[Steve Bentley]

Ok I think I get it now.
What about using a parallel camera and rendering the object itself from above instead of trying for the shadow. Turn shadows off, and use a volumetric shader (like the pyrocluster shader) but instead of going for shaded lighting on a puff, go for the luminence channel only so that the thicker the volume the more brightness you get.
Make sure to turn on off the default light in the options in the render settings so that this light doesn’t affect the look.
Because you mentioned using only the one camera I don’t think you want to see the object AND the “shadow”. You just want what amounts to a volumetric alpha. Right?

If those are metaball objects you may have turn off the metaball effect and let the volumetric shader create the volume cloud based on the “points” or particles that are generated.

If that’s all geometry (and not particles with metaball function on them) or an imported obj mesh sequence you might be able to get the same kind of look with an edge shader but use it in the alpha or transparency or luminence channels – like a fresnel shader.

Another way we’ve done this is to cheat. Use a ring of lights just behind the objects (the camera looks through the ring opening) and use a white material with no specular and no other lights. If the camera has to move you can link the ring of lights to the camera so they are always perpendicular to the view and at an appropriate distance. Make all the lights in the ring about 30% brightness as they will all add up.

I don’t think the transparency in a C4D material builds up with depth of object. So a wave shape viewed from above would be just as clear at the peaks as the troughs (assuming the same index of refraction as your environment) even with a semi transparent material.
And with that thought you could use a material that is spatially wrapped to the object – so the density of the transparency gets “thicker” as the object gets thicker.

Octane or Vray may have something better for this over C4D ‘s default shaders.

I also wondered if a subsurface scattering shader might do the trick with a back light. Thats about a close as you can get with volume density in C4D.

Jim, Brian or Adam will probably have some other brilliant ideas that might help. It’s an interesting problem.
I can play with it later if you want to upload the file.

[Steve Bentley]

A couple of things:
First, use a “straight” alpha when you export. Straight extends the image past the alpha’s edge and then when you composite later that alpha acts like a cookie cutter trimming the image back to where the image glow really ended. A “premultiplied” alpha can leave a black band around glowing things because the alpha ends exactly where the glow ends, so you are cutting the glow off in the RGB image and then cutting it off again in the alpha, and this can leave a little halo of the background color when composited.

Second thing is that you probably used a “screen” function to put the glow over the neon object. But screen doesn’t always render properly when there is no background layer. Screen is a math function but when there are no pixels in a nonexistent background to do the math on you get errors in the image.

One way to go is to render your image as it should be with no alpha but make the glow bigger than it should be. You could also render on the color of the glow instead of black. Then render it again but turn all the color off and crush the levels of some of the elements so you get what amounts to an alpha of the neon with glow – all grayscale but in the RGB channels. Then use that as the track matte when you composite later on.

For both of these techniques make sure to interpret your footage as “Straight” alpha when you import the elements back in to composite. Also make sure if you render out to a movie (like QT) that you choose Millions+ so you get an alpha or RGBA for some other formats.

[Steve Bentley]

Let me see if I understand what you’re after.
In essence you want to cast a shadow from a volumetric object. Where the whispy edges of the shadow wouldn’t have much density but the thicker core of the object would cast a more dense shadow. Is that right?
Do you really want to cast the shadow right behind the object as though the camera was a light? Or do you want a light somewhere else to do the casting?
Will the object actually be a volumetric shape? Or are you just looking for a more realistic shadow that GI tends to give where the ambient light “fogs” the shadow?
Can you describe the scene? There are a number of ways of achieving this but it depends on what your objects are, what the lighting is like and the overall look you are after.

[Steve Bentley]

If you are using the Bend deformer, make sure to try the “Keep Y axis length” button in the attributes panel. The Bend deformer defaults this to off and tends to stretch the objects along the curve, so the text might take on an Extended Bold kind of look that you may not want. Clicking the button will snap the size back to its original length along the curve.

[Steve Bentley]

Sorry, I, like Dave took it as the text resolve you were looking for.
The Emma Watson fade is two elements (or more). The background comes in first with Emma (as a separate layer) on green screen and keyed or rotoscoped, holding as a silhouette before she is brightened. She may have interactive lighting on her that was shot on set or there was some color correcting with masks done to brighten different parts of her before others.
The front elements are another layer – either shot out of focus (this is called “bokeh”) or done with a filter like Lens Care’s Frischluft so you can animate the amount of out of focus – AE’s Lens Blur can sort of do this but no where near as well.
All of these elements are done as 3D layers placed at different depths and the camera is pulled back through them to get some parallax and depth – things farther from the camera recede more slowly than things close to camera. (or they are all flat layers at the same level and scaled in concert to give the multiplane camera effect)

If you are going to color correct down to a silhouette, in other words, a severe crushing of the levels, use a high color depth comp (16 or 32 bit) so there is lots of head room when you start pushing those sliders around, otherswise colors will band and oversaturate. Also try putting a desaturate (HSB effect) before your color correct in the effects stack so that crushed colors don’t bloom and posterize.

[Steve Bentley]

Don’t try and fit in, you’ll end up looking like everyone else – the name of the game is “get noticed”, stand out.

Fox’s searchlights, sky and big hulking art-deco logo is iconic because we associate it with iconic movies; I still expect “a long time ago in a galaxy far far away …” to pop up after I hear that fanfare no matter what movie it’s in front of. Their styling also hearkens back to the golden age of Hollywood, (as does WB, MGM and Columbia), but the current generation has no clue of that era so it’s a wasted reminder. (and, ahem, what century is it now?)
Same goes for the other logo openers engraved on your soul: your brain has glued the brand of the company onto the movie that you love. Are you drawn to the logo because of the movie that follows?

