Twixtor Test Video (Canon 7d) - 1000fps

[Jonathan12uiz]

loved the video great results...

Very nice test, I very much enjoyed it. Yeah, I could see artifacts, but really, I was expecting a lot worse. Have you ever tried Optical Flow in Apple Motion? Just curious how the two worked compared to each other. I've been using optical flow for going from 24fps to 60fps 1080p on my 7D and it has done a pretty good job. But, that's NOTHING like how far you pushed that. I would easily use that kind of footage in a music video.

I use motions optical flow also, so far i like the results better than twixtor. he is using premiere so I don't know if he is using an apple machine, but if you want to see some slow motion test, I just uploaded some with lots of movement here: http://www.dvxuser.com/V6/showthread.php?231436-slow-motion-testing

 

[dbagatelle]

Couple questions..

Couple questions..

I understand shutter speeds in terms of photographs, but how does that relate to video? does the shutter open and close 1000 times per second.

I'm not sure my video cameras allow me to adjust that. I have the T1i and a Sony CX550V.

What do you mean by nesting the first sequence into the second. And, what does that accomplish.


I have been testing twixtor's sony vegas plug-in and am not getting the results I was hoping for. I'm trying to understand if my footage isn't good enough, or if not using the plug-in correctly.


Dave

--

Shot with: Canon 7d - 720p, 60fps
Edited with: Adobe Premiere Pro CS5 + Twixtor plug-in

I wanted to test out the Twixtor plug-in for Premiere so I shot this test footage of myself doing flips and such in my backyard to try it out. I've read a lot about the quirks of this plug-in from other people on here, and decided to test it in tricky detailed situations as well to see where it would and wouldn't hold up. I know there are things that aren't perfect with the use of twixtor in this video for those of you that are looking for artifacts and such, but I was pleasantly surprised with the leaf footage which was shot assuming that twixtor would slaughter the footage. Some of it was unusable, but I had better luck than I had anticipated. As you've all probably heard before, the less detail the better. (I should have maybe thought about that before wearing that winter coat of crazy colours) haha.

*everything was shot with shutter speeds between 1/1000 & 1/2000 and slowed down to either 3% or 5%*

Through a bit of trial and error, the workflow that I got the best results out of in Premiere consisted of the following:

- import 720p 60fps footage into a 720p 60fps sequence **extend the length of the sequence to what it would be with your footage slowed down**
- make another sequence with the same settings
- nest the first sequence into the second sequence
- add the twixtor plug-in to the nested footage
- slow down to whichever speeds you like, check 'smart blend'
- render it out and see if twixtor has made magic with your footage

I'm going to be doing some more twixtor tests with other footage I have shot over the past week, so I'm sure I'll be learning more. If anybody has any questions, feel free to ask.

Cheers!

 

[dbagatelle]

oh...

oh...

And your footage is great.. exactly the type of stuff I'm looking to produce.

might be that the 7D output is better (or more flexible) than what I can do with the T1i.

--

Shot with: Canon 7d - 720p, 60fps
Edited with: Adobe Premiere Pro CS5 + Twixtor plug-in

I wanted to test out the Twixtor plug-in for Premiere so I shot this test footage of myself doing flips and such in my backyard to try it out. I've read a lot about the quirks of this plug-in from other people on here, and decided to test it in tricky detailed situations as well to see where it would and wouldn't hold up. I know there are things that aren't perfect with the use of twixtor in this video for those of you that are looking for artifacts and such, but I was pleasantly surprised with the leaf footage which was shot assuming that twixtor would slaughter the footage. Some of it was unusable, but I had better luck than I had anticipated. As you've all probably heard before, the less detail the better. (I should have maybe thought about that before wearing that winter coat of crazy colours) haha.

*everything was shot with shutter speeds between 1/1000 & 1/2000 and slowed down to either 3% or 5%*

Through a bit of trial and error, the workflow that I got the best results out of in Premiere consisted of the following:

- import 720p 60fps footage into a 720p 60fps sequence **extend the length of the sequence to what it would be with your footage slowed down**
- make another sequence with the same settings
- nest the first sequence into the second sequence
- add the twixtor plug-in to the nested footage
- slow down to whichever speeds you like, check 'smart blend'
- render it out and see if twixtor has made magic with your footage

I'm going to be doing some more twixtor tests with other footage I have shot over the past week, so I'm sure I'll be learning more. If anybody has any questions, feel free to ask.

Cheers!

 

[Rick Burnett]

You can adjust shutter speed on that camera, it's the length of time that the imager exposes for, before reading it out. For the time before the next frame, it will just ignore that information.

So, if you shoot in 24p that means you are getting 24 frames a second. That means your max time per frame could be 1/24 seconds (except, that is not possible in the camera).

If you look through the settings on exposure setting, most people use 1/50 (which is close to 1/48) which means you are exposing for half the time of the frame. This is what 180 degree is on real film. In the camera, it shows a '50' to denote 1/50

As you shorten the time to expose, which is raise the bottom number, say like 1/120 (which shows as 120 in the camera). The imager is only on for a short amount of time. This eliminates motion blur, but, has a strobing like effect.

The best thing to do is shoot with different settings on moving objects and see what they look like.