Darby Edelen
Forum Replies Created
-
Do you have your Z-axis locked? Check the XYZ buttons in the panel at the top of the C4D window, or just press ‘z’ on your keyboard to try and toggle it.
Darby Edelen
-
[Craig Ricker] “2. How do you join a point (vertice) to an edge? Is there a tool that does this?”
If you have a point already that you’d like to join to the edge then I believe you’ll need to create a point where you’d like to join the other vertex to the edge as the only tool I’m aware of that joins points is the Weld tool (shortcut: m-q) that Brian mentioned. That tool requires at least 2 points to weld them together.
There are a few tools for creating points on an edge including the Add Point tool (m-a), the Edge Cut tool (m-f) and the Knife tool (m-k). These all work differently. I prefer to use the Edge Cut tool whenever possible and the Knife tool in most other cases.
For joining edges or faces together there’s the Stich and Sew tool (m-p) and the Bridge tool (m-b). These also work slightly differently. The Stitch and Sew tool generally works better on edges while the Bridge tool generally works better on faces. The Stitch and Sew tool normally moves one edge to another and combines them, but it can also be modified using the Shift key (create a new face between the two edges, like the bridge tool) or the Command/Ctrl key (move both edges toward each other and join them at the mid point).
Darby Edelen
-
Shucks, now I’m curious. Mind sharing the question AND answer with us? 🙂
Darby Edelen
-
The particles are still alive.
Doing some more triage I found that the cloned geometry appears to be collapsing down along the X and Y axes after the first frame. There are no keyframes for this, no particle size variation and no scale keyframes. This only seems to happen with the “Stretch TP” behavior so I’ve switched to “Standard TP” and made some modifications to get that to work.
Everything is okay with the animation now but I’m curious why Stretch TP is behaving badly 🙁
Darby Edelen
-
Since you’ve separated them into CMY components you should start by just playing with the RGB curves. Throw an S-Curve on the Cyan layer and see what you get. The shadows should darken and take on a red tone while the highlights brighten and become more cyan. I would only move to the individual RGB components in the Curves effect if you can’t get the look you want with the master RGB curve.
You could theoretically do any of this with only one copy of your footage and one Curves effect, but splitting it into CMY components makes it easier to achieve the kind of color interactions you should expect to see from a subtractive color process like Technicolor.
Here are a couple of example images. The left side of the image is altered and the right is original.
With an S-Curve on the cyan layer:
With an S-Curve on the magenta layer and the range compressed on the yellow layer:
Darby Edelen
-
[Erik Waluska] “Director’s love to screw things up and then just say “ah, well, we’ll just fix it in post”, but it’s generally more expensive and time-consuming to fix things in post than to re-shoot.”
This depends a lot on how many people need to be recalled to the set to re-shoot. It’s a pain to have to fix something seemingly obvious like this, and the director should definitely get an earful for being oblivious, but in terms of combined man hours 1 person working for 12 hours to fix a shot may well be less costly than 8 people working for 2 hours.
Shots should be better planned from the beginning, but the correct solution after a shot has been flubbed is going to be dependent on a lot of factors and I think it’s shortsighted to always default to “reshoot.”
Besides… these are more billable hours for you 🙂
Darby Edelen
-
If you can achieve an acceptable roto of the subject in front then you might be able to save the shot with some blurring (or median, or possibly minimax) but there are no guarantees here. I’d look at trying to fix the background before doing any roto work and seeing how it looks.
Darby Edelen
-
I’m not going to recommend totally emulating a three strip process, but you can get close without TOO much work.
Start by splitting your footage into Cyan, Magenta and Yellow components. I like to do this with Shift Channels turning Red, Green and Blue to Full Off (for C, M and Y respectively).
If you’re working in 32bpc (which I recommend) then apply a Curves effect to each of the component layers and set their blend modes to Add. Your footage will look blown out at this point but you can easily return to its original range by adding an adjustment layer above the CMY layers, applying a Levels effect and reducing the Output White to 0.5.
If you’re not working in 32bpc then you’ll need to apply this Levels effect to each of the component layers (CMY) individually.
Now experiment with the RGB curves on each of the CMY layers, this is roughly analogous to changing the response curves of each film strip in a 3 color process.
There’s no attempt to emulate a dye transfer process here, but I think the results are pretty good without this step.
On the upper adjustment layer I’d recommend adding a very small blur (less than a pixel radius) as most 1970s films I’ve seen have a slightly soft look. Also the RGB Curves on the CMY layers don’t allow you to adjust saturation very well so I’d recommend either applying a Hue/Saturation effect to the adjustment layer or using a round tripped RGB -> YUV + Curves + YUV -> RGB workflow to adjust saturation.
Darby Edelen
-
[Michele Poggi] ”
* “enable continuos rasterization” – What do you mean by that?”This is a layer switch in the timeline that is specific to vector file formats such as Illustrator files. It allows AE to read the vector data and “continuously rasterize” it, or put another way: your vector file will look as good as it can at any scale.
The downside of this switch is that if you need your layer to be scaled to a very large resolution AE must rasterize to an appropriately large size and you can end up running out of available memory quickly.
More information here: https://help.adobe.com/en_US/aftereffects/cs/using/WS3878526689cb91655866c1103906c6dea-7f49a.html
[Michele Poggi] “The credits are now rolling at 3 pixels per frame, but I’m really interested in the text height you mentioned.. Is there a better way than simple logic to understand what speed it’s best to use? For example, I found vertical blur to be less useful than the horizontal+vertical one. Same goes for using a little grayer text.”
To use an extreme example imagine that your letter ‘i’ is 10 pixels tall. If it moves 4 or 5 pixels in one frame then it may be harder for someone watching to perceive continuous motion. Generally speaking I would recommend shorter text move slower, but this is not a hard and fast rule and instead something I made up on the spot 🙂
Darby Edelen
-
How big is this AI file? You can import AI files directly into AE and enable continuous rasterization on the layer in order to preserve the highest quality at any scale. There may be a reason this won’t work (such as needing the file at such a high resolution that AE chokes on it) but it would generally be my preferred method. If the file is too big then you could separate it into several sections using Illustrator or even Photoshop (the file will be rasterized in Photoshop, but you’ll be able to choose an appropriate scale).
As for the speed of the credit roll I’d say that depends a lot on the text height. Four pixels per frame would probably work okay if the text is at least 20 pixels tall. If your text is shorter then I’d run it slower.
You might also add a small amount of vertical blur to the text.
If you’re going to use an expression to drive it I’d highly recommend that you avoid the timeToFrames() function and instead use thisComp.frameDuration to convert the pixels per frame value like so:
s = -4 / (thisComp.frameDuration);
o = [0, time * s];
value + o;This will avoid stair-stepping in the position values and is probably better practice. It will also help if you decide to use motion blur to generate your small vertical blur.
Darby Edelen
