Noise calculation using the peak-to-peak formula

 

Illustration of peak-to-peak noise with drift

The drift is first calculated by determining the linear regression using all the data points in the time range of a peak. The linear regression line is subtracted from all data points within the time range to give the drift-corrected signal.

The peak-to-peak noise is then calculated using the formula:

N = I_max - I_min

where
N	Peak-to-peak noise
I_max	Highest (maximum) I_x value in the time range
I_min	Lowest (minimum) I_x value in the time range
I_x	Intensity of the signal, corrected by the drift (drift is calculated using the LSQ formula)

For European Pharmacopoeia calculations the Peak-to-Peak noise is calculated using a blank reference signal over a range of -10 and +10 times W₅₀ flanking each peak. This region can be symmetrical to the signal of interest, or asymmetrical if required due to matrix signals.

 

Determination of noise from the chromatogram of a blank sample

Where

20 W₅₀ is the region corresponding to the 20 fold of W₅₀.

h_n is the maximum amplitude of the baseline noise in the 20-fold W₅₀ region.