That was exactly his evidence: a lens sees things differently from the human eye. The paintings in question have the effects and perspective of a lens. I am familiar with one particular work cited by Mr. Hockney (I have seen the actual work--not just pictures--at the Memphis Brooks Museum). Aesthetically, I don't like it as much as a like many of William Bouguereau's works. While it certainly includes a level of extraneous detail not typically found in classical realism, and the perspective would not be inconsistent with a camera obscura, by my understanding Hockney goes further:
- He seems to suggest that painters mostly relied upon the camera obscura, rather than it being used as an occasional experiment by a few painters.
- Some people who are familiar with the way images are projected can reproduce effects such as foreshortening "accurately" without cameras; I won't quite way "freehand", since the techniques often involve guide lines. Such lines may not appear on a finished work if the design is first rendered as a drawing.
I won't dispute that some 19th-century painters probably used the camera obscura, at least experimentally. I do, however, dispute what I understand to be Hockney's broader claim--that the Classical Realist artists weren't so impressive as they seem, since all they did was trace projections.