We listen to your stream every minute from 4 locations on 3 continents. When a majority of them hear a problem, you hear from us by email, Slack, Discord, or webhook, with the reason it failed.
No card, no trial clock. Works with any Shoutcast, Icecast, or HLS stream.
A single checker can't tell a dead stream from a bad network path between the two. So every stream is checked from four places at once, and they vote.
Every minute, our probes in Canada, the US East coast, France, and Australia each open your stream and decode real audio. It's not a ping or a port check.
A check only counts as bad when most of the locations that answered agree. If one region can't reach you, your dashboard shows it, but nobody gets woken up.
Three bad checks in a row trigger the alert. The all-clear waits for five healthy checks, so a stream that keeps dropping out doesn't flood your inbox.
Verdict: stream is up. No alert sent.
Australia's result shows up on your dashboard, so you can see a regional problem without being alerted about it.
A stream can be up and still be broken. We check more than whether it answers.
Refused connections, server errors, a missing mount point, a server that stalls. The alert tells you which, so you know where to look.
The server is up, but the stream is silent. You get a separate alert for silence, not an outage alert.
When the same audio keeps repeating across checks, the encoder is stuck. We flag it even though it isn't silence.
Loudness, peaks, noise floor, and dynamics over the last hour, plus a stereo phase check, all explained in plain language.
We email you 14 days and 3 days before your HTTPS stream's certificate expires, and again if it lapses, before players start refusing to connect.
Connection time from each location, charted over time. Slow starts show up before listeners start complaining.
Works with Shoutcast, Icecast, HLS, Radio.co, Zeno.fm, Live365, or any HTTP audio stream.
Every check decodes real audio, so we measure it too. Your monitor shows loudness, peak level, noise floor, and dynamics, with a plain verdict for each, and checks that your left and right channels are in phase.
When something is off, the tile tells you the likely fix in your studio: the encoder gain is too low, the input is clipping, or the line is picking up hum. If one channel is wired upside down, we tell you that too: in stereo it sounds fine, but phone and smart speaker listeners hear near silence.
Usually the encoder gain is set too low in your studio.
Every monitor keeps its uptime for the last 30, 90, and 365 days, and since the day you added it. The 90-day summary gives you the incident count, total downtime, and the longest outage.
Every outage and dead-air incident stays in the history with its start time, duration, cause, and the status message your listeners saw. Use it for sponsor reports, or to show your host a pattern.
Send alerts to as many people as you need, and to the tools your station already runs on. Each alert says what happened, why, and from how many locations it was seen.
Setup takes a minute. See alert notifications.
{
"event": "monitoring.alert",
"transition": "went_down",
"severity": "emergency",
"title": "Night Owl FM is OFFLINE",
"message": "Your stream stopped and listeners cannot hear your station right now.",
"stream": {
"id": "01a0ce99-4cb8-7d41-9484-8762b6e3c2e0",
"name": "Night Owl FM",
"source": "external"
},
"occurred_at": "2026-09-26T14:02:11Z",
"monitoring_url": "https://studio.cloudrad.io/monitoring/..."
}Add a status banner to your website with one line of code. It stays hidden while everything is fine and appears on its own when your stream is down or broadcasting dead air.
During an incident, type an update on your monitoring dashboard and it shows up on your site immediately. Your listeners know you're on it instead of wondering why the music stopped.
Press the button to simulate an outage. A status banner appears at the top of this page. Then type a message to update it live, just like you would from your dashboard.
There's no paid monitoring tier. Every broadcaster gets every feature, whether your stream is hosted with us or somewhere else.
Start monitoring freeA single checker can't tell your stream failing from its own network path failing. With four locations voting, one bad route or one busy region doesn't trigger an alert, and when all four agree, the problem is at your server.
No. That region couldn't reach your stream, but the others could. We show it so you can spot regional trouble, and we only alert when most locations agree.
Checks run every minute and an alert fires after three bad checks in a row, so usually within about three minutes of the problem starting.
If your server takes more than 5 seconds to answer or to start sending audio, new listeners can't start the stream. We treat that as unhealthy, and the alert says it was a slow response, not a refused connection.
Yes. Every monitoring feature on this page is free with no card and no time limit, for streams hosted anywhere.
More in the monitoring help center.