Originally Posted by Lee Stewart
As we agreed, the 1920x2205 frame is coming out of the 3D BD player at 24FPS. Then the display steps those frames up to 120Hz with a 5X multiplier. No frame judder there.
Then it is just showing each L & R frame in alternating sequence to each eye at 60Hz syncing up with the active shutter glasses.
OK, I think I now see where you're coming from, so I should explain myself another way.
Strictly speaking there is no telecine judder involved since after all the 3D video has not gone through a telecine conversion process. There we are agreed.
Also as you have indicated there is no judder _of a complete frame_. So in any case we will not be seeing something which looks exactly like the stutter we are used to seeing on 2D 60Hz displays.
The problem I see in showing the 24p 3D on a 120Hz (i.e., 60Hz devoted to each eye) TV is that there will be a quasi-telecine effect in this respect: for each 1920x2205 frame, half of the frame will get more display time than the other half.
So e.g. for every of those 24fps frames, one eye gets 3 views and the other eye gets 2 views.
This is not ideal.
Since each view is 1/120 of a second, for each frame, half the frame is on screen slightly longer than the other half. There's slightly more R eye than L or vice-versa.
I believe that though this is not telecine judder properly so-called, it will produce an artifact which roughly resembles it (is a kind of 3D version of the telecine stutter).
Now I think the observation you made with respect to frame doubling and the rest (in the post from a couple days back which got me going on this subject) -- I think that observation simply reflects the fact that indeed you can get a 24p 3D signal working on a 120Hz (i.e. 60Hz per eye) display. It does not however show that this will work without artifacting.