> If the film 90 percent practical special effects aka SFX, then that would suggest only about 10 percent were DVFX/CGI.
But... is that true? If a real live stuntperson jumps from a real live car to another real live car, but a safety wire is digitally removed and mountains are added in the background, is that a practical effect or a digital one? The article insists it's a binary choice, and seems to say this should not be considered a practical effect at all. And the author's definition of digital effects really is that expansive:
> The film’s color grading process entailed using DVFX to modify certain pre-existing recorded elements within the frame, such as the skies and background environments.
Later in the article a bit of nuance appears:
> What Jackson and the article make evident then is that Fury Road actually had a LOT of CGI quite literally on top of the practical SFX and stunt sequences, which were often enhanced, revised, transformed, augmented, supplanted, etc. through DVFX manipulation. Such changes were, for the most part, microscopic yet their scope was considerably larger than audiences were led to believe.
There are some rhetorical devices that should raise flags, especially the odd "microscopic yet... considerably larger" construction.
Ultimately, when I hear that "80 to 90 percent of the effects were practical", I take it to mean the cars were real and not CGI, the people swinging on poles were real and not CGI, etc. No complaints if the author wants to construct definitions that make those shots no longer "practical", but I don't think many people ever interpreted the claims to mean there was zero color grading or background replacement for continuity.