Fluera

Troubleshooting

Performance tuning

Fluera is built to keep 60 FPS on canvases with hundreds of thousands of strokes. Here is what to check when it doesn't.

Updated 20 April 2026

Fluera’s engine is designed to stay at 60 FPS with sub-15 ms stroke latency on modern hardware — up to hundreds of thousands of strokes per notebook. When it doesn’t, one of a small set of things is usually responsible.

Diagnostics

Open Settings → Advanced → Performance. You’ll see:

  • Current FPS — should hover at 60 (or 120 on ProMotion devices)
  • Stroke latency — should read <15 ms on iOS/macOS, <20 ms on other platforms
  • Memory used — absolute and as percentage of your device’s budget
  • Tile cache hit rate — should be >90 % during normal zoom/pan

If anything here is red, the likely culprits are in the list below.

Large canvases (> 200k strokes)

The engine is fine with this volume, but two settings matter:

  • Memory budget. On devices with less than 6 GB of RAM, Fluera’s default memory budget is conservative. If you have lots of RAM and want faster zoom responses, raise it in Settings → Canvas → Memory budget to Generous.
  • Level of detail. The default LOD policy simplifies strokes at very low zoom levels. If you notice jittery-looking strokes when zoomed out, switch LOD to High detail — but expect ~10 % more GPU load.

Audio-synced recording (Time Travel)

Time Travel stores audio alongside stroke timestamps. For a 60-minute lecture, that’s roughly 20 MB of audio. It’s not the audio that slows things down — it’s the index. Try:

  • Settings → Time Travel → Index granularity set to Standard rather than Fine
  • Clearing old Time Travel sessions you don’t need (they are kept locally by default; the clear-cache button removes them safely)

Apple Pencil specifically

On iPad, the default Predicted touch mode keeps stroke latency under 8 ms. If you’ve disabled prediction (in Settings → Canvas → Stroke latency → Standard), put it back to Predicted. Prediction is an Apple Pencil API, not a guess — it uses the hardware’s own trajectory prediction.

When nothing helps

Run Settings → Advanced → Export diagnostics. You get a small JSON file with FPS samples, memory usage, and (anonymous) device information. Email it to support@fluera.dev with a short description of when things slow down.

We log ≈95 % of user-visible perf regressions to a specific piece of code within one hour of receiving a diagnostics export. The remaining 5 % are genuinely hard and we’ll tell you so.

Next