WEBVTT

00:00:00.250 --> 00:00:08.192
Some published desktop labels remain in Spanish; this English
narration guides the controls. We reopen a copy of

00:00:08.192 --> 00:00:15.252
the synthetic project and go directly to Probes before selecting
an active series. This independent study

00:00:15.252 --> 00:00:20.106
recognizes all three ramps. Review the proposed reference and its
conditions.

00:00:21.106 --> 00:00:28.437
Choose peak opens the spectrum. A double click centers the window
on the synthetic signal's local maximum. Check

00:00:28.437 --> 00:00:32.918
the region and possible overlaps before using it as a reference.

00:00:33.918 --> 00:00:41.569
Calculate all slopes displays the three ramps with separate
timings and fits. This comparison does not yet

00:00:41.569 --> 00:00:46.970
require a reference diffusion coefficient. The calibrated
gradient has not been determined.

00:00:47.970 --> 00:00:54.947
The first ramp contains a deliberately perturbed observation.
Refine R² retains a minimum point count and

00:00:54.947 --> 00:01:00.617
gradient range. The search is bounded and does not guarantee
reaching its target.

00:01:01.617 --> 00:01:08.822
The proposal compares the current fit with a twenty-two-point
selection. It suggests excluding one observation

00:01:08.822 --> 00:01:16.028
without changing intensities. Review that exclusion: the current
selection remains unchanged until you apply it.

00:01:17.028 --> 00:01:25.199
Apply selection updates this ramp's fit and enables undo. A
perfect R² in synthetic data does not physically

00:01:25.199 --> 00:01:31.100
validate the probe. Review the reference, medium and temperature
before deriving a calibration.

00:01:32.100 --> 00:01:38.008
Undo refinement restores all twenty-three points and the original
fit. This walkthrough demonstrates a

00:01:38.008 --> 00:01:45.182
reversible selection. We have not saved a physical calibration or
sent a synthetic probe to the cloud.