
About this product
hum to search is a browser-based music recognition tool for the specific moment when a melody is stuck in your head and you cannot name the song. Instead of typing lyrics or describing a tune in words, you perform it: hum with closed lips, sing the notes, whistle, or play a recording you already have. The site listens, returns song candidates, and leaves the final confirmation to you, because a candidate is only useful once you play it and recognise the track.
The product separates two situations that look similar but need different handling. If you remember how the song goes but cannot play the original, you use the hum or sing mode. The site asks for one continuous section rather than disconnected fragments, and suggests a chorus or repeated hook because that is the part people reproduce most consistently. Lyrics and the original key are optional, and the guidance is to sing at a comfortable pitch rather than strain for notes outside your range. If the real recording is available, either playing nearby or saved as a file, you switch to audio recognition and upload the clip instead.
Recordings are short by design. The site states that clips run about eight to twelve seconds and that audio is not stored. Microphone access starts only after the record button is pressed, so the page does not listen in the background while you read the instructions. For anyone who is cautious about voice recordings, that order of operations is easy to check on the page itself before granting anything.
Results are presented as candidates rather than a single answer. The workflow the site describes is deliberately two-step: get possible matches, then confirm the title and artist by listening. That is the honest shape of melody recognition. A hummed chorus can sit close to several songs, and a tool that returns one confident-looking title without asking you to verify it would be harder to trust than one that shows its shortlist.
Access follows a simple pattern. Each person gets one combined trial, and after that the product offers optional one-time credit packs rather than a subscription. There is no account requirement to try the recognition flow, no software to install, and the whole interaction happens in the browser on desktop or mobile. The site also publishes a pricing page and a blog that explains song identification terms and walks through recognising older or less familiar tracks.
What the product does not claim matters as much. It is not a lyrics search engine, it cannot work from a spoken description of a song, and it does not accept an album cover or an artist name as audio evidence. Those limits are stated on the page rather than discovered after a failed attempt, which makes the tool easier to judge before you use it.
For a commuter with a chorus on repeat, a teacher identifying a piece for a lesson, or anyone who has ever recorded a few seconds of music in a shop and forgotten about it, hum to search offers a direct path from the tune in your head to a shortlist of real songs. The free trial is enough to test whether the approach works for your voice and your memory of the melody, and the credit packs keep the commitment small if it does.
Key features
- Hum, sing or whistle a melody and get song candidates without typing lyrics
- Upload or play an existing recording when you have the real audio instead of your own voice
- Short 8-12 second clips, with audio not stored and the microphone opening only after you press record
- Candidates are shown for you to confirm by listening, not presented as one unverified answer
- One combined free trial per person, then optional one-time credit packs instead of a subscription
Pros
- Works from a melody you hum, sing or whistle
- Accepts an existing recording when you have the real audio
- Short 8-12 second clips and audio is not stored
Cons
- One combined free trial per person, then paid credit packs
- Needs a recognisable melody, not a spoken description
Reviews
Reviews are moderated before appearing publicly.
No reviews yet — be the first to share your experience.
ProductReveal