A radio station ID is a short voiced ident that says your station name (and often a slogan), mixed with a sting and a music bed so it sounds like a real station—not a bare voice memo. Below is a five-step method you can finish the same day: write the script, pick the voice, build the mix, check loudness and rights, then drop it into your playout.

Small internet, community, hospital, and hobby stations need IDs constantly: top of hour, after a stop set, when a new show starts. Hiring a full imaging package still works when you want a custom sung package and can wait. Many volunteer-run stations just need a clean spoken ID they can revise when the slogan changes.

We will use one fictional example end to end: Copper Pines FM, a small country-leaning community stream. The script we aim for is: “Copper Pines FM — classic country, closer to home.”

Step 1: Write a script that fits the clock

Decide how long the ID should speak. A practical rule of thumb used in radio-oriented AI tooling docs is about 13 characters per second of speech. That gives you a planning table you can stick next to the mic:

Target length	Approx. characters	Example use
5 seconds	~65	Name shout / legal ID
10 seconds	~130	ID + slogan
Up to ~17 seconds	≤220	One short production credit on some AI voice tools

Figure 2. Size the script before you record or generate.

For Copper Pines FM, the ten-second target fits the slogan without rushing. Spell names the way announcers should read them (FM as letters if that is your on-air habit). Avoid jokes that expire in a week; IDs get replayed for months.

Why: a script that is too long forces announcers to clip words. A script that is too cute becomes embarrassing at 7 a.m.

Check: read it aloud once at a natural pace. If you run out of breath before the slogan ends, cut words—not consonants.

Step 2: Choose the voice role, not just “a nice voice”

Station IDs usually want an on-air role: warm community, classic announcer, or light hype—not a cinematic movie trailer. Listen to three short samples in the same role family before you commit. Keep one voice across your ID, sweeper, and show tag so the station feels consistent.

For Copper Pines FM, a warm or clear announcer read fits a country-leaning community stream better than a nightclub hype voice. If your station name includes numbers or unusual spellings, listen specifically for those syllables; a pretty voice that mangled Pines is still unusable.

Why: listeners remember the station, not your tool. Switching voices every week sounds like a demo reel.

Check: play the sample into the same speakers (or headphones) your board ops use. Phone-speaker approval lies.

Step 3: Build the mix — sting, bed, voice, outro

A finished ID is rarely voice alone. The common anatomy looks like this:

Figure 1. Station ID anatomy — sting, voiced slogan over bed, outro.

1. Intro sting (about half a second to a second and a half)
2. Voiceover over a music bed, with the bed ducked under the voice
3. Outro hit so the piece does not end in dead air

You have three realistic production paths:

Figure 3. Choose a path that matches budget, turnaround, and whether you need singing.

Path	What you get	Honest tradeoff
Voice-talent imaging package	Custom delivery, often with sung options	Packages commonly start around £399 for a small set; turnaround can be days to weeks
Generic text-to-speech + a DAW	Cheap voice file	You still edit the bed, ducking, and loudness yourself
Radio-oriented mixer (voice + intro/bed/outro)	Air-shaped ID faster	Spoken jingles only—no sung hooks; you still choose a good script

If you want the third path, JingleLark’s AI jingle maker is an AI station ID and radio voiceover generator for small radio stations: type the script, pick one of 24 on-air voices, and it mixes intro, bed, and outro for you. You pay for the voice (one-time credit packs from $14.90 for 50 credits; no subscription); mixing, remixing, previews, and MP3/WAV downloads are free. It does not sing jingles, and it does not clone your own voice today—those remain voice-talent or specialist jobs.

Check for Copper Pines FM: the slogan is fully audible over the bed; the country bed does not fight the consonants in Pines.

Step 4: Normalize loudness and export a format your playout accepts

Many web and community streams aim near −16 LUFS with true-peak headroom so the ID does not slam louder than the song that follows. Whatever tool you use, bounce a file your automation actually accepts—usually MP3 for convenience or WAV when your chain wants an uncompressed drop-in. Know what your exporter really writes (compressed vs uncompressed) before you archive the master.

Play the ID between two music beds in your real playout software (RadioDJ, mAirList, PlayIt Live, AzuraCast, or your host’s cart wall). If the ID ducks the following song or sounds whisper-quiet, fix gain before you schedule a hundred plays.

Check: A-B the new ID against an older one your audience already accepts.

Step 5: Confirm you can air it, then schedule the rotations

Before the ID goes into top-of-hour rotation, confirm three paperwork facts:

1. Rights: you need permission to air the voice and the bed on-air (and usually on podcasts or social clips if you simulcast). Royalty-free downloads with no per-play fee are the usual requirement for volunteer stations.
2. Legal ID rules in your country: some licensed stations must say call letters and city at defined times. A branding ID is not always a substitute for a legal ID—ask your compliance person if you have one.
3. Naming consistency: Copper Pines FM should be spoken the same way in the ID, the website header, and the stream directory listing.

Then schedule: top of hour, after stop sets, and as a sweeper-adjacent piece if you cut a shorter variant later. Keep the long ID and a five-second name-only cut in the same cart category so ops can pick under time pressure.

If the slogan changes for a fundraiser week, revise the script and rebuild—do not stretch an old read with extra reverb and hope nobody notices.

Check: one cold listen the next morning. If you cringe, revise the slogan—not the volume.

What belongs in the first ID vs later imaging

Your first week on air only needs a solid spoken ID and maybe one sweeper. Show tags, holiday liners, and specialty drops can wait until the name is stable. Stations that try to launch with a twenty-piece sung package often stall waiting on revisions while the stream directory still shows a silent mount.

Copper Pines FM’s first shippable set: one ten-second ID, one five-second name shout, and one sweeper that teases the next country block. Rebuild later when the volunteer PD settles the positioner.

Do you need a voice actor for every station ID?

No. A voice actor or imaging house is the right call when you want a sung package, a celebrity read, or a tightly branded multi-piece launch. For everyday spoken IDs on a small station, a careful script plus a consistent on-air voice and a ducked bed is enough—as long as you verify loudness and rights before rotation.

FAQ

What is a radio station ID?

A radio station ID (or station ident) is a short on-air piece that announces the station name—often with a slogan—usually as a voiced line mixed with a sting and music bed so it can rotate through the hour.

How long should a station ID be?

Most spoken branding IDs land around 5–10 seconds of talk; use roughly 13 characters per second to size the script, and keep legal IDs compliant with your regulator even if the branding cut is shorter.

Can I air an AI-voiced station ID on my community or internet station?

Yes if the download’s license allows on-air use without per-play fees and the file meets your loudness and format needs; confirm the tool’s license text and your local legal-ID rules before heavy rotation.

Author note

Written for managers and volunteers at small internet, community, and hospital stations who need spoken IDs they can revise the same day. The worked example uses the fictional Copper Pines FM.