GY-HD100U, and Xl-H1
 

Ok, I would like to try and help those who are on the fence on what to buy HDV wise. There for, Come Janunary 9th, I will have both a JVC GY-HD100U, and Canon XL-H1 to play with. I have kind-of lurked here for the last few months, and have decided to buy the bullet to go ahead and help my fellow HDV'ers out. If you guys want to give me a list of set-up's you want me to shoot, and then I will set the camera's up the exact (or as close to) exact same as I can get them.

Peter

Temporal motion is what interests me most about the XL-H1 - since it uses interlaced chips.

So here is a simple test for both cameras in 24P and 24F modes:
Set both to 1/48th shutter and move your hand quickly from one side of the frame to the other.
When you review the footage from the HD100, the distance the hand travels per frame will be the same. This is because the HD100 captures true 24 frames per second.
Now compare to the footage from the XLH1. If Canon derives its 24 frames from 60i fields the same way Sony does, you may see a couple frames of equal distance, and then a frame where the hand suddenly travels farther than it should.

I'd also like to see a resolution chart shot in 24F.

Another way to conduct that test is to use a constant-speed object, say, a car on the street. Point the camera out at the street, at 90 degrees to the street, and tape some cars driving past. When examining the frames in post, on the HD100 you can count the pixels and see that the cars moved exactly (or almost exactly) the same distance each and every frame. On CineFrame 24 on the Sony, you'll see that's not true -- it's more like for the total distance moved in two frames, the car will move about 1/3 in one frame, and 2/3 in the second frame. So over the course of two frames it'll move as far as it did on the progressive-scan camera, but it won't do so smoothly.

It would be most interesting to see how the XLH1 handles that test. If what Canon is saying is true ("it's just like progressive scan"), then the motion should be equal between frames. If what Mike Curtis is saying is true ("the Canon motion is just like CineFrame 24") then the motion won't be equal between frames.