0

I am currently working on an importer for captured Shape Key motions. To fasten up the process, I am using the method suggested here. It works perfectly and saves up to a few minutes per import.

I am confused about the way the keyframes are displayed in the viewport. The keyframes show up in the Graph Editor and Dopesheet (Shape Key Editor view) like this:

enter image description here

Here is what I am doing:

if object_with_shape_keys:

keyblocks = object_with_shape_keys.data.shape_keys.key_blocks

for i, key in enumerate(keyblocks):

    if i == 0:
        continue
    # values captured in nano seconds:
    values = self.get_values_for_animation(shape_key_animation_values, frame_count, index = i-1)

    dp = 'key_blocks["{}"].value'.format(key.name)

    fc = self.get_fcurve(mocap_action, data_path = dp) or None

    # When there is no fcurve existing for that datapath,
    # then we can simply create a new one and populate it
    if not fc:

        fc = mocap_action.fcurves.new(dp, index = i)

        # Create new keyframe points and populate captured values
        fc.keyframe_points.add(count = frame_count)
        fc.keyframe_points.foreach_set('co', [x for co in zip(frames, values) for x in co])

The issue I am having with it is that the keyframes do not show up in the data tab in the object properties and I cannot get them to show up. It looks like this:

enter image description here

When I manually add a keyframe to one of the Shape Keys then I the interface looks like this:

enter image description here

What's worse is that after manually adding a keyframe, the previously working motion is overwritten for that specific Shape Key...

To summarize: There is a discrepancy between the keyframe_points stored in the individual fcurves and the keyframes displayed in the viewport. I am wondering if I am missing something or if this is intended behavior. I hope I made myself clear!

Thanks a lot in advance!

f.bra
  • 132
  • 9
  • I can't reproduce, can you give more details? This is what I'm doing: https://pastebin.com/GG3rue3u – scurest Oct 05 '20 at 15:14
  • Hey scurest, thanks a lot for taking a shot at this.. I edited the question to include a snippet. The issue did not arise in your example, so it must be something I am doing in my script. The animation itself is working as it should. Worth noting that the data I am using is captured in nano seconds range which results in many subframes. I tried verifying this by modifying your example to subframes, but it still worked as expected... Let me know if you need more information. – f.bra Oct 05 '20 at 19:40

1 Answers1

1

In

fc = mocap_action.fcurves.new(dp, index = i)

Delete index = i.

That sets fc.array_index which is for things like locations (index=0 targets X, 1 targets Y, etc.). You don't need it for shape keys.

scurest
  • 10,349
  • 13
  • 31