Larry S. evans ii
Forum Replies Created
-
You can start with Effects> Distort> Mesh Warp. You’ll likely need to isolate sections of your image with a mask and make duplicates of the footage with varying masks to control how the warp affects given individuals.
The puppeteer tool will also do some of this effect, and is a bit more free-form interactive. You just place pins on the “pivot points” you want to manipulate and then move them over time. You can adjust the motion paths for each pin so there’s a great deal of fine tuning possible, and you don’t have to do as much masking.
Digieffects makes Freeform AE which is designed to do exactly this sort of thing. I have not had much chance to work with the demo version, but if it’s in line with their other plug-ins, I’d say it will have some very nice features. I think it’s in the $99-$199 range, but you can try the demo first.
Larry S. Evans II
Executive Producer
Digital I Productions -
There’s not really a good answer for that. After Effects has some high-quality frame blending (assuming version CS3/CS4) for dealing with changes in the timescale, where it will actually try to create new frames based on the “gap” between the frames you have and where you’d want them on the time line.
The idea behind frame blending is that the intervening frames make a smoother visual transition between your existing frame 1 and frame 2, so that as they playback you don’t get “jerky” motion. The concept, put in simplest terms, determines the average value of a pixel by looking at it’s values in frame 1 and frame 2. If you re-time things and get a gap of two frames, then it calculates that mid-point, and moves backward toward each of the original frames, and caluclates an average value between the midpoint and the original.
The more frames you skip between originals, the more that has to be “made up” by the computer, so that long stretch operations can take significant render time and also look rather “phony” when done.
Even if the results are very accurate, what you will get is still a “slow motion” version of the original clip, it will just be smoother.
Now, you are talking about taking a 1 second clip and spreading it to 4 seconds, so all the motion in it will slow down to 1/4 of it’s original speed. If that’s how fast you want it to appear, then that’s fine, and you can turn on the frame blending so After Efects will try to render accurate interframes to fill the gap (in this case, three new frames will need to be created for every existing frame).
Without frame blending, After Effects will simply copy each existing frame three times before moving on to the next one, resulting in a jumpy or jerky playback.
You turn Frame Blending on in the timeline, by checking the box under the little icon that looks like frames of film for each clip you want to use it on. You will also need to click the larger film frame icon in the top of the timeline bar to enable it for all checked layers. Checking the box enables a layer for frame blending. The larger icon turns on frame blending for preview, RAM Playback, etc. so that you can turn it off if you want to work on other parts of a project without having to wait for it to render. Just remember to turn it back on before sending your final work to the render queue.
Now, all that being said, I personally would say that if you have the opportunity to output the original image files at a speed more in line with what you want, that is the best solution. For example, I use Lightwave to create 3D animations that get comped into some video projects. The animations are rendered out of Lightwave to individual Photoshop files with their alpha channel, and then imported as a sequence into After Effects, which is fairly close to what you are doing.
If the timing of the animation is off by a little bit- and I would say never more than 25% or so- it is probably safe and more practical to use time stretching to get the animation to hit it’s mark.
Think of it in these terms. If I need to make the animation 25% longer, then it will run 25% slower. This may be just fine, presuming that all other parts of my animation still match up with any other cues in the video sequence.
If they don’t I may need to do time-remapping, which essentially moves specific points in a clip to a new point on the timeline, and adjust speed of playback to accomodate that. We actually see this used as a visual effect these days, particularly on CSI, where the motion is superfast, then goes to normal, then back to superfast. It’s completely unrealistic, and if you have very large variances in your imported animation when you try to match your background video, then the final shot will look like an episode of CSI (which may not be your intention).
Thus any time-stretching or time-remapping is going to be an “unnatural” effect, and something you want to keep as subtle as possible. To keep it subtle, you have two choices, you can slow the whole thing down a lot and live with the clip being slow motion, or you can have a clip that only needs a bit of slowing down, so that the “unnatural” flow of time isn’t noticed.
