Recording setups by job
The right recording setup depends on the job: what you are capturing, and where the file goes afterwards. Each page below covers one job end to end.
Two decisions drive everything else
- Will you edit it? If yes, record WAV — every editing step works better on uncompressed source, and you encode once at the end. If no, record MP3 and keep the file small.
- Is it speech or performance? Speech benefits from the browser's noise suppression. Singing and sound recording are damaged by it, so those pages request the raw signal with it switched off.
Everything else — microphone distance, level, what to do with the file — follows from the job. That is what these pages cover.
Publishing and performance
- PodcastingSolo episodes and the double-ender technique for remote guests.
- VoiceoverClient delivery specs, and why the room matters more than the microphone.
- Voice acting demosReel length, take-based recording and handling loud character reads.
- SingingRecording with processing disabled, and working to a backing track.
Conversations and study
- InterviewsConsent on tape, two-voice setups and transcription-friendly formats.
- MeetingsAnnouncing the recording, and capturing an online call locally.
- LecturesInstitutional permission, file sizes across a term, and laptop placement.
- Language practiceHearing what you actually said, and tracking progress monthly.
Recording for another app
- Google SlidesThe Drive permission step that silences audio for everyone else.
- PowerPointEmbedded audio, deck size, and exporting to video.
- WhatsAppThe honest trade-off: review before sending, but no native voice-note bubble.
- DiscordSoundboard limits of five seconds and 512 KB.
Choosing by where the file ends up
The destination usually decides the format before anything else does.
| Where it is going | Format | Why |
|---|---|---|
| Into an audio editor | WAV | Editing compressed audio and re-exporting stacks two lossy passes |
| An email or a message | MP3 | A minute is under 1 MB rather than 5 MB |
| A slide deck | MP3 | PowerPoint embeds audio, so WAV inflates the file badly |
| A transcription service | MP3 | Accuracy depends on clarity, not on bitrate |
| A client who specified it | WAV | Almost every voiceover brief asks for uncompressed |
| Straight back to you, later | MP3 | Smaller, and quality is far beyond what review needs |
What changes between jobs
- Distance and level. A lecture recorded from row twenty is mostly room; a vocal take at 15 cm needs gain set against your loudest note, not your speaking voice.
- Whether processing helps or hurts. Noise suppression cleans up a meeting and mangles a sustained sung note.
- Whether consent is involved. Anything capturing another person carries legal obligations that vary by jurisdiction — those pages carry a notice, and the disclaimer explains its limits.
- What happens after the download. A Discord clip has a five-second cap; a Google Slides file has to live in Drive with the right sharing permission.
Before any of them
Set your level once with the volume meter and confirm the right microphone is live with the microphone test. Those two minutes prevent most of the problems that make a recording unusable.
Questions
Which format should I record in?
WAV if the file will be edited afterwards, MP3 if it is going straight to someone. MP3 vs WAV covers the trade-off in full.
Can I record two people into separate tracks?
Not from one page — this records one microphone. For two clean tracks, both people record locally and you sync afterwards with a clap. The podcast page describes the method.
Why do some pages turn off noise suppression?
Because it is tuned to remove everything that is not a speaking voice. That helps a call and damages singing, instruments and ambience, so the singing and sound recorder pages request the raw signal instead.
Do I need different settings for each job?
Each page is preconfigured for its job — default format, and whether browser processing is on. You can override the format on any of them.