音声ファイルをオンラインで圧縮する3つの手順
Shrinking your audio file size does not require installing bulky desktop software or registering for expensive subscriptions. Our browser-based audio file compressor allows you to re-encode high-resolution files into lightweight, universally compatible MP3s in seconds.
音声ファイルを読み込む
Drag and drop any MP3, WAV, M4A, FLAC, AAC, or OGG file directly into the intake box. Your file is decoded instantly in browser memory without sending a single byte to an external cloud server.
品質プリセットまたはビットレートを選択
Pick one of five curated bitrate presets ranging from 32 kbps to 192 kbps. Review the live size preview instantly to verify the reduction percentage before starting compression.
即座に圧縮してダウンロード保存
Click "圧縮を実行 [Bitrate] kbps" to run our client-side LAME MP3 Turbo engine. Audition the compressed file in the preview player and save your optimized audio with one click.
音声圧縮の仕組み：音響心理マスキング技術の解説
To understand how an audio file compressor achieves up to 90% file size reductions without destroying audio intelligibility, one must examine the difference between uncompressed linear Pulse Code Modulation (PCM) and psychoacoustic lossy encoding.
Uncompressed formats such as 16-bit 48 kHz WAV record 48,000 discrete amplitude snapshots per second across two stereo channels. This yields a massive bitstream of 1,536 kilobits per second (kbps), consuming approximately 11.5 megabytes of storage for every single minute of sound. When you run audio through an intelligent compressor audio engine, it applies mathematical models based on human psychoacoustics (known as Fletcher-Munson equal-loudness contours):
- Simultaneous Frequency Masking: When a loud sound occurs at a specific frequency (such as a deep bass drum kick), human ears cannot perceive quieter sounds occurring simultaneously in adjacent frequency bands. The encoder eliminates these imperceptible bits entirely.
- Temporal Masking: For a few milliseconds immediately before and after an explosive transient sound (such as a snare hit), human hearing is temporarily desensitized. The compressor reduces quantization depth during these microscopic time slices.
- Stereo Phase Optimization: In stereo mode, identical acoustic energy shared between the left and right speakers is mathematically encoded as a sum-and-difference matrix (Mid/Side stereo), eliminating redundant spatial data.
ファイルサイズ圧縮と音楽制作におけるダイナミクス圧縮の違い
A frequent point of confusion among podcasters, sound designers, and musicians is the distinction between file size compression (data encoding) and selective compression music engineering (dynamic range control). While both terms share the word "compression", they address completely different acoustic challenges.
Audio File Size Compression: This process reduces the digital footprint (megabytes) required to store or transmit an audio stream. By re-encoding a 100 MB WAV master to a 128 kbps MP3, you produce a 9 MB file that downloads 10 times faster, streams seamlessly across cellular networks, and fits easily as an email attachment.
Selective Compression in Music Production: In a recording studio or digital audio workstation (DAW), selective compression refers to multiband or sidechain dynamic range processing. Audio engineers apply selective compressors to specific frequency ranges—such as taming 3 kHz harshness on a lead vocal without dampening low-end warmth, or sidechaining bass synth frequencies to duck only when the kick drum strikes. Selective compression alters volume dynamics and tonal punch; it does not reduce file size.
Our online audio tool specializes in high-fidelity audio file compression, giving you full control over target bitrates, channel downmixing, and sample rate decimation so your tracks load lightning-fast everywhere.
