Bracket calibration
Cal sets around samples, per-bracket curves.
Bracketing splits the sequence by acquisition time. Each cal set gets its own curve. Each stretch of samples is read from the cal set before it and the cal set after it.
When to use it
- Long sequences where the instrument drifts.
- You already run opening and closing multi-level sets (and maybe mid-sequence sets).
- You want a bad cal in set B to stay off set A.
Method Calibration mode = Bracketing. Levels and ISTD links are the same as ESTD or ISTD.
How Quanto cuts the sequence
Samples are ordered by Acquired. Contiguous calibration bottles form a cal set. Everything else (Sample, QC, Blank) forms a sample block.
| Block | Curve points |
|---|---|
| Cal set | That set only |
| Sample block | Immediately preceding cal set and immediately following cal set |
| Leading samples (no cal yet) | The first cal set only |
| Trailing samples (no cal after) | The last cal set only |
QC and blanks ride with the sample block they sit in. They do not start a new bracket.
Independent fits
Each bracket stores:
- Its own Use flags
- Its own replacement responses
- Its own coefficients, R², and residuals
Turn a point off in bracket 2 and bracket 1 is unchanged. The calibration plot for a sample shows the points of that sample’s bracket.
Set it up
- Build the method as ESTD or ISTD, multi-level.
- Set Calibration mode to Bracketing.
- Add the sequence in run order (or trust Acquired).
- Mark every standard Type Calibration and Dose C1…Cn. Repeat the same dose names in every cal set.
- Analyze.
If a sample block has no neighboring calibrations at all, that block is not quantitated.
What bracketing does not do
- It does not reuse responses saved in the method. Only live batch calibrations are fitted. Partial method update is an initial-calibration feature.
- It does not average two curves into one. A sample between A and B is fitted on the combined A+B point cloud as one bracket.
Open the result’s calibration rows to see which files belong to the current bracket.