So the relationship between the sprocket hole and the image never varies?
I'm also a fan of the light table idea. Although the room would have to be dust free.
Telecine without sprocket holes!
Moderator: Andreas Wideroe
-
MovieStuff
- Posts: 6135
- Joined: Wed May 01, 2002 1:07 am
- Real name: Roger Evans
- Location: Kerrville, Texas
- Contact:
Not really. The trajectory of the hub would be mechanically precise and the film would align in tracks to either side.Alex wrote:Two things I'm wondering about.
The bigger this sprocketed hub is, the more likely one may get sideways weave, no?
Actually, it does a bit due to stretch, which is why there would be a bit of breathing in the film and why there is a limit to the size of the drum but it is all workable within certain tolerances. Again, the purpose of the drum isn't to create better registration than you'd get using other methods but, rather, to allow transfer of film that would otherwise be untransferrable due to massive loss of sprocket holes.Alex wrote: The other thing is if for some reason the distance between the sprocket holes on the film isn't consistent, than couldn't the big sprocket gear gizmo turn into a giant sproket hole puncher punching new sprockets between the existing ones?
So the relationship between the sprocket hole and the image never varies?
The light table idea has always been a clever one and variations on that have been used over the years for scrolling titles. Copying film frame by frame would be a fun variation. The problem is that you really can't have the camera advance a predetermined amount over a long strip because of the stretch problem. In other words, 50 feet off one roll is likely to contain more frames than 50 feet off another roll. You would need something that read each sprocket hole and told the camera where to stop. I supposed you could have a variety of sensors ahead and behind the camera to average the placement of missing sprocket holes.Alex wrote: I'm also a fan of the light table idea. Although the room would have to be dust free.
Roger
http://www.moviestuff.tv
-
Actor
- Senior member
- Posts: 1562
- Joined: Mon Nov 25, 2002 2:12 am
- Real name: Sterling Prophet
- Location: Ohio, USA
- Contact:
Interesting that it's been used on something besides the Zapruder film. It's intrigued me ever since I first heard about it.MovieStuff wrote: The light table idea has always been a clever one and variations on that have been used over the years for scrolling titles. Copying film frame by frame would be a fun variation. The problem is that you really can't have the camera advance a predetermined amount over a long strip because of the stretch problem. In other words, 50 feet off one roll is likely to contain more frames than 50 feet off another roll. You would need something that read each sprocket hole and told the camera where to stop. I supposed you could have a variety of sensors ahead and behind the camera to average the placement of missing sprocket holes.
Roger
http://www.moviestuff.tv
The stretch problem is why I suggested that an A.I. program realign the camera on each frame.
I've done some experimenting with stepper motors and computers to control film advance through a gate in lieu of the traditional claw mechanism. The results are mixed at best. The biggest challenge is in getting a smooth stop and repeatable registration of the frames.
The ancient mechanical claw design gracefully accelerates the film to advance and similarly gracefully decellerates the film as the frame comes to rest in the gate. It's always very repeatable with very little vertical jitter, frame to frame. However, simple stepper motor systems have a "step function" start and stop with relatively high torque. As a result, the film is jerked to advance and then the stepper drive instantly stops - then -- surprise! 8O -- the film overshoots tiny amounts on each frame, but not the same amount frame to frame. Hence, the image jitters vertically in front of the telecine camera. An utterly unacceptable situation!
So, stepper motor drives have to be very sophistocated to provide a gentle start and stop of the film. In other words, even though the motor stops with very precise angular location, the film sprocket hole tolerances allow the film to stop in a place different from where the motor driven sprocket stopped.
My conclusion was that even though it is simple to make a stepper motor drive that has precision angular rotation increments, it is still going to be very difficult to get the film to move as well as it does with traditional mechanical claw mechanisms. It certainly can be done with appropriate motor controls, but it's not going to be cheap or easy.