Recommended Bitrate & Quality Settings Guide
Selecting the optimal bitrate ensures you achieve the greatest file size savings without introducing noticeable acoustic artifacts like swishing or metallic high-end distortion:
| Preset Bitrate | File Size per Minute | Avg. Savings vs WAV | Recommended Use Case |
|---|---|---|---|
| 32 kbps (Smallest) | ~0.24 MB / min | 98% Smaller | Long voice notes, low-bandwidth radio logs, and strict size restrictions. |
| 64 kbps (Voice) | ~0.48 MB / min | 96% Smaller | Speech recordings, audiobooks, lectures, and WhatsApp audio sharing. |
| 96 kbps (Podcast) | ~0.72 MB / min | 94% Smaller | Spoken podcasts with intro music, interview archives, and web demos. |
| 128 kbps (Balanced) | ~0.96 MB / min | 92% Smaller | The universal web standard. Excellent balance of voice clarity and music fidelity. |
| 192 kbps (高音質 (192 kbps)) | ~1.44 MB / min | 88% Smaller | Music tracks, broadcast mixes, sound design, and critical listening. |
| 320 kbps (Max MP3) | ~2.40 MB / min | 80% Smaller | Maximum possible lossy MP3 fidelity. Suitable for DJ sets and high-end stereo streaming. |
Common Real-World Audio Compression Use Cases
Email & Discord Attachment Limits
Gmail, Outlook, and Discord enforce strict 25 MB file size caps. Uncompressed audio from smartphones and Zoom calls easily exceeds this limit. By compressing your recording to 128 kbps or downmixing to mono, you can fit up to 50 minutes of crisp dialogue into a single email attachment.
Podcast RSS Feed Bandwidth Savings
Every additional megabyte in your podcast audio multiplies your server hosting bills when thousands of listeners stream your episodes. Re-encoding spoken audio at 96 kbps or 128 kbps cuts bandwidth consumption by over 60% with zero perceptible loss in vocal presence.
Whisper AI & Speech Transcription Prep
Automatic Speech Recognition (ASR) engines such as OpenAI Whisper and Otter operate natively on 16,000 Hz mono audio. Uploading multi-hundred-megabyte stereo files slows down transcription pipelines. Use our 詳細設定 to export a compact 16 kHz mono file for lightning-fast transcript processing.
Freeing Device Storage on Mobile
Voice memos and uncompressed DAW session bounces can quickly monopolize gigabytes of internal smartphone or laptop storage. Batch compressing archived voice notes to 64 kbps reclaims over 90% of your disk space while preserving every spoken word.
よくある質問（FAQ）
No. All audio decoding, sample rate resampling, channel matrixing, and MP3 compression occur 100% locally inside your web browser via the Web Audio API and client-side WebAssembly encoders. Your recordings never leave your computer or phone, ensuring strict confidentiality for sensitive voice notes, corporate meetings, and unreleased master tracks.
An audio file compressor lowers the bitrate (the amount of data processed per second) using psychoacoustic perceptual algorithms. It strips away imperceptible high frequencies and sound details masked by louder tones, shrinking file size by 50% to 90% while preserving perceived vocal and musical clarity.
Audio file compression reduces digital byte size and storage requirements through lossy encoding (e.g. MP3 at 128 kbps). In contrast, selective compression in music production refers to dynamic range compression applied to specific frequency bands or instruments (such as multiband sidechain compression) to smooth volume dynamics without reducing file size.
For speech, voice memos, and podcasts, 64 kbps to 96 kbps provides crystal-clear dialogue at tiny file sizes. For music, commercial podcasts, and complex acoustic recordings, 128 kbps to 192 kbps is recommended to maintain transparent high-frequency air and stereo imaging.
To fit under 25 MB, choose 128 kbps for recordings up to 25 minutes long, or 64 kbps for recordings up to 50 minutes long. You can also downmix stereo tracks to mono in our 詳細設定 drawer, cutting data bandwidth in half.
At 128 kbps to 192 kbps, human psychoacoustic masking ensures that the difference between uncompressed WAV and compressed MP3 is virtually imperceptible to standard headphones and smartphone speakers, while saving roughly 85% to 90% of file size.
If your input file was already encoded at a very low bitrate (such as 32 kbps or 64 kbps) and you selected a higher bitrate like 192 kbps, re-encoding creates larger MP3 frames without restoring the original acoustic frequencies lost in the first compression cycle. Always select a target bitrate lower than or equal to the source file.
Explore Other Free In-Browser Audio Tools
Automatically detect and strip dead air, pauses, and silent intervals from spoken recordings.
Audio TrimmingCut, slice, and trim audio clips with millisecond precision and instant local export.
Tempo AdjustmentSpeed up or slow down recordings from 0.25x to 4.0x with pitch-preservation DSP.
Multi-Track JoinerCombine multiple audio tracks into a seamless file with drag-and-drop reordering.
Format ConversionConvert audio between MP3, WAV, M4A, FLAC, and OGG formats with custom encoding bitrates.