Skip to content

FIX: Improve channel name rendering in plot_alignment #13773

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
aman-coder03 wants to merge 5 commits into
base: main
Choose a base branch
from
+22 −7
Open

FIX: Improve channel name rendering in plot_alignment #13773

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
afe6247
reduce text scale and improve contrast
aman-coder03 Mar 21, 2026
a9eca8f
switch text color back to white
aman-coder03 Mar 22, 2026
6ab310d
use add_point_labels for show_channel_names
aman-coder03 Mar 22, 2026
11462b9
add changelog entry
aman-coder03 Mar 25, 2026
e834a28
address review comments
aman-coder03 Mar 26, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions doc/changes/dev/13773.bugfix.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
Improved channel name label rendering in :func:`mne.viz.plot_alignment` when using ``show_channel_names=True`` by reducing overlap, adding a shadow for contrast, and offsetting labels outward from sensor positions, by `Aman Srivastava`_.
Copy link
Copy Markdown
Member

@drammock drammock Apr 1, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
Improved channel name label rendering in :func:`mne.viz.plot_alignment` when using ``show_channel_names=True`` by reducing overlap, adding a shadow for contrast, and offsetting labels outward from sensor positions, by `Aman Srivastava`_.
Make channel name labels easier to read in :func:`mne.viz.plot_alignment`, by `Aman Srivastava`_.

28 changes: 21 additions & 7 deletions mne/viz/_3d.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -955,13 +955,27 @@ def plot_alignment(
# transform to current coord frame
pos = apply_trans(to_cf_t["head"], pos)

for ch, xyz in zip(chs, pos):
renderer.text3d(
*xyz,
ch["ch_name"],
scale=0.005,
color=(1.0, 1.0, 1.0),
)
# offset labels outward from centroid so they clear the sensor glyphs
centroid = pos.mean(axis=0)
directions = pos - centroid
norms = np.linalg.norm(directions, axis=1, keepdims=True)
norms = np.where(norms > 0, norms, 1)
offsets = pos + 0.01 * directions / norms

labels = [ch["ch_name"] for ch in chs]
renderer.plotter.add_point_labels(
offsets,
labels,
font_size=10,
text_color=(1.0, 1.0, 1.0),
font_family="renderer.font_family",
shadow=True,
show_points=False,
shape=None,
always_visible=True,
Copy link
Copy Markdown
Contributor

@scott-huberty scott-huberty Mar 25, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have concerns about circumventing mne.viz.backends._pyvista.PyVistaRenderer.text3d and reaching straight for renderer.plotter.add_point_labels, because the aforementioned text3d does some checks/sanitizing, e.g.:

mne-python/mne/viz/backends/_pyvista.py

Lines 755 to 756 in ba12118

if "always_visible" in signature(self.plotter.add_point_labels).parameters:
kwargs["always_visible"] = True

@drammock WDYT?

Copy link
Copy Markdown
Member

@drammock drammock Mar 25, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

for PyVista questions I always ask @larsoner 😆 But looking at the PyVista docs, I don't see any indication that add_point_labels is especially low-level or dangerous. That said, I agree that since we seem to have taken pains to add a wrapper for it, it was probably for a good reason and we should probably use the wrapper.

Copy link
Copy Markdown
Contributor

@scott-huberty scott-huberty Mar 26, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK! - my specific concern is just that here we hardcode always_visible=True whereas in text3d we check to see if that parameter exists (given the users version of PyVista(??..) ). Maybe that check is outdated and our lower pin guarantees the existence of this parameter, but someone would need to check...

@aman-coder03 is there a specific reason why we need to use add_point_labels instead of text3d here? What was your motivation for using switching to it?

Copy link
Copy Markdown
Contributor Author

@aman-coder03 aman-coder03 Mar 28, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@scott-huberty good point...
the reason I switched to add_point_labels directly was to pass all labels at once as a batch and to use shadow=True and show_points=False which aren't available through the text3d wrapper but yeah bypassing text3d to get those params isn't ideal. I think the cleanest fix would be to just extend text3d in _pyvista.py to accept shadow and show_points as parameters and pass them through, that way we keep the version check for always_visible and everything stays within the wrapper.
happy to do that if @drammock and @larsoner are okay with it!

Copy link
Copy Markdown
Member

@drammock drammock Mar 28, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

SGTM

render=False,
reset_camera=False,
)

if src is not None:
atlas_ids, colors = read_freesurfer_lut()
Expand Down
Loading