Steve Roberts
Forum Replies Created
-
… and if you have a dual-processor box, you can really speed up your renders by setting up a dual-instance render:
Set up the render to creata a still sequence. Be sure to check “skip existing files”. You may want to close comp windows to minimize clutter. Save the project.
Run this command in the Windows command line:“C:Program Files:Adobe:After Effects 6.5:Support Files:AfterFX.exe” -m
(use backslashes as usual — they don’t show up here)
It works on mac too — if I recall, you need to click on the AE app in the applications folder.AE will launch again in a new window. Open the project you just saved, but in this new window. (just open the file in the usual fashion)
You should now have two AE windows open, each showing the render queue, ready to go. This is the “two instances” of AE.
Press the render button for each instance. The first one you press will render frame 1, and the other one will render frame 2. When the first one is done, it checks to see if frame 2 is done (or busy), skips it, and renders frame 3.AE then renders the sequence in that “leapfrog” pattern until the sequence is done. For fun, watch your processor usage go up. 🙂
Steve
-
Hmm … not through altering the Prefs text file. I haven’t touched that in almost a year.
It’s quite odd — while it’s running ad infinitum, I have no access to the AE interface. It’s unresponsive, yet looping happily.
Steve
-
File>export is generally only used for special cases. For AVIs, Quicktimes and still sequences, use Make Movie, which leads you to the render queue. Though it may be hard to spell, the queue is very powerful. It allows you to fine-tune your render and output settings quite nicely and save those settings. Also, you can queue (!) up a bunch of renders then go away for lunch.
For more info, check the manual, then get Trish and Chris Meyer’s AE books.
Steve
-
After Effects always recompresses every frame of imported footage. The only way to minimize image degradation is to render to a high quality “lossless” codec such as Quicktime’s Animation codec in the intermediate stages, and delay recompressing to the final lossy codec (HDV?) until the end of the entire process.
That’s my 2 cents, never having worked with HDV. Hope it helps.
Steve
-
The effect is commonly called “bullet-time”, and in the first Matrix movie, it involved 180 Canon EOS still cameras if I recall.
Forget software for the moment — it’s a small part of the process. If you’re talking about a real person, and you want to see her from the front, back and side … where are you going to get those images of the model from those angles?
I’ll assume you don’t have a 180-EOS rig. If you have access to a dolly track, you can track around the model to go from back to side to front and get your images. Keep in mind, the model can’t hang in space as you do this unless she’s rigged for it. Once you have the images, and they’re sequentially numbered, you can import them into AE and speed them up, slow them down, whatever.
My point is: if you want to do it old school like Matrix #1, with real people, and you want them to float or freeze, you have to shoot with the multi-still-camera rig, or a special ($$$$$$) film rig that does the same thing. Software can then be used to fill in the spaces between the cameras, turning 180 cameras into 360 or 720 and allowing for slo-mo. AE might be able to do that with a plugin such as Twixtor.
If you want to do it new school, you need 3D animated characters, probably motion captured and a 3D animation app to pull it off. That’s how they did the burly brawls in Matrix #2 & 3. Once you do the 3D animation in the 3D app, you can bring the animation into AE for compositing, but AE wouldn’t really have anything to do with the bullet-time effect.
In short, “no”. 🙂 AE can only do this if the model is a 3D object created in Zaxwerks Invigorator, text created with the shatter effect, or maybe a sphere, cylinder, particles or a stroke created with Trapcode’s 3D Stroke.
For more info, Google “bullet-time”.
Steve
-
The effect is commonly called “bullet-time”, and in the first Matrix movie, it involved 180 Canon EOS still cameras if I recall.
Forget software for the moment — it’s a small part of the process. If you’re talking about a real person, and you want to see her from the front, back and side … where are you going to get those images of the model from those angles?
I’ll assume you don’t have a 180-EOS rig. If you have access to a dolly track, you can track around the model to go from back to side to front and get your images. Keep in mind, the model can’t hang in space as you do this unless she’s rigged for it. Once you have the images, and they’re sequentially numbered, you can import them into AE and speed them up, slow them down, whatever.
My point is: if you want to do it old school like Matrix #1, with real people, and you want them to float or freeze, you have to shoot with the multi-still-camera rig, or a special ($$$$$$) film rig that does the same thing. Software can then be used to fill in the spaces between the cameras, turning 180 cameras into 360 or 720 and allowing for slo-mo. AE might be able to do that with a plugin such as Twixtor.
If you want to do it new school, you need 3D animated characters, probably motion captured and a 3D animation app to pull it off. That’s how they did the burly brawls in Matrix #2 & 3. Once you do the 3D animation in the 3D app, you can bring the animation into AE for compositing, but AE wouldn’t really have anything to do with the bullet-time effect.
In short, “no”. 🙂 AE can only do this if the model is a 3D object created in Zaxwerks Invigorator, text created with the shatter effect, or maybe a sphere, cylinder, particles or a stroke created with Trapcode’s 3D Stroke.
For more info, Google “bullet-time”.
Steve
-
Sounds fine if you choose the AVI route. I go the Quicktime route using FCP, Compressor and DVD Studio Pro, but that’s my preference.
But … never, ever stretch in the output module if you are rendering fields. It will mush up your fields and you’ll get ugly horizontal lines everywhere. If you want to render fields, drag the 540 comp into a DV 480 (preset) comp and scale it 90%, then render the DV 480 comp.
You should note that there are other issues at work here: for instance: 1)if your source comp contains field-rendered video sources filling the frame, I think it’s best to work in a comp that matches the frame size of the video source. To temporarily make it look normal, use Pixel Aspect Ratio Correction. Switch PAR correction off when rendering. 2) if you take 720×486 field-rendered video into a 720×480 comp, move the video up one pixel, so there’s 2 pixels above and 4 below. Don’t use it with 3 px above and 3 below: it will flip the field order. (you can also go 4 above and 2 below)
Get Trish and Chris Meyer’s books to decode the preceding. 🙂
Steve
-
Unless I’m wrong, doesn’t Cinema4D offer cameras with non-perspective projection? That sort of camera can render an object with a “flat” front and “angled” sides.
If that’s not an option, I could picture using an animated mask to reveal two graphics: the “line” and the “shadow”.
Anybody else?
Steve -
Unless I’m wrong, doesn’t Cinema4D offer cameras with non-perspective projection? That sort of camera can render an object with a “flat” front and “angled” sides.
If that’s not an option, I could picture using an animated mask to reveal two graphics: the “line” and the “shadow”.
Anybody else?
Steve -