I’m looking for a way to expand the range of the input attribute on the lookupchanPOP. In the below example you can see that the twist effect being driven by the output stops right at about the y-unit bound in the local coordinate space (0,1,0).
I assume that the loockupchanPOP is only interpolating the chop input values over a normalized range? Is there a way to expand this or do I just want to scale everything down before sampling it and then size it back up?
Seems like it’s more of a conceptual conundrum: how does one go about defining the range bounds when sampling from a CHOP to a POP Attribute? Using a normalizePOP to properly sanitize your input value ranges seems like it would be a best practice.
Indeed the index in the lookup channel POP is expected to be in the 0-1 range, except if “lookup attr is sample index” is on, then the lookup index is treated as chop sample number.
It could be an RFE to have an option to change the range for the index for convenience.
For this specific example though you can define the behavior outside of the 0-1 range with the two extend parameters, so in your case you would want extend right to be “cycle”
Since the noiseCHOP feeding the lookupchanPOP does not loop cleanly, there would still be a discontinuity with cycle. I suppose something like ‘mirror’ would work.
EDIT: just looked at my own screenshot and I see that it’s a waveCHOP. So yeah, totally.
As with my lookuptex question, this seems like one of those things that just requires the user to be mindful of the ranges of the underlying data they’re working on.
It would be amazing if there were a feedback widget (a bounding box?) or something overlaid in the OP viewer that shows the user how the input data is being mapped. This would only really work with P attribute mappings, but perhaps it’s worth considering drawing some kind of overlay when possible.
