[Alon_a]

Ah. So the linear falloff affects one axis, and the effector’s spline still effects the other axes – got it.

Thank you!

[Alon_a]

This has actually been discussed here many times. One common solution is to use a particle system such as Trapcode Particular. Typically what you do is set the particle emitter to follow the source – say, the phone being whipped around – and the particles are made to stay put, namely be emitted with 0 velocity. You then set the particles to start fairly bright and glowy, decaying over time, and you allow enough of them to get emitted so you get the appearance of a continuous streak.

Of course, if you need something fancier like the new iPod commercials, this may not be workable (you may need to work with actual 3D paths in a 3d app, deforming the streaks as actual geometry or extruding some geometry along the paths). but in many situations the particle trick works really well.

AA

[Alon_a]

Parent a null to each object, and scale the nulls.

AA

[Alon_a]

Parent a null to each object, and scale the nulls.

AA

[Alon_a]

I’ve used XSI for quite a while, and I’m a big fan. I’ve just started using C4D as well, so perhaps I’m in a position to provide some personal perspectives on comparing the two.

XSI is a powerhorse that’s a pleasure to work with. It’s very logically and consistently laid out, and the bang you get for your buck with the FND package is just incredible. Consider you’re getting a top-notch renderer (mental ray), full-fledged particles, superb character animation, soft and hard body dynamics – all that would cost you quite a bit with C4D, and in some cases you won’t be getting the same level of quality.

If you do any amount of modeling, you’ll probably need to work with either subdivision surfaces or NURBS or both. For Sub-D’s, XSI is perhaps the best out there, and from my experience it’s definitely easier to work with than C4D’s HyperNURBS. I don’t work with NURBS that much. Anyway poly modeling in XSI seems to me to be a lot more advanced than with C4D.

C4D has its strong points, no doubt. It’s great for setting stuff up quickly since it has a huge number of useful presets and ready-made components (splines, shaders, rendering components, etc.). Some parts of it are easier to learn, although I must say its reputation as the easiest-of-all-3D-packages may not be true throughout all aspects of the program. To give one small example, the whole “make editable” ritual is not necessary in XSI, which gives you a lot more flexibility in revising your work, and makes a smoother workflow. Personally I find that XSI’s construction history and explorer make much more sense than C4D’s object manager.

For contemporary motion graphics work, the Mograph module in C4D has no match, not by a long shot. I can get most of this stuff done with XSI but it requires a considerable amount of scripting, which may not be up everyone’s alley. Even then, I would get it done much more quickly in C4D, and speed often breeds inspiration (at least for me). Plus, C4D’s integration with AE is still the best, although there are some helpful free plug-ins for XSI in this arena.

Don’t skip checking out XSI. Get the demo (or just shell out the few hundred $$$ it costs to get FND) and spend some time with some good tutorials, there are excellent ones at digital-tutors, 3dtutorial, 3dquakers and elsewhere. You can get help from the community over at XSIbase.com and the XSI thread on CGtalk. Of course C4D has a great community as well (Cgtalk, c4dcafe).

Hope that helps, if you have any specific questions just fire away.

– AA

[Alon_a]

I guess it depends on what you want to do with AE – visual effects, compositing, motion graphics, etc. For motion graphics, my personal favorite is Trish and Chris Meyers’ “Creating Motion Graphics with After Effects”. Vol. 1 is updated for 6.5 while Vol. 2 is still for 5.5 I believe, and they are both absolutely invaluable.

You shouldn’t worry too much about working with 7.0, just make sure you get familiar with the new user interface. Even the keyboard shortcuts are mostly still there.

AA

[Alon_a]

This may not be easy in AE, but one thing I would throw into the mix on top of transfer modes is a Displacement Map. Use the fog layer (or, better, a copy of it) to drive the displacement, so the projected movie will get slightly distorted by the fog – as it should. You may need to tweak the levels of both the fog and the movie, and it may be better to make the visible background fog less contrasty than the fog used for displacement (this is why it’s better to duplicate the fog layer). I think you should be able to get a decent result this way.

Another useful addition might be a volumic light representing the projector beam. Trapcode Lux may be useful here.

By the way, this may not necessarily be easy even in a 3D app, unless you have a dedicated plugin for making 3D fog and are comfortable with texturing, lighting, volumic lights etc.

Hope this helps,

AA

[Alon_a]

You can use Layer->Create Outlines on a text layer to create a mask out of text. However, this will create multiple masks, one or more per character in the text string. Radio Waves needs a single mask, so you’ll want either multiple RW effects or manually connect the different text masks into a single one. You can do this in AE by hand, or in Illustrator, or use some trickery like adding a white solid that slightly overlaps with the bottom of the text string applying Convert Outlines.

Once you have the mask you need, copy it into the layer on which Radio Waves is applied and select it in the RW options.

AA

[Alon_a]

My suggestion would be really to do this in a 3D app… achieving this quality in AE would be next to impossible. For example, there are nice, soft contact shadows in the clip you linked to which were probably created with Ambient Occlusion (or Final Gathering, or a “Dirtmap”…) Recreating such a 3d scene in AE would be very hard work, and will most likely not look quite as good.

If you insist on AE, animating the camera would probably best be done along a path, with separate control over the point-of-interest. There’s an excellent tutorial on that here on the cow (I seem to be pointing people there a lot recently…), search for “Animating the Camera” and “Expressing the Camera” by Rick Gerard.

If you manage to recreate the cube objects and the floor in AE, motion blur and DOF should be manageable with the standard AE MB switch and camera properties. Expect long rendering times!

Good luck…

– AA

[Alon_a]

I’m guessing the “optimise vectorlayer” switch is the one called “Continuous rasterization” in the English version, in which case you’re doing the right thing.

I think the problem is that background fills are not ideally imported as .EPS files. EPS is a vector format and the metal texture is probably (I’m guessing) just a bitmap. So, if for example you scale your logo text you would not want the bitmap to scale as well since that would look bad.

If the texture is repetive, perhaps the vector program in which the EPS was created (Illustrator?) was used to tile a small bitmap, and when you import the tiling is not retained perfectly and you see those thin lines.

Basically what I would do is try to get hold of the metal texture *separately* as an image file, not vector, and if necessary re-tile it in AE. Also get the logo alone in EPS format and then you would get a clean rasterized vector for the text, at any scale, and the texture would remain nice and clean.

I hope I’m on the right track here, just trying to help…

– AA