Media Library

What the Media Library is

A central place for all the audio your phone system uses:

• Voicemail greetings
• IVR prompts ("Press 1 for sales...")
• On-hold music
• Queue announcements
• Outbound campaign audio
• Whatever else needs to be played to a caller

Three ways to add audio: upload an existing file, generate from text-to-speech, or record from your browser or a phone.

Where it lives

Top nav → Media Library.

The list shows every audio file with columns:

• Display name
• File type (WAV / MP3 / OGG)
• File size
• Duration
• Tags
• Used by (which call flows, voicemail boxes, queues, campaigns reference this file)
• Actions (play, download, edit metadata, delete)

Search by name, filter by tag, sort any column.

Upload an existing file

Click New → Upload tab.

Drag a file onto the drop zone, or click Browse. Supported formats:

• WAV — best quality, no conversion needed
• MP3 — converted to WAV on upload
• OGG — converted to WAV on upload

Form fields:

• Display Name — auto-filled from filename, editable
• Description — optional
• Tags — comma-separated, for organization

Click Upload. The file is processed (transcoding if needed), then appears in the library.

Audio specs

For best quality:

• 16-bit PCM WAV
• 8 kHz mono (telephony standard) or 16 kHz mono (HD voice)
• No DC offset
• -3 dB peak max (avoid clipping)

Audio you grab from YouTube or other web sources is usually fine but loud — normalize before uploading or callers will be blown out.

Generate from text (TTS)

Click New → Text-to-Speech tab.

Form fields:

• Display Name
• Text — what to say (max ~5000 characters)
• Voice — pick from available voices (typically Polly voices: Joanna, Matthew, Salli, etc.)
• Language — en-US, es-ES, fr-FR, etc.
• Tags

Click Preview to hear it before saving. Adjust the text or voice as needed.

Click Generate & Save. The file is generated and appears in the library, ready to use.

Writing for TTS

A few tips:

• Use full punctuation — periods, commas, ellipses. They control pacing.
• Spell tricky words phonetically — TTS often mispronounces names and acronyms. "EE-DAH" instead of "Ada"; "S-Q-L" instead of "SQL"; "yor-NIK" instead of "Yarnyk".
• Keep sentences short — long sentences sound robotic
• Test with the actual voice — voices interpret the same text differently. Joanna might pronounce "Yarnyk" wrong while Matthew gets it right. Pick the best voice for your text.

For greetings and IVR, TTS is usually good enough. For polished marketing-style intros, consider human voice talent and upload the resulting WAV.

Record from the browser

Click New → Record tab.

Browser asks for microphone permission. Click Start Recording, speak, click Stop.

Preview by playing back. Re-record if needed. Save with a display name and tags.

Works in Chrome, Firefox, Safari (latest versions). Mobile browser support varies.

Quality depends entirely on your microphone. A laptop's built-in mic is okay for personal voicemail greetings; a USB headset is better; a dedicated condenser mic is best.

Record from a phone (call-out method)

For people who don't want to do the browser thing or who want a "studio" feel using their desk phone:

Click New → Record tab → Use Phone option.

Form:

• Target — pick an extension
• Caller ID — which DID to use for the outbound call
• Display Name for the resulting file

Click Start Recording. The platform calls the chosen extension. The user answers, follows the voice prompt: "Recording starts after the beep. Press # when done." The recording is saved to the Media Library automatically.

Editing metadata

Click any file. Update display name, description, tags. Save. Audio content can't be edited in place (re-upload to replace).

Replacing audio in place

Click a file → Replace Audio. Upload a new file. The old file is overwritten; everything that referenced this file (call flows, voicemail boxes, queues) automatically uses the new audio.

Use this for seasonal updates ("Happy Holidays" greeting that switches back in January) without changing call flow logic.

Deleting files

Click Delete on a file. If the file is in use anywhere, the platform warns you. Either:

• Remove the references first (in call flows, voicemail boxes, queues)
• Delete and live with the breakage temporarily

Organizing with tags

For a small library (< 30 files), no organization needed. For a larger one:

• Tag by use: voicemail-greeting, ivr-prompt, hold-music, campaign-audio
• Tag by team: sales, support, marketing
• Tag by campaign: q1-2026-launch

Filter by tag using the search bar.

Library quota

Your service provider sets a storage quota for media. Most tenants stay well under it (audio files are small — a 30-second WAV is ~500 KB).

If you hit the quota, the upload form warns you. Delete unused files or contact your service provider to raise the limit.

Common patterns

IVR prompt set

For a multi-prompt IVR, generate all prompts as TTS at once with the same voice, tagged ivr-main. Then your IVR sounds consistent. Don't mix TTS voices across prompts — it sounds disjointed.

Hold music

For one polished hold music track, source from a royalty-free music library (Bensound, Free Music Archive, Epidemic Sound). Upload as MP3. Tag hold-music.

Don't use copyrighted music. Performance Rights Organizations (ASCAP, BMI) can fine you for putting unlicensed music on hold.

Voicemail greetings

Personal: each user records their own via *97. No Media Library entry needed.

Shared boxes: TTS-generate the greeting and assign in the Voicemail box settings.