I was ‘cleaning’ up my music collection last night (clearing out songs I don’t like to reduce the space taken up by the collection) and I came across quite a few tracks, some as long as 40 minutes, where the song would last a normal amount of time (say 5 minutes) and the rest of the track would be silence.
I can have a guess at the answer to my GQ because these tracks are usually the last tracks on the CD I got them from. But I’m still curious.
I neglected to check the file size of these songs, but unless MP3 compression is good at compressing silence into a very small space (even 35 minutes of it!) I am assuming these tracks take up a lot of file space. Which would be wasteful.
Why do some end-of-cd tracks have such long periods of silence?
Case-in-Point “Miss Blue” from a Filter album I have.
Re pulykamell’s suggestion - I own that Filter CD, and I’m pretty sure that “Miss Blue” has some music at the very end. I took it off of my iPod as I was annoyed by the silence, but the Wiki listing of the CD does confirm there’s an untitled song on it.
Odd. Considering that Miss Blue is my favourite track on the CD I was fairly sure I’d listened to the whole thing for the very purpose of seeing if there was anything at the end. I guess I was mistaken.
But for many of the tracks I was sorting last night that had this phenomenon I did seek to near the end and heard nothing.
I shall listen again tonight when I go home.
Hmm, maybe you’re right. I wonder if it’s a variant on Wilco’s “Less Than You Think” (off A Ghost Is Born), where an electronic hum trails off for several minutes at the end of a song, letting the listener choose when to end it. Maybe the long silence that isn’t followed by a hidden track is to lull the listener into just listening to silence as a calming effect or something.
In my case it just causes the listener to think his MP3 player has crashed again.
In the case of Miss Blue I suspect you are probably right. It’s quite possible I haven’t listened to that one right til the end (I just thought I ‘probably’ did). I’ll listen to it tonight.
A slight hijack, but if you use itunes, you can make 2 copies of the song, and end one of them right after the song ends, and then start the other one right before the hidden track. In the options you can choose start/end times. It creates 2 different tracks. Then you can listen to both of them, but not have the annoying silence. Hopefully that makes sense.
Lobsang, I just checked, the “Miss Blue” track just has a little bit of speaking and some yelling, starting 1:12 from the end (at about 18:36 through the track). I don’t think it’s worth keeping the last part on.
The Indigo Girls track “Faye Tucker” has a long stretch of silence after the song ends. Then (as if to add an extra layer of disguise) there’s a short reprise of part of another song from the album, “Sister.” Then there’s more silence. Finally after several minutes, there’s the hidden song: “Philosophy of Loss,” which heavily criticizes Christianity for its inequality toward LGBT people (sung by Emily Saliers, a lesbian Christian). Plus, it has a strong antiwar message. I’m guessing she wanted to release a controversial song without stirring up too much public controversy, just share an Easter egg with her dedicated listeners.
Depending upon the exact encoder you use will change the amount of space used for the track.
Essentially you can have two sorts of compressed audio encoding. Constant bit rate, and variable bit rate. (There is an intermediate - average bit rate, which is essentially variable but with a constraint to keep to an overall averge value.)
The deal is simple information theory. Most MP3s are encoded with a constant bit rate encoder (160b/s for instance.) So no matter what you encode, the rate of data creation is the same. So many seconds long, so many bits of data. Easy. The down side is that if the music is complex, the encoder may not really have enough space to compress down the nuances of the sound, so the quality degrades. And on the flip-side, if the sound is very simple - like silence - it has no mechanism to avoid spitting out lots of redundant data. So you want a variable rate. When the music is complex it creates more data, when it is simple it creates less.
One neat result is that you can create a useful lossless compressor if it is allowed to be variable rate. It is not possible to create a constant rate lossless compressor. (Well not unless you count 1:1 compression.)
I never use anything but lossless. Disk is cheap, and unless you insist on using the $2 heaphones that came with your iPod/MP3 player it is pretty easy to hear the difference. Against 160b/s it is trivial. 320b/s is pretty good, but still not there.