Audio cleanup: noise, wind, levels and loudness

Eddie can clean up the sound under a voice. It can also fix audio levels. Here is what each one does.

Remove noise, wind or music from under a voice

Ask Eddie to clean up the sound of a clip. Eddie keeps the voice and removes everything else: wind, hum, hiss, room noise, traffic, crowd noise and music that plays under someone talking.

"There is wind on the second interview. Clean up the audio on that clip."

"The music under the guest is too loud. Take it out of the first three clips."

Other ways to ask: "get rid of the hum", "remove the background noise", "isolate the voice".

• Eddie cleans the whole source file, once. It does not clean only the part you named. Eddie keeps the cleaned file. Use it on more clips from the same file, or in another edit, and it costs nothing.
• It runs in the background. An hour of audio takes a few minutes. Until the cleaned sound is ready, the clip plays its original sound. Eddie posts in the chat when it is ready.
• It is fully reversible. Your original file is never changed. Ask Eddie to put the original sound back at any time. This is free.
• Multicam cuts. If the cut plays a mix of several microphones, Eddie cleans each microphone and bills each one by its own minutes. A microphone that cannot be cleaned stays in the mix with its original sound.

What it costs

• 18 credits per minute of the source file. Eddie bills the whole file, not only the clip, with a minimum of 1 second. A 10-minute interview file costs 180 credits.
• You pay once per file. The charge is taken up front. If the clean-up fails, the credits come back automatically.
• Reusing a cleaned file is free. Turning the clean-up off is free.
• Eddie quotes the price before it runs. Ask "how much would it cost to clean up the audio?" and Eddie gives the files, the minutes and the credits, and changes nothing. See the price list.

Where the cleaned sound carries

• It carries to the player in the editor, and to every cloud MP4 render. An MP4 made by the desktop app also uses it when the app can download the cleaned file. If it cannot, the export says so.
• It does not carry to Premiere Pro, Final Cut Pro, DaVinci Resolve, Avid and OTIO exports, including the .prproj and .drp files. Those exports point to your original recordings. Clean the sound in your editing app if you need it there. Final Cut Pro and Resolve have voice isolation tools of their own. See Export to Premiere, Final Cut, Resolve, Avid & OTIO.

What it cannot do

• It keeps voices and removes the rest. It is the wrong tool for a music performance, or for room sound or effects that you want to keep.
• It does not repair distortion or clipping. A sound that was recorded too loud stays too loud.
• Echo and reverb removal, EQ, de-essing and click removal. Eddie has no tool for these.

What Eddie can do with levels

• Even out the dialogue. Auto-level speech (−16 LUFS) brings each source's speech to the same level before the mix, so a quiet speaker and a loud one sound even. Tick it in the Export menu, or ask "level the dialogue". It applies to MP4 renders, and Premiere Pro, DaVinci Resolve and Final Cut Pro exports get it as clip gain.
• Set the final loudness of an MP4. Final loudness sets the level of the whole mix: Web & social (−14 LUFS), the default, Broadcast EBU R128 (−23 LUFS), Broadcast US ATSC A/85 (−24 LKFS), or Off. It applies to cloud MP4 renders only.
• Change one clip. "The third clip is too loud" or "mute the drone shot". A level goes from −24 dB to +12 dB, and changing it is free. See Make one clip louder, quieter or silent.
• Use better sound from another file. If a lav or recorder file is synced with the clip, Eddie can take the clip's sound from it. See Make one clip louder, quieter or silent.
• Keep music under speech. Music ducks under dialogue automatically. See Music, voice-over and audio.
• Mute a track. Mute the clips' own sound, the voice-over, the sound effects or the music.
• Smooth the cuts. Ask for an audio crossfade at every cut. It is baked into cloud MP4 renders. The player does not preview it.

Good to know

• A level or loudness setting applies to the next render or export. Export again to hear it.
• Final loudness does not change a timeline export. Auto-level speech does.