Knowledge Base

Problem

Users have questions about interpreting FaceReader's data output, including: what the Quality metric means and when data is unusable, how downsampling works when exporting at lower frame rates, and how HRV values (RMSSD, SDNN) are calculated.

Details

Covers common data interpretation questions arising during research analysis. Affects users at the data export and analysis stage across multiple research contexts.

Solution

**Quality metric:** The Quality value indicates how reliably FaceReader detected and modeled the face in each frame. Low quality scores mean the face was poorly captured; data from frames below the acceptable quality threshold should be treated with caution or excluded. Contact Noldus Support for the specific threshold recommended for your version.

**Downsampling:** When exporting analyzed data at a lower frame rate (e.g., reducing from 30 Hz to 1 Hz), FaceReader averages the scores across the frames within each output interval — it does not simply skip frames.

**HRV calculations:** FaceReader 10 calculates RMSSD and SDNN and supports both standard short-term (5-minute) and ultra-short-term (under 5 minutes) windows. The exact windowing implementation (trailing vs. cumulative) for custom periods is not fully documented; contact Noldus Support for clarification specific to your analysis design.