api-load-sample

name	api-load-sample
description	Load a sample from the open project into a tracker instrument slot, then optionally conform it to a specific row count. Walks subdirectories. Project-only — fails with noProject if no project is open.

Load a sample via the Local API

When to use

The user has files in <project>/samples/ (or wants to know what's there) and wants to bring one into an instrument slot, optionally snapping its length to a number of rows.

Project-required

Every call in this flow returns { ok: false, errors: [{ code: "noProject" }] } if no project is open. Tell the user to open one via PROJECT > EXPLORER. Don't retry.

Three steps

# 1. See what's available (recursive — includes subfolders)
nanotracker_assets_list

# 2. Load a file into a slot (1..31). fileName can contain "/" for subfolders —
#    pass the full 'name' field from list() verbatim.
nanotracker_assets_load { "slot": 1, "fileName": "Drums/break.wav" }
nanotracker_assets_load { "slot": 2, "fileName": "kick.wav" }

# 3. (Optional) Snap a loop to N rows at the project's bpm/speed
nanotracker_execute {
  "commands": [{ "op": "conformSampleToRows", "sampleId": 1, "rows": 16 }],
  "opts": { "undoDescription": "conform amen to 1 bar" }
}

Subdirectories

assets.list walks <project>/samples/ recursively (max depth 8) and returns entries with three name-like fields:

{
  "name":      "Drums/break.wav",              // full path, forward-slash, pass to load
  "leafName":  "break.wav",                    // just the filename
  "directory": "Amens",                        // parent folder (empty at top level)
  "size":      1494512,
  "lastModified": 1473590967000
}

For a library organised by pack (e.g. Drums/, Basses/, One-Shots/Synths/) the list is flat — one entry per file — but the directory field makes it easy to filter or group in your own code.

conformSampleToRows computes the stretchRatio so the sample plays for exactly rows rows at the project's current bpm/speed. Useful for drum loops that should fit one bar (16 rows by default), two bars (32 rows), etc.

Direct stretch

If you know the exact ratio:

{ "op": "setSampleStretchRatio", "sampleId": 1, "ratio": 2 }

Plays at 2× speed. ratio: 0.5 is half speed. Range 0.01..100.

Other sample-metadata edits

{ "op": "renameSample",  "sampleId": 1, "name": "KICK" }
{ "op": "setSampleMeta", "sampleId": 1, "patch": {
    "baseNote": 60, "finetune": 0, "volume": 64, "pan": 128,
    "loopStart": 0, "loopLength": 0
  } }

baseNote: 60 (MIDI middle C) means tracker note 49 plays the sample at its natural pitch. Higher tracker notes pitch up, lower notes pitch down.

Unloading

nanotracker_assets_unload { "slot": 1 }

Frees the audio engine voice and removes the sample. The instrument table slot resets to a default empty entry.

Out of scope

• Uploading new sample files via the API. Files have to be in <project>/samples/ already. The user adds them via PROJECT > EXPLORER (which uses the browser's File System Access API).
• Loading from URLs. Same reason.

Common mistakes

• Calling assets.load before checking nanotracker_health. If the page isn't connected, you'll get an opaque error.
• Forgetting undoDescription on the conformSampleToRows execute batch. Required.
• Confusing slot (1..31, instrument table) with sampleId (the loaded sample's id, which is normally equal to the slot but may differ in unusual projects).