In film restoration, you cannot make overly simplistic, blanket statements about "what is used". What is used is determined by the job at hand and what survives, even with such famous titles as "Gone with the Wind" and so on...
Added to this is the preference and capabilities of the various labs doing the work.
Sorry but wrong. True scanning from this period is exactly how it is done today, only a bit slower; Single RBG image passes per frame (usually with a Alpha channel or Infrared pass for dust) saved in a log-space file such as .cin or .dpx or Open EXR.
Originally Posted by hrst
Telecine work should not be confused with real scanning.
The point was that true scanners (or devices having 2D CCD but with traditional scanner quality) have become much cheaper (and faster as you say). CCDs have better dynamic range and lower noise today and RGB LEDs in conjunction with monochrome CCD's can be used.
Originally Posted by Kino
This way, scanning the original material has become cheaper than making special prints for cheap telecine equipment.
In 1999, scanning was a big deal as you can read from the document; today, it's just a normal operation.
I stand by my post because I do this for a living; I digitally restore motion pictures. Six of films from the 1910's that I restored are playing the Whitney Museum of American Art as you read this.
I built my own 3K scanner so I know a bit about the subject.
You'd do well to read more and further afield and beyond the PR engine that is Millimeter.
how'd you build the scanner? I've only read about them, never seen one in the flesh or in pictures.
Kino; I can't see what is the actual subject you are trying to argue with me and why. I can see no discrepancy nor controversy anywhere. I say just that scanning (not telecining, thank you for the correction) is nowadays easier and cheaper than 10 years ago. Am I wrong if I say that scanning is used in film restoration nowadays more usually than it was 10 years ago?
Anyhow, I've also built my own 16mm "scanner", though it's quite easy and simple design and maybe doesn't deserve to be called a scanner at all. So I know also a bit about the subject, yet probably less than you. Considering the easiness of the build, and time (a few days) and money (nothing) my "scanner" took, it gives quite pleasant results at about 1,5K. My unit is a modified 16mm projector with DC motor drive, enlargened image port, removed audio rollers, adjustable precision RGB led source (to scan color negs) and a normal digital camera (Fuji S5600), using the projector's original lens as a close-up lens. I've opened the digital camera and soldered wires to shutter release button to allow shooting automatically every frame. This unit scans about 1,5 frames per second. I've also written a piece of software on PC to postprocess images acquired. This is the same design as in some units called "workprinters" that are sold at quite high prices. The difference is that they usually use low-res video cameras.
Please tell us something about your scanner! I'm quite interested, too.
hrst, the point I was trying to make, and which you acknowledged, is the difference between telecine/datacine workflows and end products VS tradtional digital scanners.
Scanning has never been "hard", it has been incredibly expensive and remains so in a commercial/post-house environment.
Sounds like you have a nice setup, but you might want to investigate a non-bayer pattern camera without a mechanical shutter.
Stepper motors, an Oxberry Aerial Image head with 16mm and 35mm gates, a 3K (8MP) monochrome, non-bayer pattern camera, RGBI LED lamphouse, some custom software and a multi processor PC with several TB array.
I can capture .dpx or .cin or .exr up to 3K or linear TIFF 16 bit files.
Anyone can make one, given the proper motivation, but using it properly, for whatever discipline, is a whole 'nother ball of wax.
In any event, I should just read the posts and save my energy for some rapidly approaching deadlines.