What format does Zoom save recordings in?

Local Zoom recordings save as MP4 (H.264 video plus AAC audio at ~128 kbps stereo) and a parallel M4A (AAC audio only) in your Zoom folder — usually ~/Documents/Zoom on Mac and C:\Users\[user]\Documents\Zoom on Windows. Cloud recordings are available as MP4 and M4A through the Zoom web portal, plus VTT transcripts and chat logs. Hosts who enable separate audio tracks also get one M4A file per participant containing only that person's microphone, useful for podcast and interview editing.

How do I convert Zoom recordings to MP3?

The simplest path is the browser-based [MP4 to MP3](/convert/mp4-to-mp3) tool — drop the Zoom MP4 or M4A in, pick a bitrate (192 kbps for general use, 64 kbps mono for transcription), and download the MP3. Conversion runs locally in the tab; no upload, no account. Command-line equivalent: 'ffmpeg -i zoom_0.mp4 -vn -b:a 192k zoom.mp3'. The '-vn' flag drops the video stream so only audio is encoded. Conversion takes seconds for a typical hour-long meeting.

Why does my Zoom recording sound low quality?

Zoom transmits audio as Opus at network-adapted bitrates (often 16–32 kbps mono in speech mode) and then transcodes to AAC at 128 kbps stereo when it writes the recording. The quality floor is set during the call, not during the save — converting the file afterward cannot recover detail that was never captured. To improve future recordings, enable Original Sound for musicians (Settings > Audio > Advanced), use a USB or XLR microphone instead of laptop built-ins, and turn on stereo audio if multiple sources are involved.

Should I record locally or to the cloud for best quality?

Local recording usually wins on quality because the audio is encoded once on your machine after the call's Opus stream arrives, while cloud recording adds a server-side processing step that sometimes re-compresses audio more aggressively. Local recording also gives you the file immediately rather than waiting for cloud processing. The trade-off is storage: a 60-minute meeting eats ~250 MB of local disk. Cloud recordings are searchable, transcribable in-portal, and shareable via link — useful for collaboration even at slightly lower quality.

How do I get separate audio files for each Zoom participant?

Enable Settings > Recording > 'Record a separate audio file of each participant' before starting the meeting. This is a host-only setting and only applies to local recordings. After the meeting ends, Zoom writes one M4A per participant alongside the main MP4, named 'audio[participantName].m4a'. Each file contains that participant's microphone signal only — no other voices, no system audio. This is the standard workflow for podcast and interview production, since each track can be processed and mixed independently in a DAW.

Audio Formats for Zoom: Recordings, Uploads, and Sharing

Zoom generates audio in three distinct contexts that each have their own format quirks: local recordings written to your hard drive, cloud recordings stored on Zoom's servers, and separate per-participant audio tracks for hosts who enabled that option. Then there is the question of how to share, transcribe, or edit those recordings — which depends on file size and what the destination tool actually accepts. This guide covers all three layers.

What Zoom Writes to Disk

Local recordings land in your Zoom folder by default — '~/Documents/Zoom/[meeting name]' on macOS, 'C:\Users\[username]\Documents\Zoom\[meeting name]' on Windows, '~/Zoom/[meeting name]' on Linux. Inside that folder, depending on your Zoom recording settings, you get up to five files:

zoom_0.mp4 — H.264 video plus AAC audio at 44.1 kHz stereo, ~128 kbps. The full meeting recording.
audio_only.m4a — AAC audio only, same bitrate, same sample rate. Roughly 1 MB per minute.
chat.txt — chat transcript, plain text.
playback.m3u — playlist file linking the video and audio for some players.
closed_caption.vtt — WebVTT captions if live transcription was enabled (added in Zoom 5.10+).

The MP4 is the canonical recording. The M4A is a convenience copy of the same audio track — Zoom extracts it during post-processing, not during the call. If you rename or move the meeting folder before Zoom finishes converting, the M4A may not appear; let the conversion progress bar finish first.

Cloud recordings live in your Zoom web portal under Recordings > Cloud Recordings. They include MP4 (video plus audio), M4A (audio only), VTT transcripts, chat log, and a JSON file with timestamps. Cloud recordings retain for 30 days by default on paid plans before auto-deletion. You can download all variants from the portal.

Separate audio tracks (Settings > Recording > "Record a separate audio file of each participant") writes one M4A per participant, named 'audio[participantName].m4a'. Each file contains only that participant's microphone — no other voices, no system audio. Critical for podcast and interview workflows where you want to edit each speaker independently.

