our usual workflow for still images is quite post-heavy, as we get much better results out of Photoshop than we can out of raw ArchiCAD renders.

We're just starting to investigate 360 panoramic 'VR' images using cylindrical renders and Google Cardboard. Of course, this means that we can't apply all of our usual post tricks (background replacement, cut-out foliage, etc)

I'm just wondering if anyone else has had any thoughts about this obstacle? I guess until Photoshop evolves to the point where it can accurately distort paintwork to match the cylindrical distortion, then we're simply stuck with adjusting levels and simple image adjustments?