Here's a standard film DI workflow...
The edit is handed over to the DI facility as data - EDLs, cut lists, whatever. And a reference movie for comparison, on tape or as Quicktime. With these, (the EDLs) the film is scanned. Scanning is a semi-dumb process. No colour correction, simply film to disk 'transfer'. No art.
After scanning, the scanned frames are and 'conformed' to the EDL. Meaning a sequence is made and compared to the original edit. This sequence is graded. meaning colour corrected. A look given, effects, graphics added. All the Art happens here. While grading one looks at the picture projected on a large screen through a projector connected to a Truelight Colour management system. Or some other software or hardware.
After grading, the graded sequence is rendered out. As a large collection of frames.
The render files are recorded back to film on an Arrilaser recorder. or some other.
While recording, the film recorder also prints out test charts which are measured for density and accordingly, printer lights set to make prints at. Normally one uses one constant printer light settings for an entire reel.
Now, if, for some reason, one takes this DI output negative, and works it in a traditional optical analyzer system, grading it all over again, as it were, then the results are unpredictable. It could be great it could be awful. No bets. Should never be done.
After DI it is crucial to see the print carefully and suggest changes if any. And, involve all those, who will be assessing the film qualitatively, at this stage. Any grading decisions that aren't satisfactory, should be regraded, re-recorded and inserted into the negative.
After the negative is struck, there is only so much that can be done in making a print. If you then feel you wish the film to look different, you almost cannot, by just printing it different.
So, it is possible to calibrate any grading system and projector show pictures that translate accurately to film as a DI negative and subsequent print. And, its generally no advisable to grade a DI negative on an analyzer. Some facilities do it, but, at the very least, it goes against the very spirit or reason for DI.
