diff --git a/rugnux/Rugnux.cpp b/rugnux/Rugnux.cpp
index e25479c79..c3626ed07 100644
--- a/rugnux/Rugnux.cpp
+++ b/rugnux/Rugnux.cpp
@@ -894,7 +894,9 @@ void Rugnux::PreScan(int start_image, int images_to_process, int frame_count, Ru
             // g', confidently. A move it cannot resolve is not evidence about the header either way.
             if (moved <= 3.0f * background_center_->sigma_pxl)
                 logger.Info("Beam centre check: that is under three times the fit's own sigma, so the "
-                            "file's centre is as good as this measurement can tell");
+                            "file's centre is as good as this measurement can tell - the second first "
+                            "pass below indexes it anyway, because a fraction of a pixel is enough to "
+                            "decide a cell and no sigma small enough to gate on would notice");
             else if (across && *across > need)
                 // Info and not a warning: `need` is where the FFT amplitude at a 200 A axis has fallen
                 // to 0.70, which is a floor under what matters and not a threshold for failure - it is
@@ -2224,8 +2226,13 @@ ProcessResult Rugnux::RunPipeline(RugnuxObserver *observer, bool write_output, b
             const float measured_x = measured.beam_x_pxl;
             const float measured_y = measured.beam_y_pxl;
             const float moved = std::hypot(measured_x - header_x, measured_y - header_y);
-            // A move the fit cannot resolve is not a different hypothesis, only the same one twice.
-            if (moved > 3.0f * measured.sigma_pxl) {
+            // No gate on how significant the move is. It is tempting to skip the second pass when the
+            // two centres are within the fit's own noise - it would save it on a quarter of the runs
+            // here - but that is deciding at a threshold before the evidence is in, and the evidence
+            // is cheap. It would also gate away exactly the case worth asking about: measured on real
+            // data, a 0.12 px change of centre is the whole difference between the deposited cell and
+            // a halved axis, and 0.12 px is well inside the sigma of any fit that could gate it.
+            {
                 const int majority = static_cast<int>(validation.size()) / 2;
                 experiment_.BeamX_pxl(measured_x).BeamY_pxl(measured_y);
                 FirstPass alt = pick_best(*indexer_pool, *indexer);