No instrument reads a perfectly clean zero — even a blank sample with no analyte
produces a fluctuating baseline signal from electronic and optical noise. Analytical
chemists quantify that noise as sigma (σ), the standard deviation of
repeated blank measurements, and compare it against the slope (S) of a
calibration curve built from standards of known concentration. The IUPAC limit of
detection, LOD = 3.3σ/S, is the smallest concentration reliably
distinguishable from the noise floor; the limit of quantification,
LOQ = 10σ/S, is the smallest concentration that can be reported as a
reliable number rather than just "present."
A signal sitting right at the blank's average is not proof of "nothing there" — it is only below LOD, meaning any real analyte at that level is statistically indistinguishable from noise. That's why regulatory reports for trace contaminants are written as "< LOD" or "< LOQ" rather than as a false zero.
A 3D spectrometer measures samples through real baseline noise while building a live calibration curve from standards — watch the IUPAC limit of detection and limit of quantification appear on the graph as the sample signal rises out of the noise.
Every real instrument has baseline noise, quantified as sigma (σ), the standard deviation of repeated blank measurements. Comparing σ against the calibration curve's least-squares slope (S) gives the IUPAC limit of detection (LOD = 3.3σ/S) and limit of quantification (LOQ = 10σ/S) — the concentrations below which a signal can't be reliably called "present" or "measured," respectively.
Adjust instrument noise and test-sample concentration, then click Add calibration point to record a fresh noisy standard and refit the calibration line. Watch σ, the slope, and LOD/LOQ update live, and see the test sample's diamond marker shift color as it crosses the LOD and LOQ lines on the graph.
A measurement right at the blank's own average isn't proof nothing is there — it's simply below LOD, meaning any real analyte at that level is statistically indistinguishable from noise, which is why trace-contaminant reports are written as "< LOD" rather than as a false zero.