[Steve Bentley]

Ya, that’s a problem with some of the multi camera photogram apps. Because each shot can only see a certain amount of geometry, the polys made from that shot get their own UVW. This is normally necessary because the image from that camera is mapped back onto only that geometry so it makes sense that you would want the UVW’s to match just those polys and just that image. And because there are mutiple images from different angle and they all restart the UVW count at 1, you will get multiple parts of the model that all have a poly called 1 and 2 and 3…. When you are painting C4D just sends all the pixels assigned to poly 1 you are painting on to all the other poly 1’s.

You may have to reorder or sort the UVWs of the whole model to get them back into some coherent order each with a unique ID#.

[Steve Bentley]

Wedges, smedges! Just print everything for 1 second at f8!

[Steve Bentley]

There’s a script on aescripts.com called Zorro the layer tagger which adds all sorts of layer selection and grouping features (including inversions).
(I am not affiliated with aescripts.com or the Zorro script)

[Steve Bentley]

Sorry that took so many posts – an ad popped up on top of the video, covering the tops of the keys, so I didn’t see the little “popcorn” particles until the second viewing. I thought you meant just the falling “notes”.
Glad its working out.

[Steve Bentley]

While what the paint tool draws is pixels, the path it follow while it paints is a vector based curve. You can’t access this path directly like you can a mask, but you can copy the paths from each paint stroke and paste it into a layer with a mask or a shape (pasting on top of a dummy mask is required sometimes) and then scale that layer. The simplest solution is then to stroke those enlarges paths and they will be clean. However, if you have some “english” on the strokes you painted originally that’s not going to come through.

What I’m wondering is if you could link the path from your original paint on the small layer to a paint effect on a large layer, and then scale (and link) the path with expressions. If you head over to the expression forum Dan Ebberts and boys might have an answer for you, to save you doing it all again.

[Steve Bentley]

Oh I see what you are talking about – the little particles at the top of the keys. Ok, that’s pretty easy. Make a mask that essentially makes a line just at the top of the keys. Then pass the footage of your descending “notes” through that mask. Precomp that.
Use that precomp layer for Particular to drive the emission of the particles. The particles will only emit when a note passes through that region and only while the note is passing, and in the area where the note crosses that line mask.
If you want you can also have particular take the color of the note that’s passing and use that as the color for the particles.
From the looks of the particles, you will probably need to turn on auxiliary particles so that the first batch of larger particles emit secondary particles. You could also use two or three copies of your layer with Particular on it and have a different style of particles for each layer

Comp all the particle layers on top of the original footage.

[Steve Bentley]

On a side note: If the image is (as is suggested by the provided pictures) from a book, there will be a dot pattern used in the printing. As soon as you animate anything in the picture it immediately looks fake due to the dot pattern not moving like film grain moves. It screams “postcard”. (watch a few bad B movies and you will see this happen frequently)

Depending on the resolution, you can sometimes use clever software to remove the dot pattern but its tricky (do they still make Ofoto?) – the dots are where the detail is; in fact, in printing, get rid of the dots and there isn’t much left at all.

Using a grain killer often makes the image soft (and the reference is already soft enough due to age and the de-rez from the printing) plus these filters were designed for removing film grain which has a very different structure to a print half-tone or color sep.

[Steve Bentley]

Well the ground plane probably doesn’t have any geometry on it so you would just be projecting on a flat plane – no distortion.
A quick cheat that can sometimes work is to use the Displacement Map effect. Use the ground plane image as the source for the displacement and that will “shift” the pixels of the flipped clouds so that it gives the illusion that they are moving out of the way along stuff in the scene. You may have to precomp your flipped clouds so the displacement operates on the flipped image after its been flipped. There is usually some masking involved too for stuff that shifts off the top of a building, say.

Add to that, or instead of that, you can apply a corner pin (I prefer CCPin as it’s sharper) to warp the flipped clouds wider out near camera (for perspective). Again you may need to precomp to have the CCpin operate on the already flipped cloud image.

If you really want to go to town, and you have C4D or C4Dlite you can do what’s called camera mapping (or projection mapping). You build simple shapes in 3D that match your picture and then you can literally project your clouds on what is now 3D geometry. The illusion is very convincing. You can also do this with Element 3D.

Can you post pic? We might have some other ideas based on what’s in the scene.

[Steve Bentley]

And just when I was finally getting over my slit scan nightmares from the optical days. Thanks Andrew!!! I can hear the stepper motors again! Ahhhhhhhhhh!

[Steve Bentley]

That’s funny. I was thinking “Jacobs ladder” as I was reading the post and then saw the image.
Unfortunately there is no Force motion blur like there is in Flame. But there are other ways.

Compound blur takes a grayscale image and blurs the underlying image by the amount specified by the grayscale map. This way the entire image need not blur just parts of it and by different amounts.

Echo will take past frames and blend them together – combine that with a density mask and you have have different parts of the image that have more echos than others.

Directional Blur (or Motion blur depending on the version of AE you have) can smear the image on specific angles – again, combined with soft edge masks you can get larger smears in only parts of the image.

You can use Particular to emit particles from different parts of the image and the image can be applied back to those particles – get enough of the right type of particles and it can produce some really cool smeary effects.

If you are animating things you can increase the default motion blur to crazy levels (Comp settings, advanced tab)

You can generate a motion blur data element and then throw that at something like RealSmart Motion blur and it will smear the image based on the element (its a color element with R,G,B denoting different directions of blur – in 3D no less!)