For an example of the former, consider that you are adding an animated spaceship flyover to your video clip, and it’s zooming by 4 times too fast. If you slow it down by making it 400% longer than the original, it will move in “slow motion”. However, if it is the only movement in it’s respective clip, and the speed is artistically a match for your video clip, then no one will actually see it as slow motion. That is, because all the audience sees is the ship flying by at it’s new speed, they assume that this is the speed it was originally produced at (we use overcranked cameras for this kind of thing all the time. A plastic model at 120 fps can look like a full sized object moving at “normal” speed when slowed down to 30fps).
In the second case, let’s presume that you are adding the same spaceship, but it also has an animated vapor trail coming out behind it. If you drop the speed by a great deal now, the expansion of the vapor cloud can look like it is moving in slow motion, and not be perceived as “real” by the audience. So in this case, you only want to have to alter the timescale by a small amount, because at some point the shot will look “fake”.
If you are trying to get animation to match preshot video (and that’s frequently what we compositors are doing), I find it is easiest to take my video (even a rough cut) into the 3D program and use it as a backdrop or projection. I think most 3D animation systems support this kind of thing nowadays, even what one would term “entry level”. This allows you to get very close, if not dead on, to the final composite, and then you are only going to have to make small adjustments.
Of course, if you are stuck with a badly timed piece from a client and there is no budget, time, or option to recreate or reshoot, you are stuck with simply making them aware that if you make the video run four times longer, it will run four times slower, and no magic can fix that.
Larry S. Evans II
Executive Producer
Digital I Productions -
All good ideas for making “old film”, but i think what you are looking for is the halftone moire pattern that comes from the low-resolution dot prints used for those old cards.
Personally I would use the CC Ball Action Effect (included in CS3/CS4 not sure about earlier versions) with a grid spacing of something like 1 or 2 and a scatter setting of 0.1, You can adjust the ball size up or down but I wouldn’t go over something like 60 or so. At 60 and below you can get what appears to be a halftone look. Above 60 it starts looking like square pixels (which is one way to get a pixelated effect) as the edges of the balls go outside the gridspace.
Larry S. Evans II
Executive Producer
Digital I Productions -
Time stretching does exactly that. If you make a clip in the comp 200% longer than the original, then it plays back half as fast (slow motion). Likewise, if you make it 50% of the original size it plays twice as fast. The length of the comp is just a “window” of time that the clip plays against. Altering the time on the clip will always produce some kind of temporal distortion.
Larry S. Evans II
Executive Producer
Digital I Productions -
You might also consider creating solid layer of the desired color, then cloning your graphic layer and using it as an alpha matte for the solid which would be “under” the original graphic layer.
In theory if you then scaled up the cloned graphic, it would expose an even amount of the solid layer and produce a stroke-like edge around the entire graphic matching the original’s edges.
This presumes you looking at having the stroke be static, but I think you might also be able to animate it using a mask and the mask shape parameter.
Larry S. Evans II
Executive Producer
Digital I Productions -
No problem. It’s good to have some external validation for the method. Feel free to pass it on to anyone else with a similar issue. -R
Larry S. Evans II
Executive Producer
Digital I Productions -
[Brian Stewart] “For Dynamic Link – you do it from Encore. Encore had a “Adobe Dynamic Link” menu, then you select “Import Premiere Pro Sequence”.”
Not sure if this is a recent update but mine has Adobe Dynamic Link on the File menu just under the Capture and Batch Capture and the top option is “Send to Encore”.
That being said, I have attempted that nifty keep-everything-dynamic-back-to-the-original-source approach to writing a DVD.
Even on a heavy horsepower machine it chugged and chugged, and eventually choked before it wrote a DVD. O
n the other hand, I was able to get the same thing out in reasonable time using the Media Encoder to put everything into mpeg2-dvd and then putting the resultant files into Encore, so I’d echo what everyone else has said. You may find that if you are looking at “Chapters” to your DVD to begin with, rendering out discreet sections and putting them together in Encore can be faster. In any case, pre-render to the final format as much as possible, because doing it inside Encore is slow and un-reliable.
