[Stuart Elith]

I’m not super experienced with optimising compression myself, but one option may be to add a bit of noise to the black – my understanding is that the compression is introduced to save space/speed, by looking at areas that seem similar and lumping them together. So as it approaches black, it may look at pixels that are very similar and just “call” them all black… so you see chunky blocks where it is compressed together. By adding some noise, you are forcing it to render the pixels as they are, which will increase size but keep more detail.

Of course, then you have noise in your black which may not be what you want either. Not sure how much you’d need to use, too.

[Stuart Elith]

I have looked into the educational licensing a few times, for myself and also to show others in arguments 🙂

I don’t know about now, but a few years ago, there were EDUCATIONAL licenses and ACADEMIC licenses. They were both available to students but with a big difference. The education version was dirt cheap, but could not be used for commercial work and couldn’t be upgraded to a full version, but the academic could. The academic was still a great price (about 30% of the real one i think) and since it could be used commercially and also later upgraded, it is a great option for students. If I were you, this is the one I would go for. You may decide it will be acceptable to use for a while, then upgrade to CS5 or CS6 or whatever later.

There is info on this somewhere on the Adobe website but I remember it being quite difficult and frustrating to find.

Also I think CS5 will be 64bit only, which will be great for those with 64bit systems (and may be worth the upgrade price alone) and useless for those without… in case that also influences you.

[Stuart Elith]

Not sure if this is exactly what you mean, but after you add your comp to the render queue, you can go to the Render Settings and at the top, select Wireframe in the Quality dropdown.

Other than that, I don’t think you can. The wireframe option in the comp viewer window is just about how you view the comp, I don’t think you cn render directly from that (though it looks very similar to the option I have described, I think).

You could try applying a Grid effect to all of your layers… wouldn’t look the same but you would be seeing a bunch of grids in 3D instead of the actual assets…?

[Stuart Elith]

It will depend on the length of video (as in, if it’s really long the amount of work needed will take a long time and perhaps be unrealistic) and also how detailed/clear the undergarments are, but you may be able to achieve a much less obvious result using the methods you have proposed. If the difference in luminance is the main issue, it may even work to make a loose mask around the area (so that the whole image isn’t affected) and simply use levels or other CC effects to bring it closer to the shirt color/luminance. You will need to animate the mask a bit if she moves a lot but it may not be too bad.

I have had a similar situation before where a nipple was showing through a shirt, and in that case I needed to track it quite accurately since the shirt was detailed and unforgiving to simple blurring.

With whatever respect necessary, I think Bogie made a pretty quick jump to some big assumptions from your OP to suggest that you is narrow-minded and restricting her human rights! But I would agree with his suggestion to check with the appropriate people – you may not even need to do it if they are happy with her dress code.

[Stuart Elith]

Michael has summed things up very well, I think.

Although it may seem like a lot of work to shoot the separate clips and layer them, i would MUCH rather do that than shoot the scene all in one and mess around with speeding it up and rotoscoping out your actor later. There are any number of comments I could include here about post work taking longer, being more painful and expensive (in time and money) than some good planning beforehand.

[Stuart Elith]

Hey,
Welcome!

I’m not going to give you step-by-step tutelage but here’s the basic idea that you can follow… read up on specific steps/techniques from other tutorials if you need, and use the adobe Help in AE, it’s really great and should get you going. AE is a big program and you can’t expect to achieve everything straight away.

Because your eye isn’t totally still, you may need to stabilize the footage. Eyes are annoying because you blink, but you will want to track/stabilize based on your pupil, because that is the point which you want to target with the dilation.

To actually achieve the dilate effect, it will depend on what you’ve got in mind but you can probably just use some of the basic distort effects, something like Bulge would be perfect. You essentially want to scale up the pupil area of your eye to make it look dilated.

Good luck!

[Stuart Elith]

OK you’ve got a few different comments/questions going on here. Let me see what I can clear up.

When I referred to the Fill effect, there is an actual effect called “Fill”. It’s in the Generate category of the effects menu, and it just lets you pick a color for the layer.
This is helpful for you, because you can keyframe the effect to change it.

To control the color, you can either change it by hand (activate keyframes for the Color attribute of the fill effect, then choose a new color at the point you want it to change).
Now, I can’t quite tell if this is what you want… you said you don’t mind doing this manually – do you mean you can just go through the timeline and choose the points where you want it to change? If so, that’s easy, because you just do it as you like.

If you want an expression to change it, you can do this but as I said before, AE can’t determine pitch as far as I know, so it wouldn’t know when to change it.

Also note, if you’re playing around with expressions for the color fill, the values need to be between 0 and 1, not 0 and 255 (like RGB). There are also 4 values… R, G, B and alpha.

So, for example, you could enter [0,0,1,100] for pure blue.

On a side note, if you want to use a shape layer instead of a solid, you can simply apply a Fast Blur effect which will give you a similar result to feathering a mask. But either option (solid layer or shape layer) works in most situations 🙂

[Stuart Elith]

If you use the Fill effect to color your circles, you will then be able to play with the color values numerically to change them (using the same data you have already extracted.
Doing it based on pitch is a tricky one though… I don’t know how you would get AE to recognise the different pitches.

[Stuart Elith]

Soundkeys is definitely a nice plugin for things like this. It’s got a visual interface which is nice, and you can control a lot of stuff with the generation of the keyframes.

If you don’t want to buy a plugin, you might be able to achieve it using an expression or two.

Off the top of my head, it’s something like this:

You want to make the circles pop on (100% opacity) when the audio hits a beat, correct? So, the points where the audio is loudest? You can look at your audio layer keyframes after you have applied the Convert Audio to Keyframes command… have a look at the values, take note of what the value is when it hits a beat (when you want the circle to be opaque).

I jsut did it with a random clip, and the highest value was about 5.

Now you need to tie the opacity of a circle to this value. Alt-Click the opacity stopwatch, use the pickwhip to drag to the keyframes you created (i think it’s called Left Slider/Right Slider or something).

If you leave it like this, it will simply use this value as the opacity value. Which isn’t what you want, probably, because it’s a low number (in my case, only 5, which would mean only 5% opacity at the highest points.

So you can then edit your expression by adding a multiplier after the code that the pickwhip created… in my case, if i added *20 afterwards, it would multiply the result by 20 and therefore be 100% opacity at the high point.

That’s the basic way to go about it, but depending on your music, this may not be enough, because there will typically be SOME audio at all points, so the circles will wobble in and out of opacity.

[Stuart Elith]

One other thing you might try is to blur the lower third/caption, to a degree where the text is very illegible, but you might still maintain the natural colors of the video.

If you did it well, it might work as a nice design element… but it could easily look smudged and half-done, in which case covering it up is a better option.