Ever wondered how Shazam does what you can't do? Remember the song? Yeah. I didnt either. But I still made a video about it lol Music by youtube themselves and Epidemic Sound Enjoy! (and maybe please suscrib)
he literally copied a youtuber named "This." and likely planning to make more content on his style. regardless i like you way of doing things and i hope you make it big on youtube (cuz i couldnt...atleast for now)
Its the Fourier Transform. Physics here. I did this project for a class once. Basically the Fourier Transform decomposes any complex sound (which is an agglomeration of diferent frequencies) and now you are able to "see" each frequency individually. Then the spectogram is a 3D graphic representation of the song (complex sound) by intensity of each frequency and by time. Of course for each song, there is a unique spectogram. The last step is to compare the spectogram in the hash database mentioned in the video and find the one that matches yours.
I study EE and just now taking Harmonic Analysis where we should study about the Fourier Transform in a few weeks. It's fascinating seeing the math come to life.
incorrect, they have a bunch of kids they kidnapped from the temu factories that are fed 1 meal a day and pee in bottles who also know every song in existence and they are the ones behind the song recognition ability that shazam has (I used to work there trust I know fr)
I think I already used it back then to find the title of a song I heard in a store, my circle is already using it. It was probably 2013 or 2014, the song was Young Blood by The Naked and Famous
I don't get how it calculates a hash out of a spectrogram, which contains noise, and uses that to find the right song, which has been stored in the database using a hash without noise. Probably some magic hash algorithm created by math wizards.
Thanks to fourier transforms, (the spectogram) you get all the frequencies picked up by the mic, so if the noise is not overwhelming the sound of the song youre trying to record, then shazam will just compare the strongest frequencies with the songs it has in the hash, not minding the noise since it pales in conparison to the frequencies of the song
Crazy how a good chunk of songs in my playlist is because I hard a song, i liked it, then Shazam. From shazam, I get to discover more songs from that artist
Especially if it's meant only for particular audiences, e.g. producers, remixers, and music artists. Fortunately, they started to identify this kind of song since circa 2021 Production music, music meant to be remixed, might throw you songs that remixed them. For example: A Letter 2 My Unborn by Tupac on Shazam and some Arabic music on the Google app music identifier. I discovered the production sample in the original form aka no voices and whatsoever in an unfinished RPG Maker game "The Legend of the Philosopher's Stone" with the filename "Dungeon nine.mp3". I don't know how the game creator knew the production sample Production music are meant to be production secret, I guess. "Your Favorite Martian" seemed to intentionally hide his production music library file names when he talked about a Fortnite song sounds the same to his
google identifies a song much more accurately and faster, giving you similar songs too.. and its been around for a couple of years now. my first choice is google, second is shazam and third is aha music
@@pbilk you can click on speak to search on normal google on your phone (not chrome) and the will be a button "search for a song". there is also a widget called 'sound search' by google, u can try if any of this works on pc peace
Damn, him saying "shazam has existed for longer than most of you have" and then me realizing it was released a month after I was born made me feel so old
I used to think Shazam was simply recognizing the lyrics of songs and then searching for those in its database of song lyrics, but then I realized it also works for songs without any lyrics at all, so that quashed my theory. As for Shazam's ability to record the audio of a song in a noisy environment, I've been impressed with its ability to single out the song audio over against the other noises.
