Jason Watson
Forum Replies Created
-
There are a number of ways one could go about it- two that come to mind immediately:
1. In Motion make multiple copies of your image, equal to the number of pieces you want assembled. Then mask each copy differently so that altogther they form the original image- pretty much like each is puzzle piece.
2. Do something similar in Photoshop beforehand, making a bunch of different pieces (in as many layers) that can then assemble in Motion later. If you pull in the .psd with all the layers, you won’t have to arrange them again.
I find Motion’s masking tool too unruly and imprecise to work with for something this detailed- Photoshop might be easier if you are comfortable with it.
For both of these methods- as you intuited, working in reverse would be easiest. You don’t necessarily have to reverse the animation, though. Simply go to the frame where you want the assembling to be complete and set a keyframe for position, rotation, etc., for each piece. Then move back in time and modify the position, rotation, etc., as desired. Once you have the start and end points for each piece you can then go in and fine-tune in the middle, which is much easier than trying to eye the whole thing from start to finish.
Hope that helps.
-
I think it’s the image mask that is doing it. When I disabled it the text shadows showed up. I don’t exactly know why, but I’ve noticed that image masks and shadows can be wonky sometimes.
I found two solutions, one of which is much easier.
First, you could clone that text layer, draw a regular mask on it (add blur, animation, etc.) and it will cast shadows.
Secondly, and much easier: check the ‘flatten’ box in the final_logo Group tab in the Inspector.
-
Sorry, I meant moving the text layer as in moving the group it was in. Since it’s being masked by that rectangle, moving just the text layer wouldn’t make a difference, beyond losing the mask. (And wouldn’t get you what you wanted anyway.) The blend mode would also have to be set to normal for both the object and its group, since you have a floor.
Sorry for my lack of clarity.
I’m glad, however, that the opacity trick worked.
Another (probably easier) way would be to move that rectangle into the final_logo group and extend the rectangle all the way to the beginning.
The easiest way, which I played around with a little more to find out, is to use the Add blending to your advantage by having the checkerboard group set to Add, rather than the just object itself.
-
I’m pretty sure the reason it’s doing this is because you have the blend mode set to ‘Add’ for the checkerboard layer. When it fills the whole screen it is more or less accidental, since having the text layer below it screws around with the positioning (at least in how it is viewed) and the blending. (If you move the text layer above it, it will cut off the whole time.)
You can easily trick it into doing what you want, however. Start the text layer at the beginning, but keyframe the opacity to 0. Set another 0 opacity one frame before the mask starts, and then set to 100 on the next. It will probably achieve what you are going for.
-
You mentioned you had some effects on some of layers. If you disable them does it do what you want? It’s possible that the effects aren’t playing nicely with the floor.
-
This doesn’t really reset the anchor point like you want and might be considered cheating, but if you group that object into a new group, that new group then has an anchor point in the center of the object (or objects, depending on if you go that route).
Naturally, you’d have to compensate for that group being your new ‘object,’ but in certain scenarios it can be helpful.
This is also quite useful for making big position changes to an already animated object without adding keyframes and screwing up the animation you already have.
-
If you want to achieve this with keyframes rather than behaviors, it’s not just the spacing of bezier keyframes that matters but the speed of each curve. I usually space out the keyframes as I want, then go into the keyframe editor and drag the handles of the middle ones to create the speeds I want. The curves will definitely end up differently than just changing the interpolation to bezier, as the attached image shows:
-
I’ve always had poor experience with RAM preview in Motion. Sometimes it works, sometimes it doesn’t. I’ve also had the experience of the real-time playback having better performance than the RAM preview.
In some of the earlier versions of Motion you could sometimes trick it into giving you a good RAM preview by slightly adjusting one parameter on one of the layers. Sometimes it would work, at least for me.
You might also try RAM previewing a very short segment and see if that helps at all. In my experience it’s still a mixed bag.
I’d be hard-pressed to say why it does this as well. I suspect it has something to do with how it still plays back in real time within the RAM preview. In AE, for example, as far as I know a RAM preview will only play with audio from the designated in point (and if you play from a different spot you lose the audio preview), whereas in Motion it is in-point agnostic for simultaneous video/audio playback. Part of me wonders if that has any effect on the RAM preview performance.
-
I have a few suggestions, listed from fastest to most tedious.
My first suggestion would be to export the project, import the resulting movie and add the Retiming>Strobe effect to it. Set the Strobe duration for the number of frames you want it to stutter before moving on.
My second suggestion is to add a camera every few frames (determine this in advance!) and very slightly shift its position and/or rotation. If you do this for about 15-20 unique cameras, you can then group them, duplicate and reuse. There will be enough randomness to make it seem random all the way through, especially if the modifications are only very slight. You can then export the project, import the movie and add the Strobe effect at a duration that matches the camera changes.
My third suggestion is far more tedious, but would give you much more realistic results, since it is essentially stop-motion animation.
Start by selecting the number of frames per second you want to use. (It significantly helps if your base timecode is divisible by this number!) For example, if your project has a frame rate of 30, you might want to have the motion be 10fps, which means each object has a new state every 3 frames.
If you want to use an already animated object as a guide, that could be useful. I think the easiest way is to duplicate forward: 1. have the first object be in its position, then duplicate it 2. move the duplicated object forward three frames and adjust position/rotation/whatever slightly 3. duplicate that, move forward, adjust, etc.) Having a smoothly animated reference object would probably help here, since you could just match the states at different points in time.
You could also add the camera adjustment from previous suggestion to really sell it, assuming you are going for a non-steady camera look as well. (that is, simulating the look of the camera shutter adding some jitter, etc.)
There may be easier ways to do this that elude me, but I hope that’s helpful. 🙂
-
There might be other ways to do this, but here’s one:
First off, make sure you are using the Time Date generator as opposed to the Timecode generator.
The animation check box is in the Inspector Window. If you uncheck it you can then set a keyframe with the value you want to start with and then move to the point you want it to end and add another, specifying the end value.
Since Motion’s generator doesn’t seem to let you not display the hours, you’ll probably want to either crop the layer or use a mask. Keep in mind that, depending on the font you use, the time might jump around a bit. Also, it doesn’t really seem to matter at what point in time you place the ending value, Motion will compensate to get it there even if it doesn’t truly represent the time elapsed. In other words, make sure your ending point is in the right spot if the timing is critical.
