I have been testing the 1640 with the TPU installed using "Positive Film".
As I have reported earlier, the backend doesn't seem to send a focus command, and the focus position is reported as 0x40 or 'focus on glass'. I have now tried several scans with targets both on the glass and above. It was actually much harder to distinguish the different cases than I expected. I am not sure what sort of *internal* autofocus the 1640 may have. But I think it is still "on glass" rather than above it. To be really sure I will have to modify the code to send an explicit command to see what difference it makes. I have never really believed the "2.5mm" above the glass: I have just used a micrometer screw gauge to measure the thickness of the film holders and the result was almost precisely 1mm. ael
