diff --git a/src/ess/reduce/live/raw.py b/src/ess/reduce/live/raw.py
index 7dd119e5..358ae7d4 100644
--- a/src/ess/reduce/live/raw.py
+++ b/src/ess/reduce/live/raw.py
@@ -110,7 +110,8 @@ def from_coords(
     def __call__(self, da: sc.DataArray) -> sc.DataArray:
         self._current += 1
         coords = self._coords[self._replica_dim, self._current % self._replicas]
-        return sc.DataArray(da.data, coords=coords).hist(self._edges)
+        # If input is multi-dim we need to flatten since those dims cannot be preserved.
+        return sc.DataArray(da.data, coords=coords).flatten(to='_').hist(self._edges)
 
     def input_indices(self) -> sc.DataArray:
         """Return an array with input indices corresponding to each histogram bin."""
@@ -247,6 +248,7 @@ def __init__(
         self._cache = self._history.sum('window')
 
     def make_roi_filter(self) -> roi.ROIFilter:
+        """Return a ROI filter operating via the projection plane of the view."""
         norm = 1.0
         if isinstance(self._projection, Histogrammer):
             indices = self._projection.input_indices()
diff --git a/tests/live/raw_test.py b/tests/live/raw_test.py
index b34f8a05..6aa997de 100644
--- a/tests/live/raw_test.py
+++ b/tests/live/raw_test.py
@@ -222,3 +222,96 @@ def test_histogrammer_input_indices() -> None:
     assert set(indices.coords) == {'x', 'y'}
     assert indices.bins.size().sum().value == nx * ny * nz
     assert indices.sizes == resolution
+
+
+def test_ROIFilter_from_trivial_RollingDetectorView() -> None:
+    detector_number = sc.array(
+        dims=['x', 'y'], values=[[1, 2, 3], [4, 5, 6]], unit=None
+    )
+    view = raw.RollingDetectorView(detector_number=detector_number, window=2)
+    roi_filter = view.make_roi_filter()
+    data = detector_number.copy()
+    data.unit = 'counts'
+    flat = data.flatten(to='detector_number')
+
+    result, scale = roi_filter.apply(data)
+    # ROIFilter defaults to include nothing
+    assert sc.identical(result, flat[0:0])
+    assert sc.identical(scale, sc.zeros(dims=['detector_number'], shape=[0]))
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(1, 2)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(result, flat[3:6])
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[3]))
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(1, 2), y=(1, 3)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(result, flat[4:6])
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[2]))
+
+
+def test_ROIFilter_from_RollingDetectorView_with_LogicalView() -> None:
+    logical_view = raw.LogicalView(select={'z': 0})
+    detector_number = sc.array(
+        dims=['x', 'y', 'z'], values=[[[1, 2], [3, 4]], [[5, 6], [7, 8]]], unit=None
+    )
+    view = raw.RollingDetectorView(
+        detector_number=detector_number, window=2, projection=logical_view
+    )
+    roi_filter = view.make_roi_filter()
+    data = detector_number.copy()
+    data.unit = 'counts'
+    flat = data['z', 0].flatten(to='detector_number')
+
+    result, scale = roi_filter.apply(data)
+    # ROIFilter defaults to include nothing
+    assert sc.identical(result, flat[0:0])
+    assert sc.identical(scale, sc.zeros(dims=['detector_number'], shape=[0]))
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(1, 2)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(result, flat[2:4])
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[2]))
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(1, 2), y=(1, 3)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(result, flat[3:4])
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[1]))
+
+
+def test_ROIFilter_from_RollingDetectorView_with_xy_projection() -> None:
+    detector_number = sc.array(
+        dims=['x', 'y', 'z'], values=[[[1, 2], [3, 4]], [[5, 6], [7, 8]]], unit=None
+    )
+    nx, ny, nz = 2, 2, 2
+    coords = raw.project_xy(
+        make_grid_cube(nx=nx, ny=ny, nz=nz, center=(0.0, 0.0, 10.0))
+    )
+    coords = sc.concat(
+        [coords.fold(dim='point', sizes=detector_number.sizes)], 'replica'
+    )
+
+    resolution = {'x': 4, 'y': 4}
+    histogrammer = raw.Histogrammer.from_coords(coords=coords, resolution=resolution)
+    view = raw.RollingDetectorView(
+        detector_number=detector_number, window=1, projection=histogrammer
+    )
+    roi_filter = view.make_roi_filter()
+    data = detector_number.copy()
+    data.unit = 'counts'
+    flat = data.flatten(to='detector_number')
+
+    result, scale = roi_filter.apply(data)
+    # ROIFilter defaults to include nothing
+    assert sc.identical(result, flat[0:0])
+    assert sc.identical(scale, sc.zeros(dims=['detector_number'], shape=[0]))
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(0, 2)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[4]))
+    assert sc.identical(result, flat[:4])
+
+    roi_filter.set_roi_from_intervals(sc.DataGroup(x=(0, 2), y=(0, 2)))
+    result, scale = roi_filter.apply(data)
+    assert sc.identical(scale, sc.ones(dims=['detector_number'], shape=[2]))
+    assert sc.identical(result, flat[0:2])