Traditionally the filenames for samples follow a basic format which includes the articulation, velocity layer, and note name. These labels are usually interchangeable--for example:

"Sustain_Soft_D4"

"C#3-Slide-Loud"

While this file naming convention includes all the elements to identify and map samples, it is far from efficient when dealing with hundreds, even thousands of samples. To speed up the mapping process, some sample development products have added an auto-mapping feature, which attempts to automatically map samples based on their file names. The downside is that generally your file naming has to conform to the way the auto-mapping identifies samples, so the amount of work it saves you varies.

Over the years, I have developed a more practical system of file naming, which does not rely on auto-mapping.

If possible, the samples should be saved in a folder structure rather than all within the same directory. The hierarchy of folders represents the main elements in each sample, rather than having each element within the file name itself. This keeps the samples neatly organized as well as reducing the length of file names. The note name label should always be at the very end of the file structure.

The key to fast and easy mapping is to use note numbers rather than note names--that way the samples sort in alphabetic (well, technically alphanumeric) order correctly. Otherwise, using note names, an A in one octave will be sorted before a C in the same octave. If you use note numbers, the lowest number corresponding with the lowest note, the samples are guaranteed to sort correctly. For best results offset numbers with fewer digits by zeros to maintain their order--so use "001" instead of just "1".

If you use note numbers in a sampler such as Kontakt, mapping the notes is as easy as selecting all of the WAV samples and dragging them into the mapping view. Kontakt will automatically map these samples in their correct order.

The rest of the folder structure should represent the other variables in each sample such as velocity, round-robin, and articulation. Here's a final example of a sample library's folder structure:

Root directory
> Samples
>> Sustain
>>> Soft
>>> Medium
>>> Loud
>>>> Round-Robin 1
>>>> Round-Robin 2
>>>>> 01.wav
>>>>> 02.wav
>>>>> 03.wav
>>>>> 04.wav
>>>>> 05.wav

etc.

If you prefer that the individual samples, despite their organization within the folder structure, have more unique names, you can put a descriptive label after the note number. In the previous example, the file names could be: "01-RR2-L-Sus.wav", which uses initials to reduce the file name length, but still making the samples' names unique.

I hope this method for sample naming proves useful for your sampling process! Please leave a comment if you have any suggestions or improvements on this method.


Enjoyed reading this blog article? Here are some related topics with more articles you'll like: Articles, DIY Sampling


Share with Your Friends