How to sample an instrument and make it playable (the full guide)

Play and record the notes you want — ideally a few notes spread across its range, at a consistent volume, with a gap between each so they can be separated later. Then split the recording into individual notes, tune and map each to a key, loop the sustained ones, and export. Recording is the manual part; Xampler handles everything after the recording — split, pitch, loop, map, export.

Just some. Sampling every third or fourth note is normal — the sampler pitch-shifts the nearest sample to fill the gaps, and small shifts are inaudible. Recording all 88 keys is overkill for most sounds. Simple, even-toned instruments need fewer samples; ones that change a lot across their range need more. Xampler maps however many you give it.

Draw or play a series of notes in your DAW (or play them live), keep them all at the same velocity for consistent volume, and leave enough silence between notes to capture the full tail of each one. Then export that as a WAV. That single WAV — or one file per note — is what you feed into the sampler tool. Xampler accepts either.

Raw recordings often have a little silence or noise before the note and an abrupt cut at the end. Trimming tightens the start to the attack; a short fade-out stops the end from clicking. Clean edges make the instrument feel tight and click-free. Xampler trims each note to its onset and can fade the edges automatically.

The root key is the note a sample was actually recorded at — its true pitch. The sampler uses it as the reference: on the root key the sample plays unshifted; on nearby keys it shifts relative to the root. If the root is wrong, the whole instrument plays out of tune. Xampler detects the root from the audio so the mapping is in tune automatically.

Only the ones you'll hold longer than the recording. Sustained instruments (pads, strings, organs) need a loop so the tone continues as long as you press the key. Short or percussive sounds can stay as one-shots. Xampler finds a sustain loop where it's needed and leaves short sounds as one-shots.

Depends on what will play it. For software samplers, SFZ (text + WAVs) or SF2 (one file) are the common choices. For Korg Pa/Kronos/Triton hardware, you need KMP/KSF. Xampler exports all of them from the same project, so you pick whatever your target reads.

Yes — that's the whole point of an auto-mapping tool. Instead of chopping, tuning, looping and mapping each note manually, you drop your recordings in and it builds the instrument. Xampler does the split, pitch detection, looping and key mapping automatically, in the browser, then exports SFZ/SF2/Korg — turning an afternoon of editing into a couple of minutes.

Yes — a voice is just another sound source. Record a sustained vowel (or a few notes of it), and the same chain applies: split, detect pitch, loop, map. Then you can play your voice chromatically. Xampler treats a vocal sample like any other pitched sample and maps it across the keys.