Forum

The 2nd one looks different to the other 2, if that’s the sort of thing you wanna hear?

Yes, the difference is absolutely massive. I’d suggest googling it for a better explanation, but heres a gist of it: Lets say you have a 240fps video and want it to be output at 30fps.

Frame Blending: It will look at your video and see 240 / 30 = 8 frames input per 1 frame in the final output. So it’ll just put the 8 frames on top of each other which looks fucking terrible when its paused or there’s a still when the number of input frames is less than 8, but when played back as a finished video looks amazingly smooth (see: bassically every decent TF2/CSS/CoD4 frag movie ever made except for Random Damage).

Pixel Motion: It’ll again work out there is 8 frames per final output. So it’ll compare each of those 8 frames and create a perfectly middle frame from that. The final image is slightly smoother, but this is really designed for when you want to convert a low FPS video to a high FPS video, not vice versa. It’ll also take literally about 4x as long to render due to the amount of CPU work involved.

There is of course the 3rd option of using the “CC Force Motion Blur” effect which can work really nice when used situationally, but it causes the crosshair in the screen to blur about when people flick. Again uses tons of CPU and increases render time loads. Bones used this on about 50% of the clips in Random Damage. Also heres a clip I made ages using Pixel Motion and CC Force Motion Blur together to see how it looked: http://www.youtube.com/watch?v=8ULO_uPOzV0 http://www.youtube.com/watch?v=0sXr_-clSA4&NR=1&feature=fvwp

CC Force Motion Blur Method is amazingly fucking smooth imo, looks fantastic, its just a shame about the crosshair thing.

I’m really tempted to do a proper explanation on a blog or something with proper video/still examples and such.

Also, tl;dr Use Frame Blending if you’re recording at 120fps or higher, otherwise don’t bother at all and just record at 30fps in-game.

Thanks Skyride!

If you’re still paying attention, would you reccomend using the motion blur method over the frame blending method? I usually prefer to leave the crosshair out anyway due to preference.

Also, what would you say is the ideal FPS to record at when frame blending?

Really good stuff Skyride I appreciate it.

Last few questions. After effects has this strange 99FPS cap to it. Is there any possible way to exceed the cap or is it not worth it. I ask because I usually use Adobe Media Encoder to compress my videos. I am wondering whether it’s worth frame blending the footage using After Effects or Media Encoder although they’re both CS5 and I’d expect it to make no difference.

Quoted from nTraum

Yes.
If you want to import a 120-fps-footage, import them and set their fps to 60.
Afterwards, double the playback speed of the layer.

Eh? :D

It depends what reason you’d want to do that. Personally the way I used to do it was take the 480fps video, import it into After Effects, then put each clip within its own Composition (Pre-Comps its called), you can do that by just dragging the video file from the project window on to the New Composition button. It’ll automatically put it in a new composition with the same name as the video and the same length, FPS, res, etc.

Then change this composition to 30/60fps (depending on what your final output video will be) and enable frame blending of the video within this layer. Also, if you are planning to put effects/colour grading/etc on the clip, do it here. This means that the final composition you actually edit your video is just tidy and only has the plain clips in it. Sometimes even a good idea to put the into text in its own composition too.

Would actually love to hear what Torden had to say on this tbh, I reckon I’ve got that down to tea pretty much by looking up how to do things in after effects rather than just “how do i get this effect in tf2 frag movie hurr durr”, but he uses it pretty much on a daily basis so is going to have a much better idea.

Quoted from nTraum

I think you got me wrong. :D

Ahhh. Ye I did, thought you meant something else. :P

The simple fix though is to just render it to AVI. TGA sequences shouldn’t be used within AE for a few reasons:

1) It puts more load on your hard disk due to being thousands of files and slows up the process.
2) The overall sequence of TGA files will be about 2-3x the file size it would be of the same sequence encoded to AVI lagarith/huffyuv.
3) its nigh-impossible to get the audio synced correctly in AE when you’re doing time remap.

http://www.youtube.com/watch?v=_cSERmdcZ88

Quickly threw this together to see the difference in recording at a higher FPS.

Now get Torden in here.