One simple solution might be to allow the images to jitter and then correct the jitter of each frame in software after capture by adjusting the registration of the frames relative to, say, the bottom of the sprocket hole (assuming the system "sees" the sprocket holes). That scheme eliminates the need for a rock solid film handling mechanism. I'm sure someone like Jeff Dodd could successfully take on a project like that!
The more I experimented with stepper drives, the more I came to appreciate the engineering excellence and elegant simplicity of the R8 S8 GAF projector film handling design!
Bob Nichol
http://www.filmtovideo.tv
The ancient mechanical claw design gracefully accelerates the film to advance and similarly gracefully decellerates the film as the frame comes to rest in the gate. It's always very repeatable with very little vertical jitter, frame to frame. However, simple stepper motor systems have a "step function" start and stop with relatively high torque. As a result, the film is jerked to advance and then the stepper drive instantly stops - then -- surprise! 8O -- the film overshoots tiny amounts on each frame, but not the same amount frame to frame. Hence, the image jitters vertically in front of the telecine camera. An utterly unacceptable situation!
So, stepper motor drives have to be very sophistocated to provide a gentle start and stop of the film. In other words, even though the motor stops with very precise angular location, the film sprocket hole tolerances allow the film to stop in a place different from where the motor driven sprocket stopped.
My conclusion was that even though it is simple to make a stepper motor drive that has precision angular rotation increments, it is still going to be very difficult to get the film to move as well as it does with traditional mechanical claw mechanisms. It certainly can be done with appropriate motor controls, but it's not going to be cheap or easy.
One simple solution might be to allow the images to jitter and then correct the jitter of each frame in software after capture by adjusting the registration of the frames relative to, say, the bottom of the sprocket hole (assuming the system "sees" the sprocket holes). That scheme eliminates the need for a rock solid film handling mechanism. I'm sure someone like Jeff Dodd could successfully take on a project like that!
The more I experimented with stepper drives, the more I came to appreciate the engineering excellence and elegant simplicity of the R8 S8 GAF projector film handling design!
Bob Nichol
http://www.filmtovideo.tv
-
ncsproducts
- Posts: 17
- Joined: Tue Jun 08, 2004 5:42 am
- Contact:
Well, you can't feed a stepper x pulses at a constant rate to advance a frame. You have to ramp up at the start then ramp down at the end. Then you'll get silky-smooth motion! A microstepping driver is very helpfull as well.Ping wrote: It's always very repeatable with very little vertical jitter, frame to frame. However, simple stepper motor systems have a "step function" start and stop with relatively high torque. As a result, the film is jerked to advance and then the stepper drive instantly stops
It's tricky software, but it has been done.
"Congrats, it's a fine piece of engineering. I would have never thought you would be able to get both such fine control and fast speeds out of an 8:1 shaft. I love the way the motor ramps up to speed for crystal speeds. Very Impressive."
Dom,
You're absolutely right. As I said, a simple solution will not work, but ramping up and back down is the key to success. I've used microstepping, but again, it's still necessary to ramp the speed and keep positive and negative acceleration of the film under control.
Cheers,
Bob
You're absolutely right. As I said, a simple solution will not work, but ramping up and back down is the key to success. I've used microstepping, but again, it's still necessary to ramp the speed and keep positive and negative acceleration of the film under control.
Cheers,
Bob
It's a very old picture!
-
MovieStuff
- Posts: 6135
- Joined: Wed May 01, 2002 1:07 am
- Real name: Roger Evans
- Location: Kerrville, Texas
- Contact:
Academic to the problem this unit is designed to solve. As noted previously in this thread, the purpose of this device is to run long sections that have no sprocket holes, not just a few damaged sprocket holes. A Rank will not do that. No telecine unit that I know of will do that presently. This unit should be able to run several feet of film that has zero sprocket holes left.DrkAngl wrote:Concerning a telecine that runs film with damaged sprockets: A rank relies on a capstan and already does a great job.
Roger
http://www.moviestuff.tv