What is AudioTag and what it does?

AudioTag.info — is a free music recognition service. It allows identifying almost any unknown piece of music quickly and easily. The use is very simple: upload a short audio fragment or an entire song (in fact, even 10 second fragment is enough), let the robot analyze it and provide you with the information about the track title, artist name, album title, etc. Your audio fragment can be in almost any file format and of almost any quality (aurally recognizable, of course): it can be an MP3 file downloaded from the Internet or a short recording made with your old tape recorder and digitized into a low quality audio file. AudioTag can also recognize audio directly from Youtube and other video sharing services.

Is it free?

Yes, the service is absolutely free for non-commercial use.

How does it work?

This service is based on a sophisticated audio recognition engine which combines a proprietary, patented audio fingerprinting technology and a multimillion songs' database. After you upload an audio file, it is analyzed by the underlying audio engine. During the recognition, an audio fingerprint is extracted from the acoustic content of the uploaded audio file, and is then identified by the engine by means of a special fingerprint matching algorithm and a large music fingerprint database. Information on any matching candidate songs is displayed in the end of the process.

What are the supported file formats?

You can upload files in almost any file format. Uncompressed, ADPCM and other waves (.WAV), MPEG-1 Layer III (.MP3), Ogg Vorbis (.OGG), FLAC, Flash Video (.FLV), .AMR, .MP4, and many more file formats are supported. AudioTag can recognize audio tracks in video files too.

Is the recognition reliable?

Yes, the recognition is reliable. Erroneous results are possible, but the probability of such event is low.

What is an optimal duration of an uploaded audio fragments?

Optimal duration for uploaded audio is 10-30 seconds. Shorter and longer audio files can be recognized as well. You can upload even very long recordings, but the robot will use a shorter fragment for recognition not to occupy the server too much and to allow others using it.

I've uploaded a short music fragment and it could not be identified, but the recognition worked once I’ve uploaded the entire song. Why?

When you upload a recording containing a fragment of an unknown song, please make sure it is representative enough, i.e. contains enough audible information to recognize it. It may happen that the fragment you selected is not informative enough and then uploading a longer recording can raise the chances to recognize it successfully.

Can the service recognize multiple songs in a single recording?

Yes, the engine can recognize different songs present at different fragments of the same audio recording. Recognition results contain information on the candidate songs with their corresponding time intervals relative to the start of the analyzed recording.

Can the robot recognize melodically similar tracks??

The underlying core audio fingerprinting technology operates with purely acoustic characteristics of sound and does not consider any melodic or musical properties. Therefore, it is unable to identify melodic similarities and can only find tracks with exactly matching sound. On the other hand, the algorithm is robust to sound distortions, noises and even speed variation, and will therefore recognize songs even in low quality audio recordings.

How many tracks in the data base?

The songs database includes more than two million tracks.

What kind of music is contained in the database?

The database contains mostly popular music (1940-s to the modern days), rock, blues, electronics, jazz and classics.

Can I contribute to AudioTag?

We do not support user submissions to the database at this moment, but we plan to implement such ability in the future.

What are the advantages of this service compared to other similar services?

AudioTag can recognize audio even by a very short audio fragment, it is available online and is free, it can report multiple songs contained in the same recording with time-stamps, and it can recognize audio directly in YouTube videos.

Does AudioTag have API?

Yes, it does. API is available to all its users and is free. Certain limitations apply to the free API usage.

Can AudioTag be used commercially?

AudioTag can be used commercially via its API.