hello all
I'm projecting pictures with camera projection (uvtexture : Perspective from Camera) on a model and rerendering it from a different point of view.
I'm having two issues atm

1/my model is relatively low def and there's a lot of “warping”.
to resolve this I cranked up the resolution of the mesh with a remesh node before generating the uv .but it gets very heavy.
Back in the days , with softimageXSI , you could define your texture projection as implicit or explicit.
I guess explicit means that the projection was done by the shader at render time.(?)
is there a way in Houdini to have this behavior?

2/the resulting renders with default parameters (micropolygon , 3x3 samples , gaussian -w 2 , ray variance antialiasing)
are not as sharp as the initial pictures.I tried a lot of different combination to try to match initial sharpness but It's somehow confusing to me.
what would be the best approach/settings to get the final render as close as possible to the initial picture (in terms of sharpness/details) …

thanks in advance.