mono files.
• MP3 files require double the bit rate of MP4 files.
• Video records an audio sample rate of 48,000.
• The Nyquist Theorem states that dividing the sample rate by 2 equals the maximum frequency response of an audio file.
• Normal human hearing goes from 20 to 20,000 Hz. Therefore, the maximum sample rate we need to cover the entire range of human hearing is 40,000 samples.
• The nearest common sample rate to 40,000 is 44,100.
• Human speech only requires a portion of the total range of human hearing; roughly 200 – 10,000 Hz. This means that for audio files that only contain people speaking, we need a sample rate of 20,000.
• The nearest sample rate to 20,000 is 22,050.
• We never need to worry about low frequencies when compressing audio.

CREATE MP3 FILES

For music files, use:

• Stereo
• Sample Rate: 44,100 (44.1K)
• Bit rate: 256 kbps
• VBR (Variable Bit Rate) encoding if possible, otherwise use CBR

For single speaker spoken word files, use:

• Mono
• Sample rate: 22,100 if your software supports it, otherwise 32,000
• Bit rate: 64 kbps (22.1K) or 96 kbps (32K)
• VBR encoding if possible, otherwise use CBR

If your spoken word file has lots of sound effects, incidental music and/or lots of stereo effects, use:

• Stereo
• Sample Rate: 32,000
• Bit rate: 192 kbps
• VBR encoding if possible, otherwise use CBR

CREATE MPEG-4 FILES

For music files, use:

• Stereo
• Sample Rate: 44,100 (44.1K)
• Bit rate: 128 kbps
• VBR (Variable Bit Rate) encoding if possible, otherwise use CBR

For single speaker spoken word files, use:

• Mono
• Sample rate: 22,100 if your software supports it, otherwise 32,000
• Bit rate: 48 kbps (22.1K) or 56 kbps (32K)
• VBR if possible, otherwise CBR

If your spoken word file has lots of sound effects, incidental music and/or lots of stereo effects, use:

• Stereo
• Sample Rate: 32,000
• Bit rate: 96 kbps
• VBR if possible, otherwise CBR

SUMMARY

While it is true that every file is different, these settings will work fine in most instances. Give them a try, then tweak as necessary.

As always, let me know what works best for you.

---

Bookmark the permalink.