What audio analysis tools do
Audio analysis tools listen to a recording and report on it instead of changing it. The output might be structured data about a song, a list of possible matches for a melody, or a reading of what a speaker feels and intends. They are used by music teams, developers, researchers, trust and safety staff and curious listeners who would otherwise tag tracks by hand, transcribe lyrics line by line or review voice recordings one at a time.
Everything on this page takes audio as input and produces descriptions, candidates or signals. None of it is about editing, cleaning or generating sound, which belongs to other categories under Audio.
How to choose an audio analysis tool
Start with the question the audio has to answer. A song that needs lyric and music data calls for a different tool than a half-remembered tune or a recorded voice.
- Input type. Full songs, short hummed or sung clips and spoken voice are handled by different tools, so match the tool to the material you actually have.
- Output type. Structured data suits cataloguing and tagging, a candidate list suits finding a track, and emotion or risk signals suit review and monitoring workflows.
- Who runs it. A browser tool fits one-off checks by an individual. An API fits a product or pipeline that needs analysis on every recording.
- Pricing. Listings here range from Free and Free Trial to Paid. A trial or free tier is enough to test a tool on your own samples before committing.
- Platform. All three tools below are web apps, so platform rarely separates them.
Run the same few representative clips through any shortlisted tool. Results on clean studio audio say little about results on noisy phone recordings.
Notable tools
Sonoteller is an AI that listens to songs and turns them into structured lyric and music data. It is listed as Free/Paid and runs as a web app.
hum to search lets you hum, sing or upload a clip and returns song candidates in the browser. It is listed as Free Trial/Paid.
Velma API by Modulate is an audio-native API that reads emotion, intent and fraud risk from voice. It is listed as Paid/Free Trial, and its API form points to developers building voice products.
Common questions
What is audio analysis?
Audio analysis is the process of extracting information from a recording, such as the content of a song, the identity of a melody or the emotional state of a speaker. The tools here do this automatically and return the result as data or a list of matches.
Can audio analysis find a song from humming?
Yes, hum to search is built for that. A hummed or sung clip, or an uploaded one, produces song candidates in the browser, so the result is a set of possibilities to check rather than a single guaranteed answer.
Do these tools require coding?
It depends on the tool. Velma API by Modulate is an API, so integrating it takes development work. Sonoteller and hum to search are listed as web apps, which suggests they are used directly in the browser.
Is there a free option?
Pricing varies. Sonoteller is listed as Free/Paid, while hum to search and Velma API by Modulate offer a free trial alongside paid plans. Current limits are set by each vendor.
When to look elsewhere
Tools that name a sound rather than describe a recording belong in Sound Identifier. Those that improve a recording, such as cleaning up weak or noisy audio, belong in Audio Enhancer. For work focused on how a person speaks, Accent Identifier covers that narrower task.
