You need to upgrade your Flash Player This is replaced by the Flash content. Place your alternate content here and users without the Flash plugin or with Javascript turned off will see this. Content here allows you to leave out noscript tags. Include a link to bypass the detection if you wish.

About Us

At BMAT we are passionate about music and technology - especially when the two come together. For this reason we decided to create a company committed to developing new software products dedicated to people’s interaction to music.

BMAT was born as a commercial spin-off of the Music Technology Group, the world's largest research lab in music & audio. In early 2006, a group of engineers, musicians, computer science PhD's, sound producers, digital signal processing experts and musicologists joined forces to exploit their passion and skills in developing cutting-edge music technology solutions.

The first thing we did was to design an intelligent, efficient but at the same time user-friendly music search and recommendation engine, Ella.

We put together a set of algorithms that we had been working on for the last eight years that enable a computer to describe and understand music. These algorithms automatically detect tempo, bpm, chord progression, key, tonality, instrumentation and mood. They were perfected and combined with context information (country, label, genre, recording date, reviews, concerts, etc.) and even with moods, just to make a search engine solution that is smart and human-like.

Next, we developed the software, Vericast, to enable the computer to identify music and audio files for monitoring radio, TV, Internet and live performances.

Using the algorithm for audio fingerprinting we patented in 2001, we rewrote it from scratch to create a robust, scalable, agile and efficient solution for audio identification. Pushing the envelope one step further, we have just released a patent-pending technology for cover song detection that is unique world-wide as of today.

And finally, we created the first-ever virtual judge that is able to analyze and evaluate singing performances.

Utilizing our ten years of experience with digital voice processing, we developed a software algorithm that is able to “Skore” singing without using a reference track and regardless of what song is sung, providing a rating based on pitch and expression.

Of course, all of this would be quite useless without a place to work from so we decided to rent a nice place near the beach in Barcelona, put some tables, big loudspeakers, a lot of computers and got started presenting our software solutions to the world.