CMOS Wiggle removal ???
- Thread starter c3hammer
- Start date
havasuphoto
Well-known member
Can you post an example??
waves randomly through a clips sounds kind of like a write error with your card.
You might also try ProDad Mercalli V2 http://www.prodad.com/home/products/videostabilizing/300391667,l-us.xhtml, which works natively or as a plug in to Vegas.
It would also help to know which version of Vegas you are using, and what camera you made the shot with.
It's not a card write issue. It's not a Vegas issue. It's not a software issue of any kind. It's rolling shutter jello.
There are corrections like this:
http://www.thefoundry.co.uk/products/rollingshutter/
But I don't know of any which plug into Vegas. I also can't vouch for how well they work.
There are corrections like this:
http://www.thefoundry.co.uk/products/rollingshutter/
But I don't know of any which plug into Vegas. I also can't vouch for how well they work.
VideoMaker
Well-known member
The stabilizer that comes with Vegas 10 has an option for correcting rolling shutter jello. I've found it to work on some, but not all rolling shutter clips.
Whoa -- didn't even know it was there! I take it back.
havasuphoto
Well-known member
Yep, ProDad Mercalli is built into Vegas 10 Pro, and works amazingly well.
But, since the original poster still has not responded, we are left to guess at what he calls "waves randomly through out the frame".
Like you, I kinda figured it was just CMOS Jello...but, the randomly part kinda through me off.
Here's an example. On the dates April 10th and May 3rd it was particularly bad as you will see.
http://www.youtube.com/watch?v=6QYakJEUd4M
At first I thought it was simply heat waves in the air, but I've found that when the wind is blowing, even ever so slightly it is much worse. This would indicate that even the slightest vibration on the tripod is adding to the issue. A friend called it CMOS wiggle as opposed to jello or rolling shutter and said there was a plugin for FCP to deal with it. The stabilizer built into Vegas 10 does nothing for the issue.
I saw the plug in from TheFoundry.co.uk and was the reason I asked if anyone here knew of a plug in for Vegas.
Cheers,
PEte
http://www.youtube.com/watch?v=6QYakJEUd4M
At first I thought it was simply heat waves in the air, but I've found that when the wind is blowing, even ever so slightly it is much worse. This would indicate that even the slightest vibration on the tripod is adding to the issue. A friend called it CMOS wiggle as opposed to jello or rolling shutter and said there was a plugin for FCP to deal with it. The stabilizer built into Vegas 10 does nothing for the issue.
I saw the plug in from TheFoundry.co.uk and was the reason I asked if anyone here knew of a plug in for Vegas.
Cheers,
PEte
havasuphoto
Well-known member
yea, i see what you mean. how far are you zooming and how long is your lens?? it almost looks like atmospheric distortion. If you have a 500+mm lens on the camera, and the tripod is extremely strong and locked down you will see that effect, even in the winter. I would think that would be impossible to remove. but, that's just my observation.
the only plug-in I know for vegas is mercalli, it's built into vegas pro 10.
the only plug-in I know for vegas is mercalli, it's built into vegas pro 10.
It doesn't look like typical CMOS artifacts . Just curious what camera was used ? Must be very sensitive if wind caused those artifacts
CMOS "Jello" related issues are poorly handled by typical "stabilizers" like coremelt lock n load, mercalli, foundry rolling shutter etc.... They work ok on CMOS linear skew , but usually not very well on "jello" type artifacts, because the motion vectors are too difficult to calculate accurately. As the image is partially scanned the vibration has already caused movement in several different axes. Direction, velocity, rate of decay are almost impossbile to calculate accurately here with different types and magnitudes of vibrations or bumps.
Regardless of what the underlying cause is, since most of your shots are static tripod (ie. background is stationary, or at least is supposed to be if the "waves" weren't there), one approach would be to use a motion adaptive temporal averager. e.g TTempSmooth in avisynth .
You can download this A/B 360p comparison of your "Apr 10" problem shot
http://www.mediafire.com/?qdizg307i5bb4r5
"negatives" include : a learning curve for avisynth, longer processing times, possible temporal averaging artifacts, and loss of image detail with stronger settings and objects in motion (this was the low res 360p example taken from youtube, but if you look at the fur/hide here are spots missing) .
You have to tweak the parameters for best results, and there are tradeoffs between smoother (less wavy) results vs. more loss of detail and generation of artifacts. e.g. there are still small residual "waves" in that example video, and fairly strong settings were already used; You would have to crank up the settings to get rid of the rest, and it will produce more side effects
One way you could improve this is do some rotoscoping (e.g. after effects) and use 2 filtered versions, 1 heavy and 1 light, and replace the object(s) in motion with the light version
I'm trying to look for something similar in concept that you could use in a vegas plugin but can't find anything. The closest thing is Re:Vision FX Denoise using the Frame Average plugin, but I'm not sure if they have a Vegas version , at least it's not on their site (I think they recently released a version of Twixtor for Vegas so maybe they have for Denoise as well)
CMOS "Jello" related issues are poorly handled by typical "stabilizers" like coremelt lock n load, mercalli, foundry rolling shutter etc.... They work ok on CMOS linear skew , but usually not very well on "jello" type artifacts, because the motion vectors are too difficult to calculate accurately. As the image is partially scanned the vibration has already caused movement in several different axes. Direction, velocity, rate of decay are almost impossbile to calculate accurately here with different types and magnitudes of vibrations or bumps.
Regardless of what the underlying cause is, since most of your shots are static tripod (ie. background is stationary, or at least is supposed to be if the "waves" weren't there), one approach would be to use a motion adaptive temporal averager. e.g TTempSmooth in avisynth .
You can download this A/B 360p comparison of your "Apr 10" problem shot
http://www.mediafire.com/?qdizg307i5bb4r5
"negatives" include : a learning curve for avisynth, longer processing times, possible temporal averaging artifacts, and loss of image detail with stronger settings and objects in motion (this was the low res 360p example taken from youtube, but if you look at the fur/hide here are spots missing) .
You have to tweak the parameters for best results, and there are tradeoffs between smoother (less wavy) results vs. more loss of detail and generation of artifacts. e.g. there are still small residual "waves" in that example video, and fairly strong settings were already used; You would have to crank up the settings to get rid of the rest, and it will produce more side effects
One way you could improve this is do some rotoscoping (e.g. after effects) and use 2 filtered versions, 1 heavy and 1 light, and replace the object(s) in motion with the light version
I'm trying to look for something similar in concept that you could use in a vegas plugin but can't find anything. The closest thing is Re:Vision FX Denoise using the Frame Average plugin, but I'm not sure if they have a Vegas version , at least it's not on their site (I think they recently released a version of Twixtor for Vegas so maybe they have for Denoise as well)
UncleBenji1977
Well-known member
Can anyone confirm that the Vegas CMOS removal works?