Recording lectures, podcasts and interviews on a Mac
Updated 2026-09-03.
Recording lectures, podcasts and interviews on a Mac comes down to one question: is the sound in the room or inside the Mac? Voices in the room are microphone input, which Voice Memos, QuickTime Player and any recorder capture for free. Voices in a stream or a call are playback, which only a system audio recorder built on ScreenCaptureKit can capture, and such a capture never includes your own microphone. After recording, MacWhisper or Voice Memos turn the file into text. One hour of speech is 144 MB as MP3 320.
Guides in this topic
- How to record a lecture on a MacIn the room, use Voice Memos or QuickTime with the mic. Online, capture system audio with ScreenCaptureKit. A 90-minute lecture is 216 MB as MP3 320.
- How to record a podcast on a MacSolo: QuickTime or GarageBand at 24-bit. Two people: an interface and separate tracks. Remote: a double-ender. A 45-minute episode is 743 MB at 24-bit.
- How to record an interview on a MacIn person, record a mic or interface at 24-bit. Remote, system audio captures only the other side, so use two takes or the call app's recorder.
- How to record an online course or training on a MacCheck the platform terms first: Udemy and Coursera license viewing only. Then record audio via ScreenCaptureKit or video with OBS. An hour of MP3: 144 MB.
- How to transcribe audio on a Mac for freeVoice Memos and Notes transcribe on macOS 15+ with Apple silicon. MacWhisper and Whisper transcribe any WAV or MP3 offline. Dictation stops after 30 s.
- How to record audio for AI meeting notes on a MacRecord locally, transcribe on the Mac, paste the text. ChatGPT (512 MB) and Claude (500 MB) list document uploads, not audio; Otter takes audio up to 5 GB.
- How to record a language lesson on a Macitalki records with both sides' consent to a local WEBM. Preply keeps lesson audio to download. Elsewhere, system audio gets the tutor, the mic gets you.
- How to record yourself on a MacVideo: QuickTime's New Movie Recording or Photo Booth. Voice only: Voice Memos, QuickTime's New Audio Recording at 24-bit, or AirCheck. Pick a wired mic.
- How to record a remote podcast guest on a MacA double-ender (Riverside, Zencastr) records each side locally in separate tracks. Capturing the call gives compressed guest audio. Run both as a backup.
How do I know whether I need a microphone recorder or a system audio recorder?
Ask where the sound is. If a person is speaking in the room, a microphone hears them and a mic recorder does the job: Voice Memos, QuickTime's File > New Audio Recording at 24-bit, GarageBand, or AirCheck with a mic or interface as the source. That covers a lecture in the hall, a podcast at a desk, and recording yourself.
If the sound comes out of the Mac's speakers, it is playback. QuickTime and Voice Memos cannot hear it. On macOS 14 and newer, Apple's ScreenCaptureKit lets an app such as AirCheck or OBS capture it with a one-time Screen Recording permission and no virtual driver. That covers an online course, a streamed lecture, and the far side of any call.
The rule that follows: system audio capture never contains your own microphone, because your voice does not play through your speakers. A conversation, such as a remote interview or a language lesson, therefore needs either the call app's own recorder, a double-ender service, or two takes (system audio plus mic) aligned in an editor. No free built-in Mac tool does both in one file; Audio Hijack ($64) can, with live routing.
What format should I record speech in and how big is the file?
Speech does not need lossless audio unless it will be edited or published as a podcast. The measured rates from the AirCheck library on the author's Mac: 24-bit 48 kHz is 16.5 MB per minute, 16-bit 48 kHz is 11.0 MB, and MP3 320 kbps is 2.4 MB. A 90-minute lecture is therefore 216 MB as MP3, 990 MB as 16-bit WAV and 1,485 MB at 24-bit. A 10-hour course is 1.4 GB as MP3 and 9.9 GB at 24-bit.
Use MP3 320 for lectures, courses, lessons and meeting notes, where the file's only future is a transcriber. Use 24-bit WAV for podcast takes and interviews that will be published, because normalizing and de-noising benefit from the headroom. QuickTime records 24-bit at a fixed 44.1 kHz; AirCheck records at the device's own rate and converts once at stop to WAV 16-bit, WAV 24-bit or MP3 320.
Length is not a limit. There is no cap in AirCheck; the longest real take in the author's library is 2 hours 5 minutes at 24-bit, 2.07 GB. The practical limits are sleep and battery: keep a MacBook's lid open, and on power turn on "Prevent automatic sleeping on power adapter when the display is off" in System Settings > Battery > Options, as detailed in how to record a lecture on a Mac.
How do I turn the recording into text or notes?
Transcription is the second half of every guide in this cluster and the tool depends on where the audio came from. Recordings made in Voice Memos or Notes are transcribed by those apps on macOS 15 or later with Apple silicon, in Apple's supported languages and regions. Any other file, including everything a system audio recorder produces, goes to MacWhisper (free tier, offline, MP3, WAV, M4A, MOV, MP4) or to the open-source Whisper model. Online, Otter.ai imports files up to 5 GB, but its free plan transcribes only 30 minutes per file with three imports per account. The options and limits are compared in how to transcribe audio on a Mac.
For AI summaries, the file limits belong to the model's upload rules, not the recorder. ChatGPT's FAQ sets 512 MB per file and lists document types without audio; Claude lists document types with 500 MB per file in chat. So the pipeline is record, transcribe on the Mac, paste text, as in recording audio for AI meeting notes. Transcribing locally also keeps a confidential interview, a paid course you may only view, or a tutor's voice off third-party servers.
Consent runs through all of it. Platforms that build recording in, such as italki, require both sides to agree; a system audio recorder is invisible to call software, so the announcement is yours to make. Remote podcast guests expect it; interviewees and tutors deserve it.
Which recording method for which spoken-word job on a Mac
| Situation | Sound is | Free built-in tool | Captures both sides in one file | Recommended format | Transcribe with |
|---|---|---|---|---|---|
| Lecture in the room | microphone | Voice Memos, QuickTime | n/a (one source) | MP3 320 or Lossless | Voice Memos (macOS 15+, Apple silicon) |
| Lecture or course streamed online | playback | none (QuickTime is mic only) | n/a, your voice not needed | MP3 320 | MacWhisper |
| Solo podcast or narration | microphone | QuickTime Maximum, GarageBand | n/a | WAV 24-bit | MacWhisper |
| In-person interview or two-mic podcast | microphone (two channels) | GarageBand | yes, separate tracks | WAV 24-bit | MacWhisper with speaker labels |
| Remote interview, lesson or meeting | playback plus your mic | call app recorder if host allows | only the call app recorder or Audio Hijack | MP3 320 (notes) or WAV (publish) | MacWhisper or Otter (paid for over 30 min) |
| Remote podcast guest | playback, or guest's local file | none | double-ender: yes, separate tracks | WAV 24-bit | double-ender transcript or MacWhisper |
AirCheck ($39) covers every microphone row and every playback row, one source per take; it does not mix the two into one file.
AirCheck records what your Mac plays, or any input, to one file. $39 once.
macOS 14 or newer. One-off payment. Fourteen-day refund, no questions asked.
Other topics
- Screen recording with audio
- Recording internal and system audio
- Recording calls and meetings
- Recording streams, music and video sound
- Virtual audio drivers and Audio MIDI Setup
- Recording with QuickTime, OBS, Audacity and GarageBand
- Recording DJ sets, vinyl and instruments
- Audio formats, quality and file size
- Fixing Mac audio recording problems
- Choosing a Mac audio recorder