MovieStuff wrote:carllooper wrote:There is digital registration being used in the transfer which is introducing the rocking.
The digital registration is required because the Logmar, as per SMPTE standard, registers the current frame by a sprocket hole offset from such by two frames, whereas the transfer system used doesn't. Normally the transfer system that Friedemann uses, uses the sprocket hole of the current frame. But as can be seen, the sprocket hole against the current frame has some vertical variation, due entirely to variations in the sprocket pitch of the film stock rather than anything to do with pin-registration.
I think the problem isn't the pitch but, rather, the film was perfed with a kind of saw tooth angle on the sprocket holes. The Lasergraphics station uses the sprocket hole edges for lateral registration and, if the sprocket holes are at an angle, then you can get an unsteady image. The pitch, which is the distance between the sprocket holes, I imagine is fine. And the sprocket hole offset isn't as important as is thought, presuming that the film is perfed correctly. Ideally you should use the same sprocket hole to register during transfer as you do when shooting but there have been many tests done long ago that showed it made no visible difference. But, again, this presumes the film is perfed correctly.
Roger
Yes, Friedemann mentioned the same thing. The sprocket is at a slight angle, and with such an error, any horizontal weave, otherwise allowed by the standard, will then affect it's vertical pitch. And it's only really the top/bottom of the sprocket that should be used for reference. The edge of the film itself should be used for horizontal registration (whichever edge is used as the 'hard' edge in the camera).
So yes, ideally a transfer system should use the +2 offset sprocket for registration, either mechanically, opto-electronically or digitally. In this case sprockets were not used at all.
Friedemann mentioned he ended up using trackers on the frame line to do the stabilisation. Looks like he didn't constrain that to XY translation - and so the occasional slight rocking.
So it still remains to be demonstrated whether the camera's pin-registration is actually working. Basically it needs to be partnered with a matching transfer system. So far the pin-registration is not being exploited in any way. That said, pin-registration does also ensure the film is not moving in any way during exposure. So if nothing else it will be improving the sharpness in that way. Freidemann did mention that when running through a projector it was very stable.
Registration clearly matters as the sprocket jiggle demonstrates, but whether that's actually resolved by pin-registration (in both camera and transfer), or must eventually involve digital means, has yet to be explored.
Carl
