How to find music if you don’t know the name?

How to find music if you don’t know the name is a free music recognition service. It allows you to easily find out the name of the composition, the name of the artist and other information for almost any audio-musical fragment. Using the service is easy and simple. It is enough to download an audio file with a fragment of the composition (even 15 seconds is enough) and the robot, having analyzed the downloaded file, will try to recognize it and give detailed information about it. The downloaded audio file can be in almost any known audio format (.WAV, .MP3, .OGG, etc.), of any size (from a short fragment of a song to an entire track), and also of any quality (from a copy of a CD track to a low-quality and noisy recordings).

Is the servise free?

Yes, It is absolutely free

How it works?

The service is based on a sophisticated audio recognition mechanism that combines audio fingerprint technology and a large database of audio and music data. When you download a file, a special algorithm converts it into the so-called “audio fingerprint”, which is used to search the database. As a result of this recognition, one or more candidate tracks are found that match the processed file with one or another probability.

What formats are supported?

The system can process the audio file in almost any format known today. You can upload files, for example, in the following formats: uncompressed, ADPCM and other wave files (.WAV), MPEG-1 Layer III (.MP3), Ogg Vorbis (.OGG), FLAC, Flash Video (.FLV), .AMR, .MP4, as well as many others.

How reliable is recognition?

Recognition is completely reliable. Of course, recognition errors are possible, but still they are unlikely. The robot would rather not recognize an audio fragment unknown to it than give an erroneous result.

What is the optimal duration of downloaded audio files?

The optimal duration of the downloaded audio files is 15-45 seconds. At the same time, there are almost no restrictions on the duration — shorter and longer files will also be processed. Recognition of audio files is carried out in fragments of 15 seconds. For recognition, the robot selects from 1 to 3 random 15-second fragments, so it makes no sense to download very long fragments or entire compositions. To save loading time and traffic, it is recommended to download relatively short, but «indicative» fragments lasting from 15 to 45 seconds.

I uploaded a 30 second audio clip and it was not recognized. When I downloaded the whole composition, the robot produced the result. Why?

When you download an audio fragment for recognition, make sure that it is «indicative», that is, it contains enough audio information so that it can be distinguished. It is not recommended to select fragments from the very beginning or the very end of the composition; it is better to choose some part from the middle. If you are not sure, download the whole song.

Is the robot able to recognize melodically similar but different compositions?

The algorithm analyzes purely acoustic information and does not consider any melodic and / or musical characteristics. For this reason, he is potentially unable to detect different compositions that are similar only to the melody. At the same time, nothing prevents the robot from coping well with the recognition of even very noisy records.

How many tracks are in the database?

The music database has over a million tracks.

What kind of music is in the database?

The database is composed of music of various genres, artists and periods. It has both modern pop music and jazz, and classical, and rock, and many other genres. In addition, a considerable part of the database is Russian and Soviet music.

The robot could not recognize the downloaded file, however, I definitely know all the data about it. Can I help replenish the database?

Today, users can not replenish the music database. However, such an opportunity is planned.

What are the advantages of this service compared to others like it?

This service does not compete with others. Its main advantages are the ability to recognize audio even by very short and very noisy recordings, as well as its free access to the Internet without the need to install any software on a computer.

Добавить комментарий

Ваш адрес email не будет опубликован. Обязательные поля помечены *