Given that spatial fields can be FLOAT64, it makes sense to allow precise clamping behavior on the parent transfer function.
Per the 5/17 WG discussion, position arrays are also removed in favor of adding them only when there exists an ANARI device implementation that will directly use the control-point version of TF colors + opacities (currently none do this).
Given that spatial fields can be
FLOAT64
, it makes sense to allow precise clamping behavior on the parent transfer function.Per the 5/17 WG discussion, position arrays are also removed in favor of adding them only when there exists an ANARI device implementation that will directly use the control-point version of TF colors + opacities (currently none do this).