Radio Imaging: What It Is and How to Make Yours (IDs · Sweepers · Liners)

What is radio imaging?

Imaging is the station's voice between the music. Songs change every few minutes; the imaging is what stays the same, so a listener who tunes in halfway through an hour still learns the station name, the slogan and what show is on. Each element has a job, and most are short.

Why does a radio station need imaging?

Radio imaging does three jobs. First, it tells a new listener the station name within seconds of tuning in. Second, it bridges songs, breaks and shows, so the stream never drops into dead air. Third, it repeats the same voice, slogan and sound every hour, which is how listeners remember the station and find it again. Internet and community stations need this as much as licensed ones: a listener who arrives from a directory or a smart speaker has no dial position to remember, only the name.

What goes in a starter imaging package?

A new internet or community station does not need a 50-piece package on day one. The table is one example starter set that covers a normal week of programming — an assumption to adapt, not a standard. The last column shows what it costs to make with JingleLark, where each voiceover of up to 220 characters is one credit and 221–900 characters is two.

• 27 credits fit in the $14.9 Starter pack of 50 credits, leaving 23 for retakes or next month's promos. At Starter prices the whole set is about $8 of credits.
• Credits are one-time and never expire, so a station can make its set in one evening and top up a promo months later from the same pack.
• For comparison, human imaging packages from imaging companies start at a few hundred pounds or dollars for a similar set.

Prices, turnaround and sound compared side by side: AI station imaging vs hiring voice talent.

How long should each element be?

Shorter is almost always better. A station ID only has to land the name before the next song; a sweeper has to finish before the listener reaches for another station. As a rule of thumb, an AI voice reads about 13 characters per second at normal speed.

• 5 seconds ≈ 65 characters — a name and a three-word slogan.
• 10 seconds ≈ 130 characters — a full station ID or sweeper.
• 15 seconds ≈ 195 characters — a liner or show tag with one extra detail.
• Up to 220 characters (about 17 seconds) is one JingleLark credit.
• 30 seconds ≈ 390 characters — a promo, which is two credits. The limit is 900 characters, about 70 seconds.

Speed can be set from 0.7× to 1.2×, so a line that runs long can be tightened without rewriting it. Read the script aloud once before you generate: if you run out of breath, the listener will too.

How is radio imaging mixed? (ducking and loudness)

A bare voiceover is not imaging yet. What makes it sound like radio is the mix: an intro hit, the voice riding on a bed, and an outro that closes cleanly so the next song can start. Imaging producers build that in an editor. JingleLark builds it for you on the server when you press Build.

1. Intro: a short hit or swoosh opens the element.
2. Bed: background music loops under the voice for as long as the read lasts.
3. Ducking: sidechain compression pulls the bed down whenever the voice speaks and lets it back up in the gaps, so every word stays clear.
4. Outro: a closing hit or fade ends the element.
5. Loudness: the finished file is loudness-normalised, so your IDs do not jump out louder or quieter than each other.

Intros, beds and outros come from a library of 88 sounds made for JingleLark, grouped by style such as news, rock, synthwave, orchestral and seasonal. You pick one of each, or swap them and rebuild as often as you like for free. The sounds are only available inside finished jingles and are not sold or downloaded on their own.

Do you need separate editing software? No. You download a finished MP3 or WAV that is ready for playout. If you already use an editor, you can still open the file in it to trim or layer it further.

Where do station IDs go in the hour?

Licensed US broadcast stations have a legal ID rule. Under FCC rule 47 CFR §73.1201, a station identifies itself at the beginning and end of each time of operation and hourly, as close to the hour as feasible, at a natural break in programming. The legal ID is the call letters immediately followed by the community of license; the licensee's name or the frequency may be inserted between the two.

The FCC rule does not apply to internet-only stations, and rules differ by country and platform, so check what applies to you. Either way, a steady pattern helps listeners learn the name. One common way to lay out an hour:

• Top of the hour: station ID (and the legal ID if you are licensed).
• Between songs: a sweeper every few songs, so the station name comes round regularly.
• Out of a break: a liner leading back into music.
• Start of each programme: the show intro or show tag.
• Inside breaks: one or two promos for upcoming shows and events.

Hear radio imaging examples

Three elements made with JingleLark, each a single line mixed with an intro, a bed and an outro. The station names are made up for these examples.

Should you hire a voice or make it yourself?

Hire voice talent or an imaging company when you want a sung jingle, one recognisable voice you can book for years, or sound design built from scratch for your format. JingleLark makes voiced imaging — IDs, sweepers, liners and drops — not sung songs. If you need a singer, a jingle house is the right call.

Make it yourself when the budget is tight, the schedule changes every week, or you need a liner for tonight's show. With JingleLark you type the line, pick one of 24 AI radio voices in 6 on-air roles or a voice you cloned from a short sample, and download a mixed element in the same session. Downloads are royalty-free for broadcast, podcasts, social and paid ads, with no per-play fees.

What tools do you need to make radio imaging?

There are three ways to get station IDs, sweepers and liners on air: produce them yourself in audio editing software, hire a voice or an imaging company, or use an AI tool that delivers the finished element. Which one fits depends on your budget, your time and how often your lines change.

With JingleLark the voice, the bed, the ducking and the loudness levelling happen in one step. A voiceover is ready as soon as it is generated and you build the mix in the same session, so there is no booking queue or delivery date to wait for.

Prices and turnaround from human imaging services, side by side: AI station imaging vs hiring voice talent.

Radio imaging tools and guides

Go deeper on one part of your imaging:

Radio imaging FAQ