rigging with matrices - part05 - soft ik

Описание: In this episode I build a node based setup for reducing the popping effect right before an ik solver reaches its max length. Also known as soft ik or smooth ik.

EDIT:
The script I use for timing the evaluation duration can be optimized to skip the file save, making it a lot faster:
def getAverageEvaluationTime(n=1000, range_size=20):
buffer_size = 100
current_frame = int(mc.currentTime(q=True))
mc.profiler( b=buffer_size )
average_duration = 0
for i in range(n):
frame = current_frame + 1 + i%range_size
mc.profiler( r=True )
mc.profiler( s=True )
mc.currentTime( frame )
mc.profiler( s=False )
event_count = mc.profiler( q=True, eventCount=True )
for event_id in range(event_count):
if mc.profiler( q=True, eventName=True, eventIndex=event_id ) == "EvaluationGraphExecution":
duration = mc.profiler( q=True, eventDuration=True, eventIndex=event_id )
average_duration += duration
break
average_duration /= n
print(f'{average_duration=} micro seconds\nPure fps={1000000/average_duration}')
mc.currentTime(current_frame)

getAverageEvaluationTime()

00:00 explaining soft ik workflow
18:28 construct the upper heigth
30:04 construct the upper target heigth
44:09 construct the upper scale value
46:34 construct the lower scale value
56:52 apply soft ik to upper and lower segments
01:03:08 fixing NaN value error
01:07:24 testing different blend and heigth curves
01:20:36 profiling soft ik performance
01:30:43 explaining soft ik with lower segment scale only

#3danimation #rigging #trigonometry #linearalgebra #matrices #maya