Randomize nodes with integers are weird, we always use on scalar values and a floor node to get proper values. Also you can put explorer on sorting to none (creation), check the groups to have the same order, if they don't you can use the reorder tool instead of recreating the groups.
On Fri, Sep 13, 2013 at 6:48 PM, Rob Chapman <tekano....@gmail.com> wrote: > Try Clamp the random value node, sometimes it gives a -1 when frankly it > shouldn't > On 13 Sep 2013 19:02, "adrian wyer" <adrian.w...@fluid-pictures.com> > wrote: > >> ** ** ** ** ** >> >> we have a problem;**** >> >> ** ** >> >> pointcloud A scatters random geometry from group A on a surface**** >> >> ** ** >> >> pointcloud B (which is a duplicate of pointcloud A) scatters random >> geometry from group B on the same surface**** >> >> ** ** >> >> with me so far?**** >> >> ** ** >> >> the purpose of this is to have the same objects (****Arnold**** standins >> of trees) on 2 clouds, Colour ones for RGB, black ones for depth pass**** >> >> ** ** >> >> (you can't override standin materials, scattered with ICE.... go figure)* >> *** >> >> ** ** >> >> problem the two clouds pick DIFFERENT random objects in each case, they >> DON'T MATCH!**** >> >> ** ** >> >> i have a scene to repro if anyone fancies a look..... this is a Soft >> problem (the override thing is ****arnold****'s)**** >> >> ** ** >> >> the workaround is to duplicate cloud A, freeze it, and in a new ICE tree, >> reset the shape using a copy of the original portion of the original ICE >> tree**** >> >> ** ** >> >> confused? we were...**** >> >> ** ** >> >> a**** >> >> ** ** >> >> ** ** >> >> ** ** >> >> Adrian Wyer >> Fluid Pictures >> ************75-77 Margaret St************. >> ************London************ >> W1W 8SY >> ++44(0) 207 580 0829 **** >> >> >> adrian.w...@fluid-pictures.com**** >> >> www.fluid-pictures.com **** >> >> ** ** >> >> Fluid Pictures Limited is registered in ******England****** and ********* >> ***Wales************. >> Company number:5657815 >> VAT number: 872 6893 71**** >> >> ** ** >> >
