Hi guys,

I wanted to pick your brains about overcoming double transforms that happen when you place a dop network within a geometry node.
As a matter of workflow, I have found over the years that trying to contain my setups within a small set of or single geometry node has been beneficial for clarity. That something that Jeff covered also in this webinar. https://vimeo.com/272793390 [vimeo.com]

However, when you do something like that, if you need to place a transform on the geometry node or parent it to a null with a transform, inevitably, it will cause the dop network within to receive a double transform. Once from the geometry node itself and a second time when the data from the geometry node is fed into the dopnetwork via opinputpath().
I've attached a simple smoke scene to illustrate what I mean.

The pyro_doubletransformed node is being translated 8 units to on the x instead of 4 units.

Is there a simple/elegant way to avoid this from happening?

nelsonlim
Is there a simple/elegant way to avoid this from happening?

there is:
simply don't transform objects containing dopnets

there is usually a simple way to compensate for this as a lot of DOPs dealing with objects have Use Object Transform checkbox that you can uncheck, however in this case Volume Source DOP doesn't seem to have it and it's always using object transform, so maybe submit an RFE

however still, you'll avoid a lot of unforeseen headaches if you just don't transform objects containing dopnets

Thanks Tomas for confirming my suspicion. At the moment, sounds like it is probably still best for setups to have dopnets that are separate objects.