Motion Mount on Epic-W

[Anton Shavlik]

I recently recieved my motion mount and was excited to try it out. I noticed a lot of double images in my still frames, almost like a double exposure, which was wrecking my motion blur and prompted me to do some tests.

I shot this lamp pole in my office at 24fps 8kFF.

The operations guide for the motion mount suggests soft shutter operate at 194.4 degree shutter. If you use that, you get a rolling shutter with a sort of split motion blur

It's not until around 105 degree shutter that you get the correct look, motion blur which is softer at the edges like a sine wave

For the square shutter it's a similar story, the recommended 180 is global until about halfway down, where it splits

While once again around 105 degrees you get the correct effect,

.

The soft shutter looks cool while the square and ND are identical and somewhat warm. Also, while the square shutter does work for most of the frame, any grossly over exposed parts of the frame will ghost

Here's normal shutter

And square shutter

and soft shutter at 194.4 degrees, which seems to be the exact amount needed to fade in and out of this ghosting by highlights. If you use soft shutter at a smaller shutter angle you will begin to see the same ghost effect as the square shutter.

Less resolution allows for a larger shutter angle. 4k works well at 144 degree shutter.

I'm wondering if this is replicable for others with the motion mount. Ideally there would be no need to choose between normal, 180 degree motion and global square or soft motion at 8k 24fps.

What it looks like is that the sensor takes more time to draw top to bottom than leaves room for 1/48 second of exposure. If that's true, and say it takes 1/34 of a second to draw the frame, then the only way to get neat global frames would be to get a faster read off the sensor, perhaps 40 or 50% faster...

 

[Stacey Spears]

Anton, thank you for sharing your results. Have been waiting on someone to test the MM with 8K. Would be curious to see if it covers VV.

Not sure if it is the technology or processing speed that is the current limitation based on higher resolution. I think we were at 5K when the MM came out, possibly 6K.

I would love a V2 type of MM, at least for ND, on VV. It is much nicer than having to change a filter on the front of the lens.

 

[Anton Shavlik]

A VV size would be great for the ND, but I think the artifact I'm seeing, the limitation here, is how long it takes to read the sensor from top to bottom.

If, while at a 360 degree shutter in a given frame, the top row of pixels has finished drawing before the bottom row has started then obviously you can't squeeze a global frame out of that since the bottom pixels have nothing in common with the top temporally.

Since ~105 degrees seems to work, there must be a delay from top to bottom of about 255 degrees, at 24 fps that's 1/34 a second. And the artifact we see, the little sliver at the bottom looking like it's misplaced, is exactly what you'd expect in that situation.