Automating SBaGen → GNaural Conversion: Tools and Tips

SBaGen to GNaural Converter — Quick Guide and Best Practices

What it does

• Converts SBaGen (.sbg) soundscape scripts into GNaural-compatible input so GNaural can generate the corresponding binaural audio.

Quick steps (presumes SBaGen script ready)

1. Inspect SBaGen script: note global settings (sample rate, buffer size, number of channels), instrument/voice definitions, and event timing/positioning.
2. Map SBaGen constructs to GNaural equivalents:
   • SBaGen Instruments → GNaural sources (sine/noise/filtered sources).
   • Events (note on/off, envelopes) → GNaural source start/stop and level envelopes.
   • Spatialization: convert azimuth/elevation/radius to GNaural’s source position format.
   • Timing: convert tick-based timing or BPM into seconds using SBaGen’s tempo formula.
3. Convert sample rate/format: ensure output sample rate matches GNaural project (resample if needed).
4. Generate GNaural input file: create GNaural’s plain-text source and event blocks with correct syntax.
5. Validate: run GNaural in dry mode or generate a short test render to check event timings and spatial placement.
6. Iterate: tweak envelopes, filters, and panning to match original SBaGen behavior.

Best practices

• Preserve tempo/tick math explicitly: document conversions from SBaGen ticks to seconds to avoid timing drift.
• Use conservative defaults: if an exact instrument mapping doesn’t exist, choose the closest GNaural source and note differences.
• Batch conversions: script the mapping in Python/Perl for many files; parse SBaGen tokens and emit GNaural blocks.
• Keep metadata: store original SBaGen comments or a header in the GNaural file for traceability.
• Test with isolated events: render single-source tests to verify spatial and filter behavior before full renders.
• Handle envelopes carefully: SBaGen and GNaural may use different envelope curve shapes—approximate visually or with small-file auditioning.
• Normalize levels: match RMS or peak levels between systems to avoid clipping or unexpected quiet output.
• Document differences: create a short compatibility note listing unsupported SBaGen features and suggested workarounds.

Common pitfalls and solutions

• Mismatched timing: ensure tempo/tick conversion is exact; add a small test tone at known ticks to verify alignment.
• Spatial inversion: coordinate systems may differ (left/right or front/back); flip axes if sources appear mirrored.
• Missing effects: SBaGen filters or FX may not map 1:1—recreate with GNaural filters or post-process with an external DAW.
• Channel-count mismatches: ensure GNaural project supports binaural rendering and matches intended output channels.

Minimal example (conceptual)

• Parse a single SBaGen event: instrument 1 at tick 480, azimuth 30°, elevation 0°, duration 240 ticks → convert tick times to seconds, create GNaural source with matching frequency/envelope and a start/stop event at computed seconds, set position to azimuth/elevation.

Tools and automation

• Use a scripting language (Python recommended) with a tokenizer for SBaGen syntax and a template engine for GNaural output.
• Add unit tests: round-trip short SBaGen examples and compare rendered waveforms or key markers.

If you want, I can:

• Produce a small Python script that converts simple SBaGen constructs to GNaural, or
• Convert a specific SBaGen file you provide into a GNaural file.