Understand lag and optimize Firestorm
FPS, graphics, cache, Hypergrid, network and simulator: identify the real cause of slowdowns.
Identify the problem by symptoms
| Symptom | Possible cause | First test |
|---|---|---|
| Low FPS, responsive scripts | Local rendering, GPU, graphics settings | Try 64–128 m and disable shadows |
| Gray or blurry textures | Asset downloads, network or cache | Wait, test network, relog |
| Jerky movement | Packet loss, ping or jitter | Compare wired/Wi-Fi and viewer stats |
| Slow scripts for multiple people | Simulator, physics, scripts or network | Compare regions and check logs |
Understanding different kinds of lag
The word “lag” covers many symptoms: low frame rate, slow texture loading, jerky movement, failed teleports, delayed scripts or physics. Identify the symptom before assuming a cause. Two avatars in the same region can have very different FPS while the simulator runs normally.
FPS and graphics card: what happens on your PC
FPS (frames per second) measures how frequently Firestorm draws new images. It depends on the GPU (graphics processor), resolution, shadows, avatars and scene. It is not a direct measurement of server load. Quick test: visit a quieter place, reduce draw distance substantially and compare FPS.
Not everyone has the same computer
Integrated graphics usually share computing and memory resources with the rest of the PC. A dedicated graphics card typically has resources designed for 3D rendering and, depending on the model, dedicated video memory (VRAM). Performance varies widely across generations and models: not every discrete GPU is powerful, while some recent integrated GPUs perform well. CPU, RAM, cooling, drivers and screen resolution matter too.
A concrete example: two avatars in the same region
Imagine two people looking at the same OpenSimulator scene at the same moment. The first uses an office laptop with modest graphics, 240 m draw distance, shadows and many detailed avatars: they get only 12 FPS. The second has a PC with a powerful graphics card and gets 55 FPS. These figures are illustrative: the same simulator can feel very different on different computers. Low FPS for the first person alone does not prove the region is lagging. Lowering draw distance to 96 m and disabling shadows may improve their frame rate substantially.
Example settings by hardware capability
| PC type | Starting draw distance | Shadows and effects | Avatars and scene |
|---|---|---|---|
| Office PC / modest integrated GPU | 64–128 m | Shadows off, low effects | Limit rendered avatars, prioritize smoothness |
| Mid-range PC / mid-range discrete GPU | 128–192 m | Test shadows and effects gradually | Balance detail and FPS |
| Powerful PC / high-end recent GPU | 192–256 m or higher if smooth | Shadows, reflections and details may be practical | Increase based on region and avatar count |
To find suitable settings, start with short draw distance and shadows off. Note FPS in the same location, then gradually increase draw distance, shadow quality and the number of fully rendered avatars. Stop when performance becomes uncomfortable. Firestorm preference names can vary with release and language.
Draw distance: suggested starting points
Draw Distance controls the area the viewer tries to render around your avatar. Higher values often increase the number of visible objects, textures and avatars. Suggested starting points: modest computer or integrated GPU 64–128 m; mid-range PC 128–192 m; powerful system 192–256 m. A distance of 512 m or more can be demanding depending on the scene. These are starting points, not universal optimal settings.
Shadows, reflections, avatar complexity and effects
Dynamic shadows, mirrors, some reflections and post-processing can lower FPS. At busy events, a few highly detailed avatars may overload a graphics card. Reduce shadows, the maximum number of fully rendered avatars, particles and expensive effects; change one setting at a time to identify what helps your system.
Understanding the Firestorm cache
The cache stores previously downloaded assets such as textures and meshes so the viewer does not need to fetch them again every visit. Reaching the configured cache size is normal: old items are replaced automatically. Clearing it every day does not speed up the viewer; it forces assets to download again. A targeted reset can be useful if textures remain corrupted or missing despite a stable connection.
Hypergrid: why textures reload
Frequent Hypergrid travel introduces new assets from many destinations. The local cache may therefore evict previously visited content more quickly. Returning to a region after many trips can trigger fresh texture downloads; this does not automatically indicate a failure on that grid.
When and how to clear the cache
Before clearing anything, restart Firestorm, check free disk space and test a stable connection. If textures remain abnormally missing across regions, open Firestorm Preferences and look under Network & Files, then directories or cache controls; names vary by release. Use the viewer's own clearing action and restart when prompted. Initial reloading is often slower. Inventory cache is a separate troubleshooting issue.
Network: bandwidth, latency, packet loss and Wi-Fi
A fast fiber connection does not guarantee smooth interaction. Ping measures round-trip latency; high latency delays responses. Packet loss and jitter (variations in latency) can cause jerky movement and disconnects. Try wired Ethernet instead of Wi-Fi, pause large downloads and compare VPN-on versus VPN-off when possible.
Firestorm: statistics and maximum bandwidth
Open statistics with Ctrl + Shift + 1 when supported. Compare FPS, Ping SIM and Packet Loss: FPS mainly describes local rendering, while Ping SIM and Packet Loss help assess the connection to the simulator. “Maximum Bandwidth” is a viewer traffic limit, not your ISP subscription speed. Do not blindly set it to the maximum network rate; use a reasonable value and compare results.
Separating viewer, network and simulator problems
If only one user has low FPS while scripts remain responsive, start with their computer and Firestorm settings. If several users experience delayed scripts, doors or movement, investigate simulator timing, physics, scripts and networking. If only one region is affected, identify its host: distributed grids may include resident-operated regions on independent servers. Avoid blaming an entire grid before diagnosing.
Reporting a problem usefully
Include the region name, approximate time and exact symptom, FPS, draw distance and graphics settings, connection type (Ethernet, Wi-Fi, VPN), Ping SIM and Packet Loss, and whether others see the same issue. Say whether only graphics are affected or scripts and teleports are delayed too. This helps operators inspect the right logs.
One-minute diagnostic checklist
1. Low FPS? Reduce draw distance, shadows and fully rendered avatars. 2. Gray textures? Check connectivity, asset loading and cache without routinely clearing it. 3. Jerky movement? Compare Wi-Fi versus Ethernet, Ping SIM and Packet Loss. 4. Slow scripts for everyone? Check the simulator. 5. Only one region affected? Identify its host. 6. Still stuck? Provide a detailed report with time and measurements.