I recently bought some groovy new tools for sampling drums and percussion. I really want to approach this seriously and see myself doing quite a bit of this going forward. But right now I'm at the point in this new venture where I am fretting about how to name my samples. I need a robust and clear convention for finding sounds at the file search level. I'm reaching out wondering if the more experienced among you have developed a good scheme that has stood the test of time? I'm using Abobe Audition and NI Battery 4.
You need to figure out what will be the defining characteristics for you. That's something nobody else can answer. In what way will the samples that you make be an addition to the millions of samples you can already find and buy and uniquely identify them? And how do you plan on using them?
If for you it's
all about genres, e.g. you want to have a basic set of samples but for 100's of genres, then genre could be a good primary identifier.
If however it's all about dynamics you might want to name all your samples by velocity range.
Regardless of the above, you want to have a different naming scheme for your source files and your finalised ones. I.e. when constructing a sample lib, versus using those samples to make music. But again, you'll need to find the characteristics that work for you. And that depends for a large part on your process. Are you someone who keeps everything, literally everything? Or on the contrary, you record tons of stuff but delete 95% and only start twiddling and tweaking with the remaining 5%? Depending on that you'll need a different naming scheme.
The important thing is that it clearly identifies where in your process it's at. E.g.:
And here the YYYYMMDD part of the filename for instance is the recording date, since probably you'll want to have many takes for the same sample. Then the final file as you'd be using it in your sample lib could be "p-short" in a "hi-hat" directory.
But this is just an example, since again, it will mostly depend on your process and what you deem will be the defining characteristics of your samples and the way you'll be using them to make music.
I would suggest including:
- Instrument name (kick, snare, hi-hat...)
- Instrument characteristics (e.g. brand and size)
- Articulation, if applicable (tip, edge, 1/2 open hi-hat...)
- For pitched instruments, note name
- Layer (velocity). Note the libraries often include more samples in loud volume range, as they are believed to be more frequently used, for round robin or other techniques to avoid "machine gun" effect, so you may need to think of some smart numbering convention.
Because percussion loops lack a key, most people just use descriptive words and a BPM. So, loop_funkybreak_140.wav is the sort of thing I see on Splice, or in Soundpacks.
Listing the instrument in question, especially if it's different than a standard drumset is also common.
Genre of music is something else I might include.
Rock_RideCymbal_100 Rock_HiHat_90 Rock_Sixteenths_Hihat_92
Not sure how I would briefly say "this loop is a whole set, with the ride leading" versus "this is a loop of ride cymbals only".
A big part of it is context, ie the directory they live in, or pack, or other organizer. If the folder says "Rock Drumset Loops" then "hihat 140" probably means the time is being kept not on hats not that it's a solo.
One real problem is, as loops enter your library from other sources other conventions will contaminate your beautiful namespace.
it's a problem man.