# Input Latency & Mouse Polling | Peripherals Lab | FPS Chaser Canonical: https://fpschaser.com/input-latency-mouse-polling-guide/ > 8000Hz polling physics, a mouse polling meter that reads your browser's pointer events, a network latency calculator, Rapid Trigger keyboards, DyAc 2 vs OLED. ## Input Latency & Peripherals Lab Live Mouse Polling Meter, Network Latency RTT Engine, 8000Hz Physics, Rapid Trigger & Motion Blur Reduction Live Hardware Polling Rate & Packet Interval Tester ### Move your mouse continuously inside the sensor pad Current Frequency 0 Hz Peak Frequency 0 Hz Average Interval 0 ms Packet Jitter (StdDev) 0 ms SWIPE MOUSE RAPIDLY IN CIRCLES HERE Detected Tier: Idle / Move Cursor Across Pad **Note on Browser Polling Caps:** Standard web browsers poll mouse events synced with screen refresh unless using high-resolution pointer events. If your 1000Hz/4000Hz mouse caps out at your monitor refresh rate (e.g., 240Hz), disable Chrome/Brave "Smooth Scrolling" or ensure rapid, continuous circular swipes to saturate the USB packet buffer. ### The 8000Hz Mouse Dilemma At 8000Hz, a gaming mouse sends an interrupt to the CPU every **0.125ms (125 microseconds)**. While this reduces sensor input delay from 1.0ms down to 0.125ms, it bombards your operating system kernel with 8,000 Hardware Interrupt Requests (IRQs) per second. The Minimum DPI Saturation Rule If you play on **400 DPI** and swipe your mouse at 10 inches per second, your optical sensor only generates 4,000 counts per second. You will *never* reach 8000Hz! Min DPI Formula = (Target Polling Hz) / (Hand Speed in IPS) Rule: Set your mouse to **1600 or 3200 DPI** and reduce in-game sensitivity proportionally to fully saturate 4000Hz / 8000Hz polling packets without aim jitter. ### Rapid Trigger & Hall Effect Keyboards Traditional mechanical switches have a fixed actuation point (e.g. 2.0mm) and require the switch to return past a fixed reset point before a new keypress registers. Mechanical Reset Distance: 1.5mm - 2.0mm travel Hall Effect Rapid Trigger: 0.1mm dynamic travel Counter-Strafe Stop Window: 35ms - 55ms faster In CS2 and Valorant, counter-strafing with Rapid Trigger instantly halts movement the millisecond your finger lifts, achieving immediate 100% first-bullet accuracy without dead-zone delay. ### Motion Blur Reduction: DyAc 2, ULMB 2 & OLED Physics Human eyes track moving targets on sample-and-hold LCDs, smearing the image across the retina (MPRT). Backlight strobing turns the backlight off while pixels transition, strobing only when the image is sharp. DyAc 2 / ULMB 2 Strobing Provides CRT-like clarity. 240Hz with DyAc 2 delivers the motion clarity of a non-strobed 500Hz display without requiring unachievable GPU framerates. **The Golden Rule:** Framerate MUST strictly lock to refresh rate (e.g., 240 FPS on 240Hz). If FPS drops, double-image strobe crosstalk ruins tracking. 0.03ms OLED Response OLED pixels switch instantaneously (0.03ms GtG) with zero overshoot ghosting. While non-strobed, high refresh OLEDs (240Hz - 480Hz) deliver pristine pixel transitions without backlight strobing flicker. Best for players sensitive to strobing flicker or headaches during extended 6-hour scrim sessions. 4:3 Stretched Resolutions Running 1280x960 or 1440x1080 stretched horizontally expands character models by ~33% in visual angle, making micro-adjustments easier while slashing GPU fillrate render times. Result: +20% to +35% higher 1% low frame rates compared to native 1080p/1440p rendering. ### Audio Pipeline & Realtek DPC Elimination Motherboard Realtek HD Audio drivers (`RTKVHD64.sys`) are notoriously responsible for high kernel DPC spikes causing random micro-freezes. The Problem: Bloated Realtek Suites - Realtek Audio Console background services continually query kernel hardware buffers. - Sonic Studio, Nahimic, and Dolby Atmos audio enhancements inject audio pipeline latency. - DPC spikes exceeding 1500µs cause audio dropouts and 1% low frame stutter. The Pro Solution: External USB DAC - Use