Update introduction

Author: rayosborn

docs/source/introduction.rst

@@ -35,11 +35,10 @@ is approximately perpendicular to the beam. Images are collected on an
 area detector placed in transmission geometry behind the sample. Many
 area detectors consist of a set of chips with small gaps between them,
 so sample rotation scans are often repeated multiple times (typically
-three) with small detector translations between each one to fill in
-these gaps. However, it is also possible to accomplish this just by
-adjusting the orientation of the rotation axis itself. *NXRefine*
-reduces the data independently for each rotation scan before merging
-them to create a single 3D data volume.
+three) to fill in the missing data, with small detector translations
+between each scan and/or changes to the orientation of the rotation
+axis. *NXRefine* reduces the data independently for each rotation scan
+before merging them to create a single 3D data volume.
 
 .. figure:: /images/experimental-geometry.png
    :align: center
@@ -187,23 +186,24 @@ from the goniometer center to the detector, at the point where the
 incident beam would intersect, is :math:`l_{sd}`. The incident beam
 wavelength is :math:`\lambda`.
 
-In the refinement procedure implemented by *NXRefine*, the orientation
-matrix, :math:`\mathcal{U}`, is generated by selecting two Bragg peaks,
-whose (*h*, *k*, *l*) values are determined using initial estimates of
-the instrument angles and the sample *d*-spacings. θ, ω, χ, and Φ are
-initially set to their nominal motor angles, while the position and tilt
-angles of the detector are estimated using a powder calibrant. It is
-assumed that the space group and approximate lattice parameters are
-known in advance, allowing an original estimate of the
-:math:`\mathcal{B}` matrix to be derived. Once the two peaks have been
+In the refinement procedure implemented by *NXRefine*, it is assumed
+that the space group and approximate lattice parameters are known in
+advance, allowing an original estimate of the :math:`\mathcal{B}` matrix
+to be derived. The orientation matrix, :math:`\mathcal{U}`, is then
+generated by selecting two Bragg peaks, whose (*h*, *k*, *l*) values are
+determined using initial estimates of the instrument angles and the
+sample *d*-spacings. θ, ω, χ, and Φ are initially set to their nominal
+motor angles, while the position and tilt angles of the detector are
+estimated using a powder calibrant. Once the two peaks have been
 selected, they are used to produce an initial estimate of
 :math:`\mathcal{U}`, from which all the other peaks are assigned (*h*,
 *k*, *l*) indices. If these assignments are reasonable, then a large
 number of peaks are used to refine both the instrumental and sample
 parameters in order to minimize discrepancies between the calculated and
 measured peak positions, allowing :math:`\mathcal{U}` to be optimized.
-If few peaks are assigned with reasonable accuracy by the selection of
-the initial two peaks, it is necessary to select two different peaks. 
+If only a few peaks are assigned with reasonable accuracy by the
+selection of the initial two peaks, it may be necessary to select two
+different peaks. 
 
 The refinement process, along with the tools that *NXRefine* provide to
 facilitate peak assignments, are described in a later section.