Thanks -
I realize in general plane waves are inappropriate but I thought in dealing with a very weak potential that they could be used.
In papers written about the density of states in curved space, they say
\int d^3 x \int d^3 p
is invariant.
I thought this meant that the \sqrt g in...
When I write down a quantum field (for instance to compute T^00 or some expectation value)
I write it as an integral over momentum space.
If I am working in curved space
should this be divided by sqrt [g]?
(and why or why not?)