An iTunes 'most popular only' playlist

Apr 21, '04 09:35:00AM

Contributed by: cedric

Well, it's rather a model than a hint. I just found a way to play my favorite music endlessly in iTunes without having to modify the playlist. Yes, of course, it uses the Smart Playlist feature. But I found most of the criteria in there were overrated software features, rather than a real useful way to manipulate your collection. The trick is the combination of all these criteria. Let's take a tour. Note that you have to check the "Match All these conditions" button.

  1. "Last played is not in the last week." You can change this of course, but I found that a week is a good buffer.
  2. Only songs that have a Play Count greater than 5. This is just because, although having thousands of songs in my iTunes library, I listen always to the same "pool" of songs. Moreover, this is the goal of this playlist.
  3. The limit of 25 songs is tricky. Theoretically, you could put 200 here, to have a large selection to choose from. But that's not the goal. This limit of 25 songs is here to keep coherence. Most of the time, I like to listen my music by albums. Not always, but most of the time. This playlist will not play music by albums, but this limit of 25 (about two times the number of songs in album) ensure that the songs of the same album will stay close in the buffer, and keep appearing close. See below.
  4. Match only the checked songs. Obvious.
  5. Live updating. Hence, as soon as a song has been played, it is removed from the list of the 25, and new one is in!
I have posted a picture of this playlist on my website. Save your play list. Select it on iTunes, and press the "Shuffle" button. Voila! This shuffle button is very important! It is meant to make a random selection in the list of 25 songs. So to explain this a bit more. When you start your playlist the first time, you have a pool of songs that are all your songs not played the last week. Among these songs, you choose only those who already have a Play Count above 5. This is just to make a rough separation between the two fuzzy set of songs I have: the ones I listen often, the ones I listen very few. Feel free to change!

Then, you play a song in this play list, and it disappears off the list for a week. Its Play Count is increased by 1, and will therefore re-appear later. Thanks to the live updating, you will have a fresh new song coming from the pool. It will therefore never finish. Thanks to the shuffle, it will choose randomly among the 25 songs in this list.

During the first week, you will have the songs coming one by one in my playlist belonging to the same album. But after one week, it starts to be a bit more random. But the limit of 25 songs is made on the "most recently played" songs. This is where it keeps coherence. It will preferably choose the most recently played. And since it has been running on a list of 25 songes only, and since your first week is composed mainly of complete albums, it will choose songs most of the time in the same album. I think. I am not 100% sure of this. I started five days ago...

What happens when you have encoded/bougth a new album? Answer: you will have satisfy the two conditions: wait at least a week before getting this into our playlist, and to listen to it at least 5 times. Then it will appear sooner or later. Since it is made on a "negative" criterium ("not played last week"), you can even go on vacation for a week (or more!) and your playlist is still valid.

If you choose to not always play this playlist (it might happen), you will increase the Played Count of new albums and songs. Some of them will go beyond the limit of 5, and will appear in the pool. That makes the pool grow along with your evolution in choice of music.

After a while, you may increase the number of 5 in the criterium, just to shrink your selection a bit. The only way I found to make some songs disappear is to uncheck them. They completely disappear from this playlist. But if you still want, one day, to listen to them, you will have to check them again, increasing their play count. Ok, you see the point.

Comments (4)


Mac OS X Hints
http://hints.macworld.com/article.php?story=20040418221951950