The question can be rephrased
like this: the curve is not exactly on the location as expected in
the original data. Some of segments may be affected by surrounding
data values, for example, the curve may be "attracted" by regions
with high data values.
This is because of the algorithm we use for the curve drawing. This
algorithm is described in the paper:
Peng, H., Ruan, Z., Long, F., Simpson, J.H., and Myers, E.W. (2010)
“V3D enables real-time 3D visualization and quantitative analysis
of large-scale biological image data sets,” Nature Biotechnology,
Vol. 28, No. 4, pp. 348-353, DOI: 10.1038/nbt.1612. ( http://penglab.janelia.org/papersall/doc...
So based on this algorithm, if there is a difference between the
drawn curve and the expected curve in the original data, the reason
is mostly because of the viewing direction that the user uses for
drawing the curve. Changing to another viewing direction may
increase the accuracy of curve locations.