Epic 300 FPS

[Damian Grady]

Hi all,

Stage 3.5 secured, I have been looking at some footage on YouTube supposedly shot at 300fps on the EPIC ...

http://www.youtube.com/watch?v=VaLpSpZ41MA

http://www.youtube.com/watch?v=Zg1JPjMvH6s

... and what is puzzling me is that the footage seems to be playing too fast in my opinion. If you playback footage shot on the EPIC @ 300fps at 25fps, that would surely mean that what you see will be 12 times slower than realtime. so something that lasts 1 second in realtime will be 12 seconds long at 300fps played back at 25fps. The clips above do not look like that at all, unless I am missing something?

 

[Eirik Tyrihjel]

The clips above do not look like that at all, unless I am missing something?

Well I think they look exactly right, if it is that important to you, contact those who made the clips and/or try (capturing and) speeding them up x12 and see what it looks like in realtime.

Also keep in mind that even if you shot something at 300 fps, you can actually choose to show it in realtime, 12x or 6x or whatever you choose, it is still shot at 300 fps. Although Ivery much think these shots looks like 300 fps played back - at 12 x

 

[Mitch Gross]

You math is wrong, as tme shift is an exponential scale, not a linear one.

25--50 = 2x
50--100 = 3x
100--200 = 4x
200--300 = 4.5x

 

[Laco Gaal]

contact those who made the clips and/or try (capturing and) speeding them up x12 and see what it looks like in realtime.

you can even download the clips from youtube, and speed them up

 

[Damian Grady]

You math is wrong, as tme shift is an exponential scale, not a linear one.

25--50 = 2x
50--100 = 3x
100--200 = 4x
200--300 = 4.5x

Can you elaborate more on this maths? I'm struggling to understand how physically recording 300 frames worth of footage played back at 25fps doesn't equal 12 seconds of playback but instead is the equivalent of 4.5 seconds.

 

[Mitch Gross]

4.5 seconds has nothing to do with it. You're now talking about something entirely different.

X number of frames is X number of frames. Shoot them at whatever rate and play them back at whatever rate. The original question asked why a clip didn't appear to be as fast as the poster had thought it would be. I explained that relationship.

To determine a running time in playback, your math is correct. Simply divide out the number of frames by the playback framerate to get runtime. But that has nothing to do with the multiplication factor of apparent shift in time. This is the difference between linear and log. The same reasoning explains why one must half or double the light in order to change it by one stop of exposure. That light change may seem huge, but it is all, well, relative.

 

[John Cordell]

Looks about right to me. If you watch the second clip, count how long a water droplet takes to fall about a foot from the top of its arc. In the video it takes about 3 seconds. In real life a one foot vertical drop takes about 1/4 second. So that matches about a 12x speed change.

Mitch, not sure what you're referring to as being log in this, but playback rate sure seems linear to me. Shoot something at 300fps, playback at 25fps, and that thing will be 12 times slower. Shoot a clock and the second hand will move once every 12 seconds on playback. Perhaps you're talking about light levels, where each doubling of the shooting frame rate costs you one additional stop of light.

 

[Eric Haase]

Looks about right to me. If you watch the second clip, count how long a water droplet takes to fall about a foot from the top of its arc. In the video it takes about 3 seconds. In real life a one foot vertical drop takes about 1/4 second. So that matches about a 12x speed change.

Mitch, not sure what you're referring to as being log in this, but playback rate sure seems linear to me. Shoot something at 300fps, playback at 25fps, and that thing will be 12 times slower. Shoot a clock and the second hand will move once every 12 seconds on playback. Perhaps you're talking about light levels, where each doubling of the shooting frame rate costs you one additional stop of light.

Right, an event that lasts for 1 second captured at 300 fps and played back at 25fps will last for 12 seconds on screen, and is thus said to be 12 times slower.

as far the original question about the videos, they look like 300fps to me. 300fps is not really that slow, especially for something that moves so fast like a wakeboard. Whenever I shoot Phantom I am amazed at how high of a frame rate I end up shooting to really slow things down like I want.

 

[Will Keir]

The clips look pretty typical to me. Would like to see more impressive 300fps shots.

Hi all,

Stage 3.5 secured, I have been looking at some footage on YouTube supposedly shot at 300fps on the EPIC ...

http://www.youtube.com/watch?v=VaLpSpZ41MA

http://www.youtube.com/watch?v=Zg1JPjMvH6s

... and what is puzzling me is that the footage seems to be playing too fast in my opinion. If you playback footage shot on the EPIC @ 300fps at 25fps, that would surely mean that what you see will be 12 times slower than realtime. so something that lasts 1 second in realtime will be 12 seconds long at 300fps played back at 25fps. The clips above do not look like that at all, unless I am missing something?

 

[Todd T]

This is what Mitch is describing (Apparent speedup)
SPD 1X.....2X...........3X......................4X.................................5X
fps 25----50--------100----------------200------------------------400

each X is twice as fast as the previous. 300 is halfway between 4X and 5X or 4.5

This is another way of looking at it (calculated speedup)
SPD 1x.....2x.....3x......4x.......5x.......6x.......7x......8x......9x......10x......11x......12x
fps 25----50----75----100----125----150----175----200----225----250----275----300
Each x is is a muiltiplier of the base rate.