Choosing MP3 Settings for Music, Speech, Lessons and Podcasts
96 kbps mono for speech. 128 kbps stereo when music is part of the mix. 192 kbps for music you sit down and listen to. Keep the sample rate at 44.1 kHz in all three cases. That settles most files. The rest of this page covers the situations where those numbers are wrong, and the one case where picking a higher bitrate does nothing at all.
Only three settings change the result
Bitrate is how much data is transmitted per second of audio, measured in kbps. It is the main control over both quality and size.
Channels means mono (one) or stereo (two). At the same bitrate, mono gives its single channel the full data budget, so a solo voice sounds cleaner in mono than in stereo.
Sample rate is how often the waveform was measured during recording. 44.1 kHz is the safe answer everywhere. Changing it is not a useful way to shrink a file.
Bit depth (16-bit, 24-bit) matters while you record and edit. It disappears once you export to MP3.
Settings by use case
| Use case | Bitrate | Mode | Channels | 1 hour |
|---|---|---|---|---|
| Voice memo, meeting | 64 kbps | CBR | Mono | ~28 MB |
| Lesson, lecture, tutorial | 64–96 kbps | CBR | Mono | 28–42 MB |
| Podcast, speech only | 96 kbps | CBR | Mono | ~42 MB |
| Podcast with music | 128 kbps | CBR | Stereo | ~56 MB |
| Audiobook, personal | 64 kbps | CBR | Mono | ~28 MB |
| Music, everyday | 192 kbps | VBR or CBR | Stereo | ~84 MB |
| Music, keeping long term | 320 kbps or VBR V0 | Either | Stereo | 108–141 MB |
Sizes are calculated with MB = kbps × seconds ÷ 8 ÷ 1024. Real files run slightly larger once cover art and ID3 tags are added.
Music
Music exposes compression faster than voice does. Cymbals, reverb tails and busy choruses are where a low bitrate starts to sound splashy or thin.
Start at 192 kbps stereo. Move to 256 or 320 kbps if you are keeping the file rather than sharing it. Variable bitrate at LAME’s V0 preset averages around 245 kbps and produces a smaller file than 320 kbps at a quality most listeners cannot separate in blind tests. Never convert music to mono to save space. The saving is real, the result sounds flat.
Speech and voice recordings
Speech sits in a narrow band, roughly 85 Hz to 8 kHz. It does not need much data. 64 kbps mono handles meeting recordings and voice memos. If a voice sounds metallic or tiring after a few minutes, step up to 96 kbps mono rather than switching to stereo.
Lessons and lectures
A 90-minute class at 64 kbps mono lands near 43 MB. The same recording at 128 kbps stereo is close to 86 MB, and learning platforms are where that difference bites.
Moodle ships with an upload cap that administrators set, commonly between 128 MB and 1 GB, and the limit is applied per course as well as site-wide (see the Moodle file upload docs). Before you export, check the cap on your own course, then work backwards to a bitrate that fits.
Mono is the right default here. A lecture captured on one microphone has nothing in the stereo field worth keeping, and students listening on a phone speaker will not notice its absence.
Podcasts
Apple publishes recommended ranges of 64–128 kbps for mono and 128–256 kbps for stereo at 44.1 or 48 kHz (Apple Podcasts audio requirements). Simplecast asks for 128 kbps or lower so the file is not re-encoded on upload, and lists 64 kbps mono and 96 kbps mono as its two standard options for spoken word (Simplecast encoding tips).
Use constant bitrate. Several hosts advise against uploading variable bitrate MP3s because some players handle seeking poorly in those files, and a listener scrubbing through an hour-long episode is the person who notices. Set loudness to −16 LUFS for stereo or −19 LUFS for mono, which sound equally loud to a listener.
The one case where a higher bitrate does nothing
Audio pulled from YouTube is already compressed before it ever reaches you. The best stream most people can access is Opus at roughly 135 to 160 kbps, or AAC at about 127 kbps. That ceiling is set on Google’s servers, not on your computer.
Exporting that stream to a 320 kbps MP3 does not add the missing detail back. Compare with our full 128kbps vs 320kbps breakdown. It produces a file more than twice the size carrying the same information, plus a second round of compression artefacts from re-encoding one lossy format into another. You pay in megabytes and get nothing in return.
So for YouTube-sourced audio, treat 192 kbps as a sensible ceiling for music and 96 kbps mono as plenty for a talk or a lecture. If a converter advertises 320 kbps, that number describes the container, not the sound — a YouTube to MP3 converter that matches the source bitrate hands you the same audio in half the space.
The same rule applies to any re-encode. Raising the bitrate of an existing MP3 inflates the file and recovers nothing that was already thrown away.
Making a file fit a size limit
Flip the formula: max kbps = MB × 8 × 1024 ÷ seconds.
| Destination | Limit |
|---|---|
| WhatsApp audio | 16 MB |
| Gmail attachment | 25 MB to send |
| Outlook.com | 20 MB |
| Telegram | 2 GB |
A 45-minute recording at 64 kbps mono is about 21 MB, which clears Gmail but not WhatsApp. Trim silence first, switch to mono second, and lower the bitrate last.
Five settings to avoid
- 320 kbps on a voice memo. You are storing silence at high fidelity.
- Variable bitrate on a podcast episode, for the reasons above.
- Mono on music, unless the source really is mono.
- Re-exporting an MP3 at a higher bitrate.
- Deleting the WAV or FLAC master after a successful MP3 export. Keep the original for editing.
Checking what you already have
On Windows, right-click the file and open the Details tab. On macOS, use Get Info or open the file in Music. VLC shows it under Tools, then Codec Information. For a precise reading, ffprobe file.mp3 prints bitrate, sample rate, and channel count in one line.
Short answers
Is 128 kbps enough?
For speech, yes. For music on good headphones, it is the point where cymbals begin to sound watery.
Mono or stereo?
Mono for one voice. Stereo for music, sound design, or anything where left and right differ.
Why did my file get bigger after converting?
The source was already compressed, and you exported at a higher bitrate than it contained.
Does 48 kHz sound better than 44.1 kHz?
Not for delivery. 44.1 kHz is accepted everywhere and the difference is not the reason a file sounds good or bad.
Sources: Apple Podcasts audio requirements · Simplecast audio encoding guide · Moodle file upload documentation · Audacity MP3 export manual.
Method: File sizes are calculated at a constant bitrate, not measured from one specific recording. Platform limits were checked in August 2026 and may change without notice.
