128 vs 320 kbps: The Short Answer
A 320 kbps MP3 stores 2.5 times as much data per second as a 128 kbps MP3. On good headphones, many listeners can pick out 128 kbps on cymbals, applause and reverb tails. Telling 320 kbps from the original WAV is much harder, and in blind tests most people cannot.
| 128 kbps | 320 kbps | |
|---|---|---|
| Data per second | 16 KB | 40 KB |
| Size per minute | 0.96 MB | 2.4 MB |
| 4-minute song | 3.84 MB | 9.6 MB |
| Share of CD-quality WAV (1,411 kbps) | 9% | 23% |
| Typical weak spots | Cymbals, hi-hats, reverb, applause | Rarely audible |
| Good for | Speech, podcasts, casual listening | Music libraries, DJ use |
From WAV, the converter above also offers VBR V0 (about 245 kbps), a common near-transparent setting that uses less data than 320 kbps CBR.
What Bitrate Means for Audio Quality
Bitrate is the number of bits per second used to represent the audio. Higher bitrate = more data = closer to the original WAV, but bigger files. Lower bitrate = smaller files, but the encoder has to discard more audio information.
A CD-quality WAV runs at 1,411 kbps (44,100 samples/sec × 16 bits × 2 channels). Converting to MP3 at 192 kbps means compressing that data to roughly 13.6% of the original size — and the encoder's job is to make that 13.6% sound as close to the original as possible.
CBR Bitrate Comparison: 128 vs 192 vs 256 vs 320
Here's how each Constant Bit Rate level compares for a typical 4-minute song:
| CBR Bitrate | File Size (4 min) | Compression Ratio | Quality | Best For |
|---|---|---|---|---|
| 128 kbps | 3.84 MB | 11:1 | Acceptable | Speech, podcasts, voice memos |
| 192 kbps | 5.76 MB | 7.3:1 | Good | General listening, streaming |
| 256 kbps | 7.68 MB | 5.5:1 | Very good | Music collections, critical listening |
| 320 kbps | 9.6 MB | 4.4:1 | Excellent | Archival, maximum MP3 quality |
The original WAV at CD quality takes ~42 MB for the same 4-minute song. Even at the highest MP3 bitrate (320 kbps), you save ~77% of the space.
VBR: Better Quality at Smaller Sizes
Variable Bit Rate encoding allocates bits dynamically — more for complex passages, fewer for silence and simple sections. This produces better quality at the same average file size compared to CBR.
| VBR Preset | Avg Bitrate | File Size (4 min) | Quality | Best For |
|---|---|---|---|---|
| V0 | ~245 kbps | ~7.4 MB | Transparent | Audiophile, archival |
| V2 | ~190 kbps | ~5.7 MB | Near-transparent | General music (recommended) |
| V4 | ~165 kbps | ~5.0 MB | Good | Podcasts, spoken word |
| V6 | ~130 kbps | ~3.9 MB | Acceptable | Voice recordings, low storage |
Recommendation: VBR V2 is the sweet spot for most people. It sounds indistinguishable from the original WAV for the vast majority of listeners and equipment, while producing files ~7x smaller. For a deeper comparison of VBR and CBR encoding methods, see our VBR vs CBR guide.
File Size Formula
For CBR MP3, the file size is predictable:
Quick reference for common durations:
| Duration | WAV | 128 kbps | 192 kbps | 320 kbps |
|---|---|---|---|---|
| 3 min | 31.8 MB | 2.9 MB | 4.3 MB | 7.2 MB |
| 5 min | 52.9 MB | 4.8 MB | 7.2 MB | 12 MB |
| 30 min | 318 MB | 28.8 MB | 43.2 MB | 72 MB |
| 1 hour | 635 MB | 57.6 MB | 86.4 MB | 144 MB |
How CleverUtils Encodes WAV to MP3
When you upload a WAV file, CleverUtils encodes it to MP3 using libmp3lame (LAME) — the gold-standard open-source encoder. By default, it uses VBR V2 (~190 kbps average). You can switch to CBR or adjust the quality/bitrate using the encoding options panel above.
Since WAV is uncompressed PCM audio, the encoder works with the full original signal — no generation loss from a previous lossy encoding step. This makes WAV-to-MP3 one of the cleanest possible conversion paths.
Tip: Need consistent volume across your converted files? You can also normalize loudness to a platform-specific LUFS target during conversion.
How to Hear the Difference Yourself: a Blind ABX Test
- Convert the same WAV twice: once at CBR 128 kbps and once at CBR 320 kbps.
- Install foobar2000 with the ABX Comparator component, or use any ABX tool that hides which file is playing.
- Load both files, play short hard passages (cymbals, applause, a reverb tail) and decide which one X is, at least 10 times.
- A score of 9 or 10 out of 10 means you hear a difference on that track: the chance of getting there by guessing is about 1% (11 of 1,024 outcomes). Around 5 out of 10 is guessing.
Keep it fair: both files must play at exactly the same volume, through the same player and headphones. A louder file almost always sounds “better”, even when it is not.