Frankly, those sky ones (you forgot TriStar’s pegasus flying out of the clouds – which I always mix up with Columbia’s, and Lionsgate just sitting in the clouds -before they added the incongruous gears. Disney starts in and ends in clouds, SKG is all clouds, and then theirs Bruckheimer’s desert clouds) but all these blend together, losing brand differentiation because of the clouds, and they are as tired as that tortured lion. I wonder too if clouds are the goto thing when the designer can’t think of anything better. I’m sure it is some Freudian thing they are trying to project on us – to do with heaven and dreaming and seeing pictures in the clouds and the view from Mulholland, but it’s been done to death.
At least Paramount is fun because they tend to blend the logo into the style of the movie, the skimming stars changing to match the movies content. (WB and Universal do this frequently too – remember “The Matrix” opening?).

Then there are the one with common themes used in others – the lighting used in Bruckheimer’s, Touchstone’s and Icon’s. The stars skimming the water in Paramount’s and Tinkerbell doing the same in Disney’s (however, Tink was fired at some point in the recent past), the revolving camera (Disney, Saturn, Skydance, Spyglass, Fox, Universal, WB), and spotlights (Weinstein, Fox, CastleRock, Searchlight), and sunrises and edge glints pretending to be sunrises (Universal, New Line, Hollywood, CastleRock, Legendary, Summit). It’s like once you get a corporate board involved, creativity goes out the 29th story window and they want what the other guy has because, hey, if they are successful, that stolen idea will make us successful too.

Remember too that most big studios are now owned by companies that don’t make movies. They are bankers or soda makers or dot-com’ers. They have corporate boards and like to play it safe. “Safe” usually means boring or as noted above – design it like that other guy’s.

Don’t forget sound design and music. Turn off the music for Universal’s logo – it’s no where near as nice without that bombastically iconic track. Or the audible reminder from THX that you are about to have your ears imploded. (Dolby keeps changing theirs trying to find something as simply perfect)

Keep in mind too that as movies become more and more co-productions to spread the risk (we just finished one that had 6 animated production logos at the front) you may get less and less screen time to make your mark, so it better make a statement fast. Only the big boys get to have their logos linger on screen. Small companies get 3-6 seconds, tops.
JJ Abrams’ Bad Robot couldn’t be farther from the clouds and searchlights of Fox and yet it has instant recognition, even in a still. To me, good motion graphics design is just as iconic when there is no motion.

I’m a fan of all the iterations of Skydance’s, as well as Focus, DMC, and of the clever Marvel comic book riffle – again this one looks good as a still too, and it evolves. DC starts with a good and different idea – the dots of comic book halftone in an eye – but falls flat at the final resolve (or is that just me reflecting on the movie that follows?).
Ridley and Tony’s is also iconic, but often copied – again, making me look twice at the also-ran’s – with me thinking, “Did Scott-Free change their opening?”. Although that style (grease paint on glass) makes any subject matter look similar.

I saw an early version of Anchor Bay’s and it was stunning. I’m not sure how they got to where they are now given the original.

Also, don’t get caught up in plugs-ins. So much of what’s hot in motion graphics comes from some new plug in that is released. The problem is, within a year, everyone is doing that effect and now you look like everyone else again – even if you were first.

Be the one others have to dig deep to copy.

[Steve Bentley]

That’s a great idea Greg. The difficulty might be that you have to put the overlap pins on the original shape before it’s been deformed so it might be tricky to get them in the right spots for the braid overlaps. But in theory you could just set each overlap pin to more and more depth. You can move the overlap pins after the fact and expand them as needed.
I think you would have to start with a shape that looks like a T so that its one big mesh (assuming a tri-braid). I don’t think the overlap tool works with puppet pins on other layers.
I’m wondering if DUIK or one of the other IK tools would help with this (DUIK is free by the way). So you get a natural look to the later position pins affecting the earlier ones as you braid.

[Steve Bentley]

There will be some clean up for sure but that’s par for the course.
If you can make a few different tracking point “designs” so that, as they appear and reappear there is some uniqueness the tracker can lock back on to. The differences can be as simple as color or shape of each dot.
Its also good to put the dots on key face geometry points so if you have to hand correct a few, you can line up the digital tracking marker to a specific facial detail. (a tear drop marker there the apex is on the corner of the eye for instance)
Remember you can also track backwards – sometimes this helps for markers that appear and then reappear.
Many times we will use multiple takes. One with just the face marked up, then one with just the hands with markers on them, then one with everything. You often get data from one take to work with another for a particular body part.

[Steve Bentley]

I don’t think the preview had reflections back in R16, and I don’t remember when the reflection tab changed in the materials editor, but the old material might not jive with the new preview engine.
You can turn reflections off in the preview – its in the Options/configure/EnhancedOpenGl tab.
It may also be the way your gfx card is working with R19 vs the way it worked with R16. You can turn Open Gl off in the C4D preferences to check.

[Steve Bentley]

I think as long as you select only a single file, you can select any file in the sequence, not just the first one, and as long as the picture sequence button is checked, those files will come in as an ordered stream – no further manipulating needed (except perhaps interpreting the time base to match your comp)
There are times that you need to force the order and times when AE will report missing frames if there is a gap in the frames numbers.