How Queue Music announces the current queue title through screen readers
For Australians who stream music through their browser, the question of how a player communicates the name of the track about to start is rarely about novelty. It affects whether a person can enjoy a playlist in a share house in Fitzroy, listen during a commute on the NSW TrainLink network, or follow along with a community radio segment from Triple J without staring at a screen. Queue Music treats the spoken title as a core accessibility feature rather than an afterthought, weaving announcements directly into the way the player queues and streams content from YouTube.
The project, which was built by Thomas Logan with funding from the Mozilla Foundation, was shaped from the start by the needs of keyboard-only users and people who rely on assistive technology. Every queue change is paired with an audible cue that names the track, the artist and the position in the queue, so the listener never has to guess what is about to play.
The polite live region behind each announcement
At the heart of the player sits an ARIA live region set to the polite value, which means the screen reader waits until it finishes speaking the current sentence before announcing new information. This avoids the jarring overlap that happens when a player fires off multiple announcements at once, a complaint often heard from listeners in busy environments like a Brisbane café or a Melbourne tram carriage. When a track begins, the live region receives a string such as "Now playing: track name by artist, number 4 of 12" and the assistive technology reads it out without interrupting other speech.
The polite setting is a deliberate trade-off. A more aggressive aria-live value would shout new titles over menu navigation or form input, which would frustrate users who rely on spoken feedback for typing. Queue Music pairs the polite region with a smaller assertive region reserved for errors, such as when a YouTube embed cannot load or the queue stalls.
| Announcement type | Live region value | When it fires | Typical message |
|---|---|---|---|
| Track start | polite | A new track begins streaming | "Now playing: title by artist, position N of M" |
| Queue advanced manually | polite | The user presses a navigation key to skip | "Skipped. Next: title by artist" |
| Queue cleared | assertive | The user empties the queue or ends the session | "Queue cleared. Playback stopped" |
| Stream error | assertive | A video is unavailable or blocked | "This track could not be loaded" |
| Saved queue restored | polite | The page reloads with a stored queue | "Restored N tracks from your last session" |
Because the polite and assertive regions sit in a hidden element near the playback controls, they do not interrupt the layout that sighted users see. A person browsing the queue with their eyes continues to watch the highlight bar move, while the screen reader user hears the matching narration.
When the title gets spoken
The title announcement is triggered by a small set of events that the player tracks in JavaScript. Moving to the next track, jumping with a keyboard shortcut, restoring a saved queue after a browser refresh, or shuffling the order all result in a new live-region update. Search results are handled differently so the page does not freeze while YouTube is being queried in the background. The journal entry explains how each keystroke debounces the request and reports completion back through the same accessibility plumbing.
A shuffle, for example, writes "Shuffled. Now playing: new title" into the polite region, giving the listener confidence that the order has changed. Restoring a session announces the first track rather than the full list, so the listener is not forced to wait through twelve titles before hearing anything useful.
Keyboard controls that trigger spoken feedback
Almost every keyboard shortcut in Queue Music has a matching announcement, the result of testing with blind and low-vision contributors from organisations such as Vision Australia and the Australian Disability Clearinghouse on Education and Training. Pressing J moves to the next track and triggers the polite announcement, while pressing K pauses without firing a long message because the screen reader already knows the state has changed. Pressing the spacebar to toggle playback tells the listener whether the queue is paused or playing, and pressing S to save speaks a short confirmation rather than dumping the whole queue back into the live region.
The combination of shortcuts and announcements means a user can sit in a park in Adelaide, plug in earphones and run an entire listening session without touching a mouse or looking at the screen. That design intent runs through every control: each visible change must have an audible counterpart.
How the player formats titles before speaking them
Before a title reaches the live region, the player passes it through a small formatter that strips smart quotes, removes emoji and trims whitespace inherited from messy YouTube metadata. Long artist names are shortened to the first twelve characters followed by an ellipsis only when the combined track and artist string would push the announcement past the eighty-character ceiling that most screen readers handle comfortably.
Unicode characters that confuse some text-to-speech engines, such as fancy dashes or full-width punctuation, are normalised to plain ASCII. Capitalisation is left untouched so the listener hears the song title as the artist intended, which matters for tracks that rely on unconventional casing.
Compatibility with local assistive technology
In Australia, the two screen readers that show up most often in feedback are NVDA on Windows and VoiceOver on macOS and iOS, with JAWS still common in offices that have not yet migrated. Queue Music has been tested against the latest stable releases of each, as well as Narrator on Windows for users who cannot install third-party software. The polite region behaves consistently across them, although VoiceOver on older iPhones occasionally swallows the first syllable of long titles, which is why the formatter keeps announcements short.
The phrasing avoids assumptions that do not travel well, such as translating American station call signs or assuming a listener knows what a regional reference points to. Track names are read exactly as YouTube provides them, with smart quotes and unusual capitalisation stripped.
Practical tips for getting the most from the announcements
Users who want to fine-tune the spoken behaviour can adjust a few settings through the help documentation, which walks through the live regions, the keyboard map and the storage preferences for saved queues. A handful of small habits make a real difference in daily listening.
- Turn on the verbose playback option in the settings panel if you want every skip and pause to be spoken, rather than only the track start.
- Save the queue at the end of a long session, so the next visit restores both the order and the playback position without you having to rebuild it.
- Use a wired headset on noisy commutes through Sydney's City Circle, since Bluetooth audio codecs can clip the first word of a polite announcement.
- Pair the player with NVDA's speech viewer if you want to see the live region text on screen while you learn the keyboard shortcuts.
- Report any unannounced events to the project through the help page, so the next release can wire more actions into the polite region.
That approach turns the queue title announcement into a dependable companion for every listening session, whether the listener is at home in Hobart or travelling through regional Western Australia.