Does anybody know how VMS does the rolling shutter correction ?
How does it compensate for different cameras ? Here's some examples ... it can even be negative : http://www.guthspot.se/video/deshaker.htm
Thank you musicvid. Some interesting discussion but the only really useful comment was re deinterlace.
How does Vegas or VMS compensate for the different amounts of rolling shutter in different cameras ?
eg Deshaker's default Sony HDR-HC1(E) is 88%
Canon IXUS 220 HS
- at 1280x720 30p: 36%
- at 1920x1080 24p: Somewhere around 75%
iPhone 4: 97%
Canon HV20, at 1440x1080 25p: 38%
I assume it uses motion / vector estimation and lookahead. It does not use fixed target factors for individual camera models as does deshaker.
The rolling shutter correction in Mercalli (including Vegas Pro 11) is on/off, and not configurable. You can download the trial version of VP11 or Mercalli V2 if you want to test it on your own material..
I ran the jello tests on the Kodak Playtouch for Gunnar last year, and rolling shutter compensation in Mercalli is better in side-by-side comparisons.
Thanks musicvid. I've done a test using my HV30. I've videoed some vertical blinds, moving the camera from side to side. The jello effect is very apparent.
VMS with or without the rolling shutter correction is simply terrible. It loses most of the movement completely.
Deshaker has a host of parameters to tinker ... I'm still working on it.
I've tried settings of 38, 60 and 88% but still can't get a good result from deshaker either. 38% is claimed to be the setting for an HV20 and 88% is the default.
Has anyone had any success in correcting rolling shutter with an HV30 ?
Have you tried playing with your shutter speed?
Jello is going to be much easier to correct at say, 1/250 than 1/30, no matter what the technique used.
Have you downloaded and tested the trial version of Mercalli V2?