How to explain data compression with lyrics
I'm making an educational video on data compression that will benefit from small doses of humor. Since most lossless compression works on the principle of repetition reduction , I would like to illustrate this idea in the form of songs.
For example, the left side of the following image is an uncompressed song (snippet) from Sia. The compressed version skips the repetitions, causing them as such.
I would say it demonstrates this type of compression technique quite clearly. But it shows that compression adds complexity (due to tags and referencing of repeated subsequences). It's also not overly compressed, which leaves it, so I can challenge viewers to explore additional ways to compress text (i.e., remove whitespace, etc.).
But there is a problem! The lyrics are boringly real (not to mention copyright). So I thought it would be more interesting to make a caricature of the lyrics that Sia could write and then sell to another artist (say Beyoncé). I thought Stackoverflow might help. Reply to this post with a lyrical caricature that:
- Suitable for twelve year olds (PG)
- Repeats in it so that the "compressed" version actually looks smaller
- Originally yours (not copyrighted)
- Makes pop jokes ( because that's a lot). Figure A
- Helps to communicate the idea better (i.e. demonstrating lossy compression?)
In response, I:
- Read all answers.
- Give all my rep (at the time of accepting the answer) up to 1500 points (Stackoverflow bail limit) to whoever posts the funniest or highest answer (discretion is mine).
- The award points to some time on April 11th.
- Give "winning" lyric credit in my tutorial video
- Link to video here
Answer if you are greedy for glasses, but also believe that STEM education should be more attractive
Update 1
I have rewarded my whole reputation
source to share
Ok ... so ... uh ...
Here's my try, fry me if you want. Now I want to notice the flow and still learn ... I can try to write more lyrics if you want, not sure if I understood all the terms correctly.
Since I'm from Germany and listen to hard gangster rap (;)), I just don't know how American songs are usually structured. I'll try.
Sia, listen, sit down we need to talk.
I know it not easy in your position
managing business and the like
check your lyrics -> 0 rhymes
not even a spark of inspiration on your mind.
take my hand, let go for a walk...
To be honest I have never thought
that you'd have more than 500 points on this site... jeez...
But while we're here, let not forget,
being a developer is something I personally regret
I would rather write songs, keep the flow...
I'm like this kid... libmozart... well, now you know...
anyways, I almost forgot – look at the clock
it time to talk about the plot in this track.
Sia, do you ever think about why there are old classic dvds
with around 10 songs and a few beats
while there are a few dozen CDs
with around 3000 songs each? mostly pop songs indeed...
because it really really cheap, to store pop songs using zlib
listen,
You wanna know why?
You wanna know why?
gzip turns your repeated lyrics into just a handful of bytes
to reduce the size it not magic am I right?
you all sound familiar am I right?
jeez, look at your eyes – cold shiver.
even an algorithm doesn't find anything that differs.
I guess I've covered all the points, should I go into more detail about compression loss? For compression, there are several whole sentences that can be compressed (do you want to know why?) And a few words (around, song, forgiveness, jeez, etc.). It is also 4 hours, so every spelling error is due to lossy compression.
If you're looking for an Eminem developer, you can publish it yourself or add extra coffee to your reward.
source to share