To BRBEATLP.958 and beyond (Epilogue)
The story so far
In order to explore the origins and prevalence of a personally-beloved drum loop sample named BRBEATLP.958, I fingerprinted and catalogued every sample of every MOD/S3M/XM/IT file in the Mod Archive so that I could hunt it down amidst tens of thousands of different songs. Now that we’ve solved that mystery, let’s wrap this up with an excursion through that data!
Disclaimer: I’ve done the best I can to clean the data I gathered and ensure its accuracy, but I’m not going to stake any rigorous claims of perfection against it. So, you know, just take all of this for what it’s worth!
What were the ten most-used samples?
Across all IT, XM, MOD and S3M files in the ModArchive torrents:
Number 10 is a tom drum, known in many sample names as Dr_Tom.
Number 9 is a closed hi-hat, most commonly named C-HIHAT.1.
Number 8 is a cymbal crash! No real clues as to the origin of this one.
Number 7 is POPSNARE2 from the ST-01 sample pack.
Number 6 is another cymbal crash! Again no clues here. Its most common name is CYMBAL.9.
Number 5, coming in at 126 bytes, is a simple chip lead.
Number 4 is the staple of many tracks, the open hi-hat. Its most common samplename is O-HIHAT.1, so I’m guessing it’s the sibling of our #9 entry.
Number 3 is a bassier chip lead, a paltry 48 bytes long.
Number 2 is a closed hi-hat, which many samplenames point to as HiHat2 in the ST-01 SoundTracker sample pack.
And the most popular sample, used in 1,592 different modules - and which I would have never, ever guessed - is the quintessential ingredient to many musical transitions: the reverse cymbal!
What were the biggest samples?
The biggest sample I found in a mod was Falseheart, although it’s reasons are fairly explainable, as it’s a large vocal track that accompanies the music.
The song with the next largest sample, and definitely the more gratuitous of the two, was ggbb.xm, which has a sample consisting of the entire 3 minutes of the 1994 Finnish Eurovision Song Contest entry, “Bye Bye Baby” by CatCat. And then proceeds to only use the first 20 seconds of it! Which is, I guess, better than zedfox_-_classic.xm, a remix of Megalovania which contains a huge portion of Dragster 5k and proceeds to not use it at all.
Purple Motion or Skaven?
Which of these two legendary Future Crew alumni are mentioned the most in sample texts?
Well, the results are so close that I think we’re going to have to just call it even:
Skaven: 415 times.
Purple Motion: 400 times
(and yes I did include skipping the space, as well as their real names)
What’s the most common sample name?
The most common sample name - with 7,429 matches - was rather the confusing ntitled (including the space).
Initially I was thinking this must have been a bug in my extraction code, but after not finding that phrase anywhere in my code - and verifying the presence of it in mod files with a hex editor - I was left a bit bewildered and started looking for an explanation.
It turns out this was very likely the result of a bug in the OpenMPT tracker which had an Impulse Tracker handling bug fixed in v1.19:
Garbage characters in sample / instrument / song names should be gone now.
This should e.g. avoid sample names like " ntitled" turning up in modules
after deleting sample names.
So, if we ignore that entry, what’s the next most common sample name? With 4,783 matches, that would be, err.. NoName!
Now, whilst my suspicions are high that it’s a similar situation of a program inserting that value as a default, I couldn’t find a culprit. So if we take a look further down the list, our third most popular sample name is… cue drum roll… the very same instrument responsible for the drum roll! That’s right, it’s SNARE.
The obligatory word cloud
So above we were considering unique sample names as the content of the entire field. What do we get if we run a word cloud generator over every individual word found in the sample texts? Well, you get something like this:
I guess bass rules the world, and I’m all for it.
The “album art”
The sample text space lets some trackers get a bit creative with their visually artistic side, so it’s not uncommon to find a few fun displays of ASCII art flair. So to sum up this epilogue, let’s celebrate a sprinkling of what I found whilst perusing the sample texts.
Will there be more to come after this? Well, probably not, unless I get some other bright ideas of what to do with this data (let me know if you’ve got some of your own!). I’ve certainly had a lot of fun with this series and I hope it was equally as enjoyable to read.
