|
|||||||||
|
Thread Tools | Search this Thread |
December 14th, 2008, 03:05 PM | #16 |
Inner Circle
Join Date: Feb 2004
Location: Brooklyn, NY, USA
Posts: 3,841
|
You might need to try using Compressor's Frame Controls Retiming feature which uses Optical Flow technology.
The first step is to check results on the computer. It should play back 25p smoothly on your computer before uploading. |
December 14th, 2008, 05:13 PM | #17 |
Major Player
Join Date: May 2008
Location: Nuremberg, Germany
Posts: 285
|
To convert 50p to 25p there's nothing more to do then to drop every second frame. You don't need advanced techiques like optical flow for that. 50p to 24p conversion for vimeo of course will benefit a lot from optical flow, but it will take a lot of render time.
|
December 14th, 2008, 07:54 PM | #18 |
Trustee
Join Date: Dec 2008
Location: Kennewick, WA
Posts: 1,124
|
Having "choppy video" on Flash video playback may be caused by the speed of your CPU. I know that on my previous MacBookPro (single processor G5) it would have to work really hard just to play a Flash video without dropping frames. One thing to check (if you're on a Mac) is to launch the Activity Monitor and then Window>CPU History. This will show you how hard your CPU is running when viewing a Flash movie. If the graph is showing that your CPU is maxed out, that may be why you're dropping frames.
On the other hand, if you're NOT having trouble viewing other people's Flash videos, then forget everything I just wrote above. :)
__________________
Sony EX3, Canon 5D MkII, Chrosziel Matte Box, Sachtler tripod, Steadicam Flyer, Mac Pro, Apple/Adobe software - 20 years as a local videographer/editor |
December 14th, 2008, 07:58 PM | #19 |
Regular Crew
Join Date: Aug 2008
Location: Melbourne
Posts: 68
|
Hey just a thought but why not upload to your own site and let them download it or play it there? I encoded a few versions of a 5 min production to mp4 and it worked well coming from a EX3 1080I source. I encoded in compressor (Mac) converting to progressive frames (25fps). If you have your own site or ftp isn't this easier?
Evan |
December 15th, 2008, 08:41 AM | #20 |
Inner Circle
Join Date: Dec 2005
Location: Rhode Island
Posts: 4,048
|
Mitchell for me it is not CPU with dual quad core 3.0 and 16GB of ram the machine is not even working when it views the footage. It is a Vimeo problem.
Evan I do upload to a hidden page but letting them download is not an option since most of the footage is right protected from a few TV shows. That is what intrigued me with Vimeo Plus you could have it play on your site only and no download. For me high quality is a must and their footage is not there yet. |
December 15th, 2008, 10:23 AM | #21 | |
Inner Circle
Join Date: Feb 2004
Location: Brooklyn, NY, USA
Posts: 3,841
|
Dominik, Carlos said he used Compressor without any special settings (optical flow) and was not happy with the results. Sometimes theory and practice are very different. If dropping every second frame doesn't look good (after all you are changing the temporal resolution) there's no reason why not trying optical flow. Maybe it's created frames based on the motion estimation will do a better job, maybe not. With technology NEVER assume the "white paper" matches reality.
In fact if dropping every 2nd frame where fine that Carlos' 50p wouldn't look too bad as a Vimeo convert. I suspect Vimeo going from 25p to 24p drops fewer frames than 30p to 24p. Basically Vimeo has a serious frame rate conversion issue and one must do the frame rate conversation yourself if you are to use Vimeo. Not of this is necessary with YouTube. Quote:
|
|
December 15th, 2008, 11:32 AM | #22 |
Regular Crew
Join Date: Mar 2008
Location: Barcelona, Spain
Posts: 33
|
Craig, I do it like you said and the results are far better. No perfect but not bad. I upload a clip to Vimeo and though that 25p>24p don't would be very agressive but it was a disaster. Now I'm uploading to expossure room. In an hour or less I post the link
|
December 15th, 2008, 11:59 AM | #23 |
Inner Circle
Join Date: Feb 2004
Location: Brooklyn, NY, USA
Posts: 3,841
|
Carlos, looking forward to the ExposureRoom link.
Other experiments might be: once you convert 50p to 25p, do a speed change to 24p and use that as your source to upload. Another might be go directly from 50p to 24p using optical flow. Of course you could try uploading your 50p to YouTube and see how that handles it. At the very least YouTube should handle 25p better than Vimeo since it doesn't do any frame conversion at all. |
December 15th, 2008, 12:45 PM | #24 |
Regular Crew
Join Date: Mar 2008
Location: Barcelona, Spain
Posts: 33
|
Because with optical flow the 50>25p conversion was better I'll try that you said, directly to 24p. You now if youtube will allow in the futures view 720 footage without upscale to 1080?
The video in expossure room. It's the better I find at the moment. Let me know if you find it jerky or something. I think is so good vs vimeo Carlos Padilla On ExposureRoom Password: Patr |
December 15th, 2008, 01:40 PM | #25 | |
Major Player
Join Date: May 2008
Location: Nuremberg, Germany
Posts: 285
|
I wouldn't accept that. ;)
Quote:
Btw, the biggest error-source there is: the user. Which upscale? Do you mean full-screen-mode? |
|
December 15th, 2008, 01:54 PM | #26 | |
Regular Crew
Join Date: Mar 2008
Location: Barcelona, Spain
Posts: 33
|
Dominique I had done the 50>25p conversion with and without using Compressor's Frame Controls Retiming and obtained better results with this option without dubt.
Quote:
|
|
December 15th, 2008, 02:02 PM | #27 | |
Major Player
Join Date: May 2008
Location: Nuremberg, Germany
Posts: 285
|
Quote:
1. You did a mistake (did choose frame-blending instead of optical flow or didn't do a 50p->25p-conversion but something else) 2. placebo-effect Can you post examples? I just tried it myself and the results are 100% identical like expected (I subtracted the frames in AFX to be sure there's no difference). |
|
December 15th, 2008, 02:23 PM | #28 |
Regular Crew
Join Date: Mar 2008
Location: Barcelona, Spain
Posts: 33
|
Dominique I delete the video without optical flow because the movements were jerky. The conversion with optical flow is the video you can see above. The difference is not little.
1- I choose optical flow in the second one and not in the first. 2- Placebo effect... ok, I was finding a solution and this seems work to me. If anyone that view the video has the same placebo effect that me, are fine. If you obtains the same results don't do it. Its easy, but believe me, in my case the difference isn't a little interpretation, is fluid movements vs jerky movements. |
December 15th, 2008, 02:44 PM | #29 |
Major Player
Join Date: May 2008
Location: Nuremberg, Germany
Posts: 285
|
I'm pretty sure there's a mistake around here. But without comparing both of your results I won't be able to tell.
|
| ||||||
|
|