You hear a familiar song in the club or in the restaurant. You’ve heard this song thousands of times for a long time and the sentimentality of the song really touches your heart. Desperately want to listen again in the morning, but you do not remember her name! Fortunately, in our incredible futuristic world, you have a phone with music recognition software installed. You can relax, since the Software told you the name of the song, so you know you can listen to it again and again until it becomes a part of you or you get desperate.
Shazam music recognition algorithm
Mobile technologies, along with the immense progress of audio signal processing, have given us, algorithm developers, the ability to create music recognizers. One of the most popular music recognition applications is Shazam. Capture a song in 20 seconds, regardless of whether it is the intro, verse, or chorus, the application will create a fingerprint of the recorded sample, query the database, and use its music recognition algorithm to tell you exactly which it’s the song you’re listening to.
I’ll give you some advice. You can click this link to buy shazam followers for your shazam application, fast, real person who will follows you, and perfect service from the customer service. Or you can click this link tobuy real soundcloud followers, the best site to buy real follower for your soundcloud, facebook, twitter, and so on.
How does Shazam really work? Shazam’s algorithm the application was first exposed by its inventor Avery Li-Chung Wang in 2003. In this article we will talk about the fundamentals of the Shazam music recognition algorithm.
From Analog to Digital – A Sampling Signal
what is sound really? Is it some kind of mystical material that we cannot touch but it flies in our ears and makes us listen to things?
Of course, this is not the case. We know that, in reality, sound is a vibration that propagates like a mechanical wave mechanical pressure wave and travels through a medium such as air or water. When this vibration reaches our ears, particularly in the eardrum, it moves small bones that transmit the vibrations to the very deep hair cells in our inner ear. In addition, the few hair cells produce electrical impulses that are transmitted to our brain through the auditory nerve of the ear.
Recording devices mimic this process by pressing the sound wave to convert it into an electrical signal. A real sound wave in the air is a continuous signal of pressure. In a microphone, the first electrical component to find this signal is translated into an analog voltage signal – again, continuous. This continuous signal is not so useful in the digital world, so before being processed, a discrete signal that can be digitally stored must be translated into discrete signal. This is done by capturing a digital value representing the amplitude of the signal.
The quantization conversion implies the quantification of the input which necessarily introduces a small amount of error. Therefore, instead of a simple conversion, an analog-to-digital converter converts digital analog conversion to very small pieces of signal – a process known as sampling
Sampling and signal
The Nyquist-Shannon Theorem Nyquist-Shannon theorem tells us what sampling rate is required to capture a given frequency in a continuous signal. In particular, to capture all the frequencies that a human being can hear in an audio signal, we must sample the signal at a double frequency than that of the human auditory range. The human ear can detect frequencies between approximately 20 Hz and 20,000 Hz. As a result, most of the time recorded audio has a sample rate of 44,100 Hz. This is the sampling frequency of Compact Discs compact discs and is also the most Used with MPEG-1 audio (VCD, SVCD, MP3). This specific index was initially chosen by Sony because it could be recorded on the modified video equipment running at a rate of 25 frames per second (PAL) or 30 frames per second (using an NTSC monochrome video recorder) and covering the 20,000 Hz Of bandwidth needed to match the professional thinking of analog recording equipment). So when choosing the frequency of the sample that needs to be recorded you will probably want to choose to go with 44,100 Hz.
Recording – Capturing the Sound
Recording a sampled audio signal is easy. Since modern sound cards already come with analog-to-digital converters, you just select a programming language, find a suitable library and set the sample frequency, number of channels (mono or stereo) and usually the size Of the sample (eg 16-bit samples). Then open the line of your sound card, just like any input stream, and write in a byte array.