The music scene was revolutionized by recommender systems for music.  Sites like Last.fm, and Pandora explore the listener’s taste and makes recommendations based on music that they already enjoy.  These systems not only help music fans find music that they will enjoy, but it also allows bands to get more exposure.  Recommender systems attempt to decrease information overkill and keep hold of consumers by selecting a division of items from a universal set based on user likings. Research in recommender systems grew out of information retrieval and filtering. It has advanced into a legitimate and tricky research area of its own.

Tim Westergren created the Music Genome Project in January 2000.  The Music Genome Project was made in an effort to “capture the essence of music at the fundamental level” by using over 400 characteristics to describe songs. 

“A given song “S” is represented by a vector containing approximately 150 genes. Each gene corresponds to a characteristic of the music, for example, gender of lead vocalist, level of distortion on the electric guitar, type of background vocals, etc. Rock and pop songs have 150 genes, rap songs have 350, and jazz songs have approximately 400. Other genres of music, such as world and classical, have 300-500 genes. The system depends on a sufficient number of genes to render useful results. Each gene is a number between 1 and 5. Fractional values are allowed but are limited to half integers.”  (http://en.wikipedia.org/wiki/Music_Genome_Project)

 Pandora is an automated music recommendation and Internet radio service.  The Music Genome Project created Pandora.  Users name a song or band that they like, and Pandora counters by playing music that is similar.  User feedback is tracked with a “thumbs-up” or “thumbs-down,” which Pandora uses for future selection choices.  The “thumbs-up” tells Pandora to play more of this selection; no response (no thumbs up or down) will not change in the preference, and the “thumbs-down” tells Pandora not to play this track again, and play fewer that are similar to this undesired song.  My only complaint about Pandora is it will only allow the user to skip so many songs in an hour, so if you keep getting songs that are “failures” as far as your taste goes, you have to switch to another “station,” to get around this skipping songs glitch. Overall, Pandora is very user friendly; anyone who can get on the Internet can use it.  It also gives the users the option of purchasing the songs or albums at Amazon.com or in the iTunes Store.  Pandora has two subscription plans.  The free subscription plan is supported by ads and the fee-based subscription without the ads.  I honestly do not see why anyone would pay for an ad-free version, I don’t mind looking at a few ads. 

Last.fm is a UK-based Internet radio and music community website.  Last.fm was founded in 2002.  Unlike Pandora, Last.fm is available throughout the world.  Last.fm uses a unique music recommendation system known as the “Audioscrobbler.”  The “Audioscrobbler” is a downloadable client that “scrobbles” music the user listens to on his or her own computer or on their iPod.  The “Audioscrobbler” will give a brief biography on the artist, list the tags associated with the artist, and a brief list of similar artists.  The users can friend one another and while you are logged in and are looking at another user’s page it will tell you how compatible your taste in music is with theirs.  Each user has a personal recommendations page known as “The Dashboard.”  “The Dashboard” is only viewable to the user of the account, and it includes lists of new music, events, journal entries and other people with similar tastes (known as Neighbors), all fitted for the user’s preferences. Recommendations are designed using a collaborative filtering algorithm so users can browse and hear previews of a list of artists not listed on their own profile but which appear on those of other users with parallel musical tastes. The page also lists music that has been suggested to the user and groups the user belongs to, journals written by users about artists the user listens to, and other users who have listened to similar music in recent times. There is also a ‘recommendation radio’ station, which will play music exclusively filtered, based on the user’s last week of listening. Last.fm also permits users to manually support specific artists, songs or albums to other users on their friends list or groups they belong to, providing the suggestion in question is included in the Last.fm database.  Last.fm is a free site, though some people chose to pay a small fee ($3, €2.50, £1.50 or ¥350) per month for “extra features” such as: no advertisements, more radio options, the ability to see who has viewed their profile page, Beta testing at beta.last.fm, and the user icon changes color from grey to blue, and says “Subscriber.”

Critical Metrics is a playlist of music, filtered automatically by the tastes of professional critics.  Some believe that these professional critics are worth listening to since they analyze and filter through music.  I’m not sure I need a “professional” telling me what is good music and what is not.  They might be able to recommend music that fits with the music I already listen to, but I do not need someone telling me what “good music” is.  Critical Metrics keeps track of recommendations and playlists across all media so you can easily find, try, and (here’s the catch) buy the “best new music.”  Honestly, I found Critical Metrics to be more of a musical search engine; it’s not nearly as user friendly as Pandora or Last.fm, and as for “recommending” music- that is limited.  I believe Critical Metrics is some kind of scam set up by the music industry.

Call me biased, but my favorite sites of the three are Last.fm. I have had my account with Pandora the longest, and as much as I enjoy the constant stream of music that comes from the site, Last.fm’s “Audioscrobbler” might be one of the best inventions ever.  I don’t like to type in every single artist whose music I enjoy.  Last.fm’s “Audioscrobbler” does all the work for you, all you have to do is sit back and listen to music, as long as the “Audioscrobbler” is up and running it will “scrobble” all your music, which will automatically be added to your profile.  It can’t get any easier than that!

Now to my next issue with these YouTube Questions and Presidential Candidates, it was COMPLETELY Biased!  Obviously candidates such as Hillary Clinton received more questions thus more exposure.  It has me wondering who she had to pay off to have more mic time?!

Many of the candidates were talking about how we need to “tell the truth” yet they continually skated around questions, and those who didn’t?… well they probably threw themselves out of the running because somehow they pissed someone off.

My previous post was a short “movie” created by my roommate, Justin, (with a little help from myself as well) of our friend, Eric’s 21st Birthday Party… which actually took place last November… so when you think about it… all the footage had been sitting around awhile before anything was made of it.

To start off, there was a lot of footage, a lot more than this 1:39 min movie shows.  If I remember correctly we had a good 90 minutes worth of footage, but as with any movie most of it ends up on the clipping room floor… or in our case, not used.

All sound from the footage was removed, and in place the sound from the “300″ trailer was placed in.  The key in the making of this video was making clips seem like they fit with the sound from the trailer.  It took several hours of going through clips and deciding which worked best where.  Yet there was one part that there wasn’t a suitable clip for, so my little digital camera was thrown into video mode, and I recorded Justin “slicing” a lime. That was actually the very last thing that was done for the whole video. So that tiny part of the lime being sliced was filmed 10pm Sunday night… but you wouldn’t have known that without me telling you.

I’d say it turned out pretty well, considering we had footage of drunk kids… not people acting out the “300″ trailer.

The two sites have similarities such as the user will name a band that they are a fan of, and in return the site suggests other bands that have a similar sound to the band that the user is already familiar with.  In general, it is a great way to discover new bands that the user is almost guaranteed to enjoy.

Pandora bookmarks bands or songs that the user types in, and plays a song by the band that has the band’s sound.  Then goes through and plays other bands that have a similar sound and the user can decide whether or not they like the song with a thumbs up or down.  Unfortunately, you can’t type in an exact song you wish to hear and you can only skip so many songs in the given amount of time.

Last.fm on the other hand has a client that the user downloads, and as the user plays the music on their computer through a program, such as iTunes, it “scrobbles” the music and adds it to your profile.  The user can look up specific bands and add it to their playlist.  The downfall of Last.fm is the “scrobbled” songs are not in their entirety.

So which site is better? I guess it depends on who you ask.  I have an account with both, although I use Last.fm more often than Pandora, but that is a matter of personal preference.

I find this quote far too amusing…. MAGIC!