Audio Quality Inside the Files

Zoom transmits audio over the call using Opus at 16–48 kHz, adapted to network conditions. When it writes the recording, it transcodes from Opus to AAC at 44.1 kHz stereo, ~128 kbps. So the audio inside your local MP4 has been through two lossy codecs: Opus on the wire, AAC on disk. That is why Zoom recordings often sound thin compared to a direct microphone capture — the floor is set by Opus's network-adapted bitrate during the call.

Two settings improve the floor:

Original Sound (Settings > Audio > Advanced > "Show in-meeting option to enable Original Sound for musicians"). When enabled and toggled on during the call, Zoom disables noise suppression, echo cancellation, and auto-gain. The Opus stream goes from 32 kbps mono speech-mode to ~96 kbps stereo music-mode. This is what musicians and podcasters use.
Stereo audio (Settings > Audio > Advanced > Stereo audio). Off by default. Required if any participant is using a stereo source (live music, ambient mics).

Converting Zoom Recordings

The right output depends on the destination:

Transcription (Otter.ai, Rev, Descript, Whisper). All accept MP4 and M4A directly — no conversion needed. If a tool insists on WAV or MP3, use MP4 to WAV or MP4 to MP3 (browser-based, no upload).
Podcast editing in a DAW. Per-participant M4A files import directly into Logic, Pro Tools, Reaper, Audition, and Audacity (Audacity needs the FFmpeg helper library for native AAC import). Converting to WAV first is unnecessary unless your DAW chokes on AAC.
Sharing the audio outside Zoom. Convert the M4A to MP3 with M4A to MP3 for maximum compatibility — email, Slack, podcast hosts, voicemail-style sharing all handle MP3 universally. The 1 MB/min M4A also works almost everywhere modern, but MP3 has zero edge cases.
Reducing file size for upload limits. A 60-minute MP4 from Zoom is roughly 200–300 MB; the M4A is ~60 MB. For Gmail's 25 MB attachment limit, transcode the M4A to MP3 at 64 kbps mono — a 60-minute meeting drops to ~28 MB.

Audio-Only Meeting Recording

If you only need the audio (interview shows, podcasts, transcription pipelines), turn off video recording entirely: Settings > Recording > uncheck "Record video during screen sharing" and "Optimize for 3rd party video editor." The MP4 is still produced but contains only audio. This shrinks file size by ~80% and removes a transcoding step if your final destination is audio.

Sharing Recording Files

Slack: M4A plays inline in the Slack desktop and web clients, MP3 plays everywhere. The Slack file size limit (1 GB on paid, much smaller on free) is generous for typical recordings.
Email: Compress the audio to MP3 64 kbps mono first or share via a link from Google Drive/Dropbox. Most consumer email caps at 25 MB.
Podcast hosts (Anchor, Buzzsprout, Libsyn). All accept MP3. Some accept M4A and WAV. Convert with audio bitrate guide by use case recommendations — typically MP3 96 kbps mono for spoken-word episodes.
Public sharing. Upload the MP4 to YouTube as unlisted for free hosting; pull the audio later if needed.

Transcription Workflow

The cleanest pipeline is M4A or MP3 fed straight into Whisper, Otter, or Descript. Speech recognition models are largely format-agnostic; they decode internally to PCM regardless. The exception is some legacy services that have parsing bugs around AAC SBR (used in low-bitrate AAC variants, sometimes triggered by Zoom's encoder). If a transcription service produces obviously bad output from a Zoom M4A but works on other audio, run the M4A through MP4 to WAV first — the underlying audio is unchanged, but PCM WAV bypasses any AAC parsing edge cases.

Storage and Long-Term Archive

Zoom recordings accumulate fast. A team running 5 hours of Zoom calls per week generates roughly 1.5 GB of MP4 plus 300 MB of M4A annually if you keep everything raw. Practical retention strategies:

Keep MP3 mono only. Convert each meeting's M4A to 64–96 kbps mono MP3 and delete the original MP4. Annual storage drops to ~150 MB.
Keep audio_only.m4a, delete MP4. If you might want video later, this is wasteful, but for audio-first archive it preserves Zoom's native AAC quality at small size.
Selective retention. Delete daily standups and routine syncs; keep all-hands, customer calls, and decision-making meetings. Saves 80% of storage with negligible loss.
Drive sync only the audio_only.m4a files. Use a Drive sync rule that includes M4A and excludes MP4 from the Zoom folder. Cheap cloud archive, expensive sync skipped.

For the dedicated MP3 conversion workflow see how to convert Zoom recording to MP3 and the parallel Google Meet workflow at how to convert Google Meet recording to MP3.