an inexpensive $9-$25 USB-C to 3.5mm DAC (Apple Dongle or Tempotec). - Runs plug-and-play on Microsoft's clean USB Audio 2.0 class driver with **0.00ms DPC overhead**. - Set format in Windows Sound Properties to **24-bit, 48000 Hz (Studio Quality)**. Packet Dynamics & Netcode Lab ### Network Latency & Round-Trip Time (RTT) Engine Calculate physical photon transit in fiber optics, ISP routing hops, last-mile media overhead, and peekers' advantage. Distance to Game Server km Common Regional Matchmaking Presets: Last-Mile Physical Medium Fiber Glass (GPON / XGS-PON) Fastest medium on earth. Photons travel inside glass core at ~204,000 km/s with zero RF interference. Local Bufferbloat / Load 0ms unqueued latency (CAKE QoS configured). Game Netcode & Tick Rate Server frame step: **4 ms** (CS2 Sub-Tick (Instant Input Timestamp)) Estimated In-Game RTT (Ping) 8.4 ms LAN-Caliber Tier Virtually instantaneous hit registration. Peekers hold near-zero latency advantage. Server Action Lag 10.7 ms Click to Server Tick Peekers' Advantage 10.4 ms Reaction time desync Packet Jitter: **±0.2 ms** Effective Fiber: **708 km** RTT Latency Decomposition Total: 8.4ms Fiber Transit: **6.93ms** ISP BGP Hops: **0.71ms** Last Mile Media: **0.8ms** Bufferbloat: **0ms** The Physics of Fiber Light Light in a vacuum travels at 300,000 km/s. Inside standard single-mode silica glass optical fiber (`SiO₂`), the refractive index slows photons to **~204,000 km/s** (approx **4.9 µs per kilometer**). Eliminating Bufferbloat (SQM) When another device uploads a video or runs a speedtest, unshaped ISP router buffers overflow, adding **30ms–150ms** of queue latency. Enabling **CAKE / FQ-CoDel Smart Queue Management** guarantees 0ms added bloat. The Wi-Fi / Powerline Fallacy Even with 800 Mbps speedtests, Wi-Fi and Powerline suffer from periodic channel re-contention and inductive noise spikes that inject **±12ms packet jitter**. Direct Cat6/Cat6A Ethernet remains the undefeated standard. Architecture Shootout Lab ### Why Monolithic Ring-Bus + Tuned B-Die Dominates 1% Lows Normalized comparison across competitive titles with identical RTX 3080 graphics. Project 9900K (Tuned B-Die + 5.0GHz) i9-9900K @ 5.0GHz Sync • DDR4-3800 CL14 Tight B-Die • RTX 3080 • Ring 4.7GHz $870 Average FPS 412 1% Low FPS 248 1% Consistency 60% Frame-Time Variance: **0.82 ms** Score: 92/100 Stock i9-9900K (Default XMP) i9-9900K Stock Turbo • DDR4-3200 CL16 Auto XMP • RTX 3080 • Ring Auto $870 Average FPS 310 1% Low FPS 155 1% Consistency 50% Frame-Time Variance: **2.45 ms** Score: 71/100 Ryzen 7 7800X3D (3D V-Cache Flagship) 7800X3D PBO -20 • DDR5-6000 CL30 EXPO • RTX 3080 (Normalized GPU) $1650 Average FPS 465 1% Low FPS 275 1% Consistency 59% Frame-Time Variance: **0.65 ms** Score: 98/100 Core i9-14900K (Hybrid Architecture) 14900K Power Limited 253W • DDR5-7200 CL34 • RTX 3080 (Normalized GPU) $2150 Average FPS 485 1% Low FPS 260 1% Consistency 54% Frame-Time Variance: **0.95 ms** Score: 95/100 Peripherals & Input Latency Lab ## Peripherals & Polling FAQs Insights on 8000Hz USB polling, Rapid Trigger magnetic keyboards, OLED motion clarity, and audio latency. Ctrl+F An 8000Hz mouse sends an interrupt packet every 125 microseconds. Front-panel case ports pass through internal ribbon cables and shared internal headers that add electrical noise and signal jitter. Motherboards have USB ports connected directly to the CPU integrated controller and other ports wired through the motherboard chipset hub. Check your motherboard manual to identify the direct CPU-connected USB ports (usually the top rear USB 3.2 Gen 2 ports) to ensure mouse packets do not queue behind other USB devices. Key Takeaway: Always plug high-polling mice into direct CPU-routed rear I/O ports—never into USB hubs, monitors, or front-panel ports. [Live Mouse Polling Rate Tester](https://fpschaser.com/input-latency-mouse-polling-guide/) Need more details on this optimization layer? [Master Zero-Latency Checklist](https://fpschaser.com/fps-optimization-checklist/)