How Vorbis Quality Works
Most audio codecs (MP3, AAC) let you pick a target bitrate — 128 kbps, 256 kbps, 320 kbps. The encoder then decides what to keep and what to discard to hit that number. Vorbis takes a different approach: you set a quality level, and the encoder determines how many bits each audio frame needs based on its complexity.
In FFmpeg and most encoding tools, the Vorbis quality parameter ranges from 0 to 10 (some tools allow fractional values and even negative values down to -1). Higher quality means:
- More spectral detail is preserved
- More bits are allocated per frame
- Larger output files
- Closer to the original lossless source
Because Vorbis is inherently VBR, the bitrate fluctuates constantly. A Q5 file might encode a silent passage at 60 kbps and a complex orchestral climax at 250 kbps. The numbers below are averages for typical stereo music.
FFmpeg syntax: ffmpeg -i input.wav -c:a libvorbis -q:a 5 output.ogg — the -q:a parameter sets the quality level (0–10).
Quality-to-Bitrate Mapping
The following table shows average bitrates for stereo 44.1 kHz audio. Actual bitrates vary by content — complex music averages higher, speech and simple audio average lower.
| Quality Level | Avg Bitrate (stereo) | Perceptual Quality | MP3 Equivalent |
|---|---|---|---|
| Q-1 | ~45 kbps | Very low — heavy artifacts | ~64 kbps CBR |
| Q0 | ~64 kbps | Low — noticeable artifacts | ~80–96 kbps CBR |
| Q1 | ~80 kbps | Acceptable for speech | ~96–112 kbps CBR |
| Q2 | ~96 kbps | Good for speech, fair for music | ~112–128 kbps CBR |
| Q3 | ~112 kbps | Good — minor artifacts on complex audio | ~128–160 kbps CBR |
| Q4 | ~128 kbps | Good — Spotify Free tier equivalent | ~160 kbps CBR |
| Q5 | ~160 kbps | Very good — transparent for most listeners | ~192 kbps CBR / VBR V2 |
| Q6 | ~192 kbps | Excellent — nearly transparent | ~224 kbps CBR / VBR V1 |
| Q7 | ~224 kbps | Transparent — indistinguishable from source | ~256 kbps CBR / VBR V0 |
| Q8 | ~256 kbps | Transparent | ~280–320 kbps CBR |
| Q9 | ~320 kbps | Transparent — Spotify Premium tier | ~320 kbps CBR |
| Q10 | ~500 kbps | Maximum — overkill for listening | No MP3 equivalent |
Recommended Settings by Use Case
The "right" quality level depends entirely on what you are encoding and where it will be used:
| Use Case | Recommended Quality | Channels | Why |
|---|---|---|---|
| Speech / podcasts | Q2–Q3 | Mono | Speech has simple spectral content; Q2 mono (~48 kbps) is transparent |
| General music | Q5 | Stereo | ~160 kbps — transparent for most listeners, good size/quality balance |
| High-quality archiving | Q7 | Stereo | ~224 kbps — indistinguishable from source, reasonable file size |
| Game SFX | Q1–Q3 | Mono | Short effects tolerate more compression; mono saves 50% vs stereo |
| Game background music | Q4–Q5 | Stereo | Balances download size vs music quality during gameplay |
| Web audio streaming | Q4 | Stereo | ~128 kbps — fast loading, good quality for background audio |
| Maximum quality | Q8–Q10 | Stereo | Diminishing returns above Q7 — use FLAC if you need true lossless |
How Spotify Maps to the Quality Scale
Spotify is the world’s largest user of OGG Vorbis. Understanding their quality tiers helps contextualize the scale:
- Low (Free mobile): ~96 kbps — roughly Q2. Audible compression on complex music, but clear for speech and simple tracks.
- Normal (Free desktop): ~160 kbps — roughly Q5. The sweet spot that sounds good to most listeners on most equipment.
- High quality: ~160 kbps — same as Normal, available to all users.
- Very High (Premium): ~320 kbps — roughly Q9. Transparent quality on any equipment.
If Q5 is good enough for 600+ million Spotify listeners, it is a solid default for most applications. Premium’s Q9 is essentially a "no compromise" setting where file size is irrelevant because Spotify handles the bandwidth.
Tip: when converting OGG to MP3, match or slightly exceed the OGG quality level. A Q5 OGG file (~160 kbps) converts well to MP3 VBR V2 (~190 kbps). See our OGG to MP3 bitrate guide for detailed mapping.
The Transparency Threshold
Transparency in audio compression means the compressed file is perceptually indistinguishable from the lossless original in blind listening tests. For OGG Vorbis, the Hydrogenaudio community — the most rigorous audio testing community online — has established clear benchmarks:
- Q5 (~160 kbps): transparent for the majority of listeners on typical equipment. Artifacts are only detectable on very specific "killer samples" chosen specifically to expose codec weaknesses.
- Q7 (~224 kbps): transparent for virtually all listeners, including trained audio engineers on reference-grade equipment. Hydrogenaudio consensus: "almost nobody hears above Q7."
- Q8–Q10: beyond human perception. These settings exist for users who want mathematical certainty, but they offer no audible benefit over Q7.
The practical takeaway: Q5 is the floor for quality-conscious listening, Q7 is the ceiling of audible improvement. Anything above Q7 is wasting storage space without any perceptual gain.
File Size per Minute at Each Quality Level
The following table shows approximate file sizes for one minute of stereo 44.1 kHz audio at each Vorbis quality level:
| Quality | Avg Bitrate | Size / Minute | Size / 4-min Song | Size / 60-min Album |
|---|---|---|---|---|
| Q0 | ~64 kbps | 0.48 MB | 1.9 MB | 28.8 MB |
| Q1 | ~80 kbps | 0.60 MB | 2.4 MB | 36.0 MB |
| Q2 | ~96 kbps | 0.72 MB | 2.9 MB | 43.2 MB |
| Q3 | ~112 kbps | 0.84 MB | 3.4 MB | 50.4 MB |
| Q4 | ~128 kbps | 0.96 MB | 3.8 MB | 57.6 MB |
| Q5 | ~160 kbps | 1.20 MB | 4.8 MB | 72.0 MB |
| Q6 | ~192 kbps | 1.44 MB | 5.8 MB | 86.4 MB |
| Q7 | ~224 kbps | 1.68 MB | 6.7 MB | 100.8 MB |
| Q8 | ~256 kbps | 1.92 MB | 7.7 MB | 115.2 MB |
| Q9 | ~320 kbps | 2.40 MB | 9.6 MB | 144.0 MB |
| Q10 | ~500 kbps | 3.75 MB | 15.0 MB | 225.0 MB |
| WAV (ref) | 1,411 kbps | 10.1 MB | 40.3 MB | 605 MB |
At Q5, a full 60-minute album takes just 72 MB — roughly 12% of the uncompressed WAV size. Moving to Q7 increases that to ~101 MB (17% of WAV). The jump from Q7 to Q10 more than doubles the file size (101 MB to 225 MB) with zero audible benefit for most listeners.
For game developers: using Q2 mono for sound effects instead of Q5 stereo cuts file size by roughly 75% (0.36 MB/min vs 1.20 MB/min). For a game with 30 minutes of SFX, that is over 25 MB saved from the download.