Larry S. Evans II
Executive Producer
Digital I Productions -
Buy Quicktime Pro and install it on another machine that does not have Premiere on it.
You’ll also need to borrow the guy’s Flip and install it’s drivers on the machine, or Quicktime won’t be able to open it, it’s an exceedingly proprietary software.
Open the files from the Flip in Quicktime Pro, and then save them out as .mov files, take that into Premiere and edit to your heart’s content.
Only workflow I know of for this device unless you edit inside the Flip’s own software (which makes Windows Moviemaker look like Premiere). . .
Important note. The reason I recommend installing both QTPro and the Flip software/drivers on a separate machine is that shortly after doing this on my work machine, I lost the option to capture HDV from Premiere. I have absolutely no idea if that caused it, but as my other Premiere workstation continued plugging along fine and the only differences in the software were the QTPro install and the Flip, I am highly suspicious. Removing the Flip didn’t help, ultimately I had to reinstall the entire Adobe stack, which corrected it, but something to beware of.
Larry S. Evans II
Executive Producer
Digital I Productions -
I have the Adobe Master Collection CS4, so in addition to PP, After Effects and Encore, I also have the “hooks” that go into the graphic design, image editing and web production apps as well.
And yes, it is stable on two systems.
I am running an HP xw8200 that is about three years old, dual 3.6 GHz Xeons, 2GB RAM, a SATA array for storage, with an Nvidia Quadro FX 4500 and Windows XP Pro SP3. I have a Vista test bed on a virtual PC on an old computer that I do not use for actual production, but I will not put Vista on any of my production machines.
The other machine I am running it on, should not actually be able to run it, but aside from having to do a reinstall because of a driver conflict with a Flip Camera, and upgrading to a better video card, I’m had no serious issue. This machine is a Sony Vaio with a hyperthread Pentium V (it’s about 7 years old), 1.5 GB RAM, a couple of external 500GB Firewire drives for media and scratch, and a Matrox G450 Multimonitor card. Also Windows XP 32 Pro SP3.
I grant that the Vaio is not nearly as zippy, but that it runs at all with the overhead amazes me. The G450 card is not what I’d recommend for video production, as it’s really more for having a 4 monitor rig in an office or data center. Still it provided an option to actually open Premiere and Encore- really to open the Media Encoder headless app that bridges them. The onboard Intel based video card was having none of it after the second or third upgrade to PP, and so I used one of the extra cards we had around. It provides sufficient rev of video interface, but no real speed.
Based on my experience “stability” here seems to be related to those darn video cards and their drivers, and will often result in problems and crashes you would never ever suspect as having anything to do with the video card. In fact the crash that prompted me to upgrade really shouldn’t have had anything to do with a video card, but it did. Codewarriors getting way too creative.
So before trashing it all. I would suggest you look at a good Nvidia board, and do a full de-install, drive de-fragmenting, and re-install of the whole suite, followed by an immediate application of the service packs. You may find it works.
Larry S. Evans II
Executive Producer
Digital I Productions -
[Clint Milner] “Also, are high def mediums all tapeless? I’ve worked with AVCHD in PP and it’s very handy to work right off our camera”
Actually I shoot with a couple of Canon HV30 hi-end consumer cameras that use standard Mini-DV tape, the HDV tape brand, and with On Location, will shoot direct to drive.
While I prefer the latter option, it’s not always practical for guerilla shots, etc. so I use a lot of tape. First generation off the Mini-DV is comparable to drive based recording. There’s still the time to transfer, but I usually end up just doing a quick review, setting some log points and then letting the system pull the cuts while I go work on something else.
Tapeless is the direction everyone wants to be heading, but I have a couple of news camera guys here that tell me they don’t shoot full-on HD in the field because they’re still having issues with getting datacards to be even 75% reliable on some of the NGC cameras. Tape, for ancient technology, still works 98% of the time.
Larry S. Evans II
Executive Producer
Digital I Productions