Blank offset is a curve setting on the Quant (Origin). It uses the blank’s response as a steady background on the calibration curve.

It does not subtract a blank concentration from NativeConc. That is a different option — Blank concentration.

Related: both options share the Blank use column on the Quants grid. Blank offset works in response space. Blank concentration works after the curve is read, toward NativeConc.

Blank offset sequence Run order: blank, C1–C3, unknown, QC, next blank. Previous and next blanks bracket the calibration and samples. Sequence — blanks around cals and samples Blank use picks non-rejected Type Blank bottles. Sequence modes use run order (acquisition time). BLK-01 previous C1 C2 C3 UNK-01 QC BLK-02 next Bracket mode averages BLK-01 and BLK-02. Previous, else closest uses BLK-01 for this span.
Blanks around calibrations and samples

What it is vs Ignore / Include / Force

Origin What the fit does
Ignore Free intercept. Blanks are not used.
Include Adds a fake (0, 0) point, then fits. Still not a measured blank.
Force Forces the intercept through zero: y = m·x (linear).
Blank offset Measures blank response b, fits the standards after subtracting b through zero, then puts b back: y = m·x + b.

A blank is not a calibration level. Type Blank bottles do not become curve points. Their response (ISTD-normalized when you use ISTD) becomes the intercept.

The unknown keeps its original response and is read from the restored curve. If Blank corr is also on, Analyze still runs: it keeps Blank offset, skips concentration subtraction (that would correct twice), and writes a warning. See Blank concentration.

When to use it

On the Quants grid: set Origin = Blank offset, then set Blank use to choose which blanks supply b.

What Analyze does (step by step)

Response Y is ISTD-normalized the same way as calibration points (y = response / istdY; for ESTD, istdY = 1).

  1. Collect non-rejected Type Blank bottles (see Blank use).
  2. b = average of those blank y values. If none (or Blank use = None), b = 0 — same slope as Force. A Warnings entry records that fallback.
  3. For each calibration standard on the curve: subtract b from its y.
  4. Fit the active model through the origin on those shifted standards (linear, quadratic, or power). Average response factors and ln fits cannot use Blank offset.
  5. Put the intercept back: effective curve y = m·x + b (or the quadratic / power form with intercept b).
  6. For unknowns, QC, and other samples: use the original y and read concentration from the restored curve. Linear example: x = (y − b) / m.

Negative amounts clamp to 0, same as other origins.

Blank offset curve construction Standards sit above blank level b. Shifted down they fit a line through the origin. The restored curve is y = m x + b. An unknown uses its raw y. Blank offset — subtract b, fit through origin, restore intercept Y is ISTD-normalized the same way calibration points use Y. Unknowns keep their original response. Concentration (x) Response (y) 0 C1 C2 C3 b unknown (raw y) 1. b = average of selected blanks 2. y' = y - b (standards only) 3. fit y' = m * x (through 0) 4. restore y = m * x + b 5. unknown: x = (y - b) / m Open circles: shifted standards. Solid line: official curve.
Response plot: blank level, shifted standards, restored line, unknown on raw y

Blank use (which blanks)

Only non-rejected Type Blank bottles contribute — same idea as clearing Use on a cal point. Type Solvent is not a blank source.

A bottle is skipped when:

Blank use Selection
None No blanks. b = 0. No measured intercept.
Average (CalGroup) Average of eligible blanks in the same CalGroup as the curve. Unlabeled samples share the default group.
Average (overall) Average of eligible blanks in the whole batch, all CalGroups.
Previous, else closest The blank before the bracket span in run order; if none, the closest blank. Stays inside the CalGroup.
Bracket (previous + next) Average of the blank before and after the span. One side is enough if the other is missing; if both are missing, same fallback as Previous, else closest. Stays inside the CalGroup.

The span is the run-time range of that bracket’s calibration points and the samples that curve quantitates.

Use Average (CalGroup) or a sequence mode when groups should keep their own blanks. Use Average (overall) when one set of blanks should serve every group.

Rejected blanks never enter b.

After Analyze: what you see

Each calibration bracket stores:

Those columns are on the calibration and result grids. Every concentration from that curve shows the same blank list: one b per bracket, not a different blank per unknown.

Method editor

On the Quants grid:

  1. Origin → Blank offset
  2. Blank use → how to pick blanks (shared with Blank concentration when Blank corr is checked)

ISTD Quants stay on average response factors and do not use Blank offset.

What this is not

Next: Blank concentration, ESTD, ISTD, bracketing, method design.