What happened?
Hi Developer,
I found a specific performance bottleneck in OpenNOW where the stream lags drastically only when the character or camera moves, whereas the official GeForce NOW client runs perfectly smooth at the exact same configuration and maximum bitrate.
The Symptom:
- When standing completely still: The stream is smooth, ping is low, and FPS is stable.
- The moment I move the character or rotate the camera: Ping spikes to unplayable levels and the stream suffers severe frame drops.
Technical Investigation:
Since the official client handles maximum bitrate without any performance drops on this specific low-end device, custom UI/HUD rendering is likely not the bottleneck.
The issue heavily points toward Network Thread Congestion or a Thread Blocking flaw. During camera movement, the sudden spike in incoming WebRTC video data seems to block or delay the outgoing input packet queue on the main network thread, leading to buffer bloat and artificial ping spikes.
Suggested Fixes for Codebase:
Please look into the networking/streaming loop architecture:
- Ensure the outgoing input handling system runs on an isolated high-priority thread to prevent incoming video packet bursts from delaying input transmission.
- Implement better packet throttling or non-blocking asynchronous socket queues during intra-frame data spikes.
Thank you for your hard work on this amazing project! Hope this insight helps improve stability for network packet handling.
Operating System
Android
GeForce NOW Tier
Free
Logs (required)
No exported log file available, but the issue occurs consistently during heavy intra-frame motion decoding on low-end hardware.
Screenshots (optional)
No response
What happened?
Hi Developer,
I found a specific performance bottleneck in OpenNOW where the stream lags drastically only when the character or camera moves, whereas the official GeForce NOW client runs perfectly smooth at the exact same configuration and maximum bitrate.
The Symptom:
Technical Investigation:
Since the official client handles maximum bitrate without any performance drops on this specific low-end device, custom UI/HUD rendering is likely not the bottleneck.
The issue heavily points toward Network Thread Congestion or a Thread Blocking flaw. During camera movement, the sudden spike in incoming WebRTC video data seems to block or delay the outgoing input packet queue on the main network thread, leading to buffer bloat and artificial ping spikes.
Suggested Fixes for Codebase:
Please look into the networking/streaming loop architecture:
Thank you for your hard work on this amazing project! Hope this insight helps improve stability for network packet handling.
Operating System
Android
GeForce NOW Tier
Free
Logs (required)
Screenshots (optional)
No response