A community app with a microphone input: servers, channels and voice chat, plus song search, karaoke with synced lyrics, and a recording vault only you can open.
In preview and free to use. Effects run on your own machine, so raw audio never leaves it to be processed.
Today the chat is in one app, the song is in another, the karaoke track is on a television, and the recording never gets made because setting it up costs an evening. Golden Karaoke puts them in the same place: a community with text and voice channels, a search box that reaches YouTube and Spotify, a karaoke screen for audio you are allowed to use, and a record button that writes to a vault nobody else can open.
It is built for people with gear and for people with a laptop and nothing else. Run a condenser microphone through an interface, or use the one hanging off your headphones — the same effects chain handles both, and it runs inside the browser rather than on a server you have to trust with your voice.
Every card says where it actually stands. The product is early, and a landing page that pretends otherwise only wastes the first ten minutes of your time.
Servers with text, voice and music channels. Roles, permissions, threads, mentions, reactions and direct messages — the structure you already know how to use, with music rooms alongside the usual ones.
In build. Accounts and profiles work today.
Search YouTube and Spotify from one box, or paste a link to identify a track. The conversation stays beside the player instead of in a second window nobody reads.
In build.
Lyrics that scroll and highlight in time, key changes of up to six semitones, tempo from half speed to one and a half, and section looping for the eight bars you keep fumbling.
In build.
A noise gate, EQ, compressor, pitch correction, reverb and delay you can reorder and save as a preset. Latency is measured rather than guessed, and the whole chain runs on your device.
Partly live. Device setup, meters and the latency test work today.
Piano, drum pads and synths played with a mouse, the computer keyboard or a controller. Map a knob to reverb, a pedal to record, a pad to a drum — whatever is already on your desk.
Partly live. Connected MIDI devices are detected and monitored today.
Multi-track takes that are private the moment they are made. Trim them, keep versions side by side, and decide per recording whether anyone else ever hears it.
In build.
Karaoke rooms, rehearsal rooms, jam rooms and open mics. A music director picks the song, the key and the tempo; everyone else joins as a singer, a player or an audience. A room clock keeps the song in the same place for all of them, so a guitarist in one city and a drummer in another are still counting the same bar.
In build.
Something you recorded, a backing track you own, or a licensed karaoke version. Drop the file in and the karaoke screen opens on it.
Choose your microphone and output, run the latency test, build an effects chain, and save the lot as a scene. It comes back exactly as you left it.
Lyrics roll, your voice goes through your own chain, and the take lands in your vault. Then decide: one friend, one community, everyone, or nobody.
They are how you find music here, and that is all they are. Playback happens inside their own players, under their own terms, and their audio never reaches the effects chain, the recorder or any analysis. No downloading, no stream capture, no exceptions for the one song you really want.
Karaoke, stems, pitch correction and everything else apply only to audio the platform is permitted to process: your own recordings, tracks you hold the rights to, and licensed material. The rule is enforced by the type system rather than by good intentions — a function that processes audio will not compile against an external track reference.
An email address is all it takes. Instruments, vocal range and the rest of your profile can wait until you care about them.