How to Record FPS and Frametimes with CapFrameX

Updated

CapFrameX records FPS, frametimes, and sensor data while you play a game, then lets you analyze and plot the results or share them through its cloud service. To do this you need three pieces in place: CapFrameX itself, Intel's PresentMon (the capture layer CapFrameX is built on), and RivaTuner Statistics Server (RTSS), which provides the in-game overlay. Once those are installed, you start a capture while the game is running and the tool logs the frame data for later analysis.

What you need before recording

Component Role
CapFrameX The capture and analysis tool; where you start/stop recording and review results
Intel PresentMon The underlying frame-capture mechanism CapFrameX is based on
RivaTuner Statistics Server (RTSS) Provides the overlay
A running game The capture only produces data while a game is rendering frames

The site describes CapFrameX as a "Frametimes capture and analysis tool based on Intel's PresentMon" with an "Overlay provided by Rivatuner Statistics Server." That means the actual frame timing data comes through PresentMon, and the on-screen overlay you see during play comes from RTSS — so both need to be present and working for a capture to succeed.

Recording a capture

  1. Install the prerequisites. Get CapFrameX, and make sure PresentMon and RTSS are available as described above. The site's primary call to action is a Download link, which is where you obtain the tool.
  2. Launch your game first. Capture depends on a game actively rendering frames, so start the game and get to the scene or gameplay you want to measure.
  3. Start the capture in CapFrameX. Begin recording while the game is running. CapFrameX will log FPS, frametimes, and sensor data for the duration of the capture.
  4. Play through the segment you want to measure. Keep the capture running for the period that represents the workload you care about.
  5. Stop the capture. The recorded data is then available for analysis.
  6. Analyze or share. Use CapFrameX to analyze and plot the results, or share them with friends through the cloud service.

The expected result is a captured dataset of FPS and frametime values (plus sensor readings) that you can review as plots and metrics rather than relying on a momentary on-screen number.

Understanding the metrics you'll see

CapFrameX's own blog posts explain that the tool reports several performance metrics, including frametimes, FPS, median, percentiles, and x%-low values. One featured post specifically addresses a common confusion: why analysis can show FPS values lower than what you saw in-game. This matters because the FPS number displayed live during play and the FPS computed from a full frametime capture are not the same measurement — the analysis reflects the complete distribution of frames, including slow ones, which is exactly why frametime-based metrics like percentiles and x%-low are useful for judging smoothness rather than just average speed.

Common sticking points

  • No overlay or no data: Since the overlay comes from RTSS and capture comes from PresentMon, a missing or misconfigured one of these is a likely cause of a capture that doesn't start or shows nothing.
  • FPS looks lower than in-game: This is expected behavior that CapFrameX documents directly — the analysis is derived from frametimes across the whole capture, so it can differ from the live in-game counter.
  • Capture without a game running: The tool records frames while a game runs, so there is nothing to capture if no game is rendering.

For the exact current download, installation steps, and any configuration specifics, use the Download link and the metrics-explained blog posts on capframex.com, since those reflect the tool's current state.