The whole path in five steps
You want an episode out, not a weekend spent learning audio software. The path is the same every time: record, cut, level, check, upload. Every step runs in a browser tab on your own computer, with no account and no watermark.
- Record with the voice recorder. Allow the microphone, run the mic check until it says Sounding good, and save a WAV.
- Cut the dead air in the audio editor, or in the silence remover if you want to review every cut.
- Level and master in Podcast Studio: pick -16 LUFS and save the mastered MP3.
- Check it. Listen from start to finish, then read the loudness results on the page.
- Upload the MP3 to the podcast host that makes your feed.
Record: one voice, one quiet room, one file
The voice recorder captures the raw audio from your microphone, so the WAV you download is the audio exactly as your browser delivered it, with no compression added along the way, and an MP3 is encoded only once, from that original. For an episode you will edit, save the WAV. It is lossless and runs about 5 to 6 MB per minute, so a 30-minute take is roughly 150 to 180 MB.
Run the mic check first. It flags silence, a level that is too quiet, a healthy level and clipping. Clipping cannot be repaired afterward, while a quiet recording can be lifted later. You can pause between takes, and everything ends up in one file with no time limit.
Clean up my voice (noise reduction, echo cancellation and automatic level from your browser) helps with a laptop microphone. With a good USB or XLR microphone, turn it off.
A guest in the same room can share your microphone. For a remote guest, the screen recorder can record your microphone and the sound of a browser tab mixed into one track (in Chrome and Edge, tick the audio option in the sharing dialog; on Windows it can also include the whole computer's sound). It is a single mixed track, so you cannot turn one speaker down later. Tell your guest you are recording. Record audio on your computer has the details.
Cut the dead air
Raw takes have false starts, coughs and long gaps. In the audio editor, drag across the part you want and press Keep only to trim the ends, or select a mistake and press Cut to remove it and close the gap. Each cut gets a 5-millisecond fade on both sides, so it does not click. Remove pauses closes the gaps between sentences, and one second is a good starting length for voice.
To review every cut first, open the file in the silence remover. It accepts MP3, WAV, M4A and FLAC and draws each pause as a striped block on a timeline that you can click to keep or drag to adjust. For conversation, a minimum pause of two seconds or more removes only real dead air, while 0.5 to 0.8 seconds gives the snappy rhythm of short video. Removing pauses from talking videos explains the settings. It cuts by loudness, not meaning, so it will not remove filler words such as um.
Level it to the loudness podcast apps expect
Listeners should not reach for the volume when your episode follows another one. The usual target for spoken word is about -16 LUFS integrated, with true peaks at -1 dBTP or lower. Apple's podcast audio requirements recommend -16 LKFS within 1 dB, and -14 LUFS suits a show that goes mainly to Spotify or YouTube. The reasoning is in LUFS for YouTube, Spotify and podcasts.
Podcast Studio does the leveling. Drop in the edited file and it measures the integrated loudness of the whole episode, applies one fixed gain, and uses a look-ahead limiter to catch only the peaks that would pass -1 dBTP, so the voice keeps its natural dynamics. The saved MP3 is measured again, and if it landed noticeably off target the gain is corrected and the MP3 is made once more. A mono recording is saved as stereo, because the -16 LUFS figure is meant for stereo files: a mono file measured as one channel would need about 3 dB less to sound equally loud.
Podcast StudioDrop in your edited episode, pick -16 LUFS and get a mastered MP3, with an optional transcript and chapters.Open itTo check any file without changing it, use the loudness meter. It reads integrated loudness, true peak and loudness range, and compares them with Spotify and YouTube (-14), Apple Podcasts (-16), EBU R128 (-23) and ATSC A/85 (-24).
What else the same file can make
Podcast Studio does not record, so it comes after the steps above. From the same episode it can also make a transcript (TXT, SRT and VTT, using speech recognition that runs on your device), suggested chapters for a YouTube description, and audiogram clips of up to 3 minutes. Chapter titles are just the first few words of each section, so edit them.
Transcription is the slow part, and it holds the episode in memory, about 0.25 GB per hour. On a phone or an older laptop, transcribe a very long show in parts. Mastering has no such limit.
Export formats and file size
| Format | Where you get it | Use it for | Size |
|---|---|---|---|
| WAV | Voice recorder, audio editor (16 or 24-bit) | Your edit master, or a file for someone else to mix | About 5 to 6 MB per minute from the voice recorder |
| MP3, 128 kbps | Voice recorder, audio editor | Speech-only shows | About 1 MB per minute, 29 MB for 30 minutes |
| MP3, 192 kbps | Voice recorder, audio editor | Shows with music | About 1.4 MB per minute, 43 MB for 30 minutes |
| M4A (AAC) | Audio editor (128 to 256 kbps) | Apple devices. Sounds better than MP3 at the same bitrate | About the same as MP3 at the same bitrate |
Podcast Studio saves its master as an MP3. MP3 is the format nearly every podcast app and host accepts, so check your host's upload rules before you choose M4A. Keep the WAV, because MP3 is lossy and editing an MP3 again re-compresses it.
Pre-publish checklist
- Listen start to finish once, on headphones, for clicks at the cuts, a mispronounced name or a private remark you meant to cut.
- Check the loudness. Within about 1 dB of -16 LUFS is close enough, with true peak at -1 dBTP or lower.
- Check the ends. No long silence before the first word and no abrupt stop. Fade out with nothing selected softens the last three seconds.
- Play the MP3 in a second app or on your phone, to be sure it opens outside the tab that made it.
- Fix the transcript and chapter titles if you publish them. Names, numbers and jargon are where recognition slips.
- Keep the WAV master until the episode is live.
What goes wrong
- The levels jump between you and a guest. One gain change cannot fix it. Lift the quiet stretches with Volume in the audio editor before you master.
- The silence remover clips quiet words. Lower the sensitivity or choose Relaxed breathing room, then check the timeline.
- Hiss or echo. The editing tools have no noise filter, so fix it at the source: a closer microphone, a softer room, Clean up my voice on.
A last limit: this is a one-recording-at-a-time setup with no separate tracks, music beds or effects, and it does not host your show. You still need a podcast host for the feed. If you recorded a video call, extract the audio from the video first. The edits in detail are in the free MP3 editor guide.