[Joe Clay]

If I had a better grasp on coding GUI programs, I’d be all over writing a new one from the ground up. Sometimes it feels like they come out with amazing features that I wouldn’t have considered. And sometimes it’s like there’s this glaringly obvious thing that should exist that doesn’t.

For example, I’ve written a script that properly reverses keyframes (maintaining their relationship to the group that is selected, not just flipping the keys in each property and leaving them where they are). It’s got some bugs so I haven’t released it yet, but why isn’t that already done properly in AE? I’d kill for Cinema’s timeline window controls in AE. Want to flip the keys as a group? Scale them negatively. Done. Slide them around. Do whatever you want.

Anyway /rant 🙂

Joe Clay
Workbench.tv

[Joe Clay]

Yeah, third parties have taken off with it, thankfully.

Ah, that makes sense as far as ray-tracing. I never use that so that explains why I’ve never noticed a boost, haha.

Joe Clay
Workbench.tv

[Joe Clay]

If you’re still working on it, it might be worth a shot to try the Smoother. It can smooth out your hand track and it might help. Or it can make it worse, haha. That’s kind of how those things go in AE unfortunately. But it’s always worth a shot.

Joe Clay
Workbench.tv

[Joe Clay]

Indeed. 🙂

I’m kind of surprised the threads don’t close here. A lot of the newest posts are in old threads. It seems to happen frequently. On the plus side, it’s all still relevant information for the most part. Have a great weekend!

Joe Clay
Workbench.tv

[Joe Clay]

Oh nice! If I get some time I might expand upon the one I built already. Worse to worse I might at least be able to get it to the point where you have an array with the data values you need to merge the two scripts.

And yes, I completely agree. Weren’t computers supposed to stop us from having to waste our time on repetitive tasks!? Haha. Have a great weekend!

Joe Clay
Workbench.tv

[Joe Clay]

No problem! I haven’t done it in a while. But from what I remember, once you have your track of the object (or even before you track if you want) you use the surface tool to add a surface. Its icon looks like a grid. It should have four corners, and I think there’s an option to display various things like the mocha logo or a grid so you can see if the track is good.

If I remember correctly, when you export the data (as corner pin data) those corners you set up will be where the corner points go for the corner pin effect. So you just set up whatever you want to replace the screens with in a comp, then add corner pin and paste that data into it.

This older tutorial is listed on their site but the link is broken so I found it on YouTube. Anyway, this should explain it better since I haven’t done it in a bit!
https://www.youtube.com/watch?v=KnHQp-y5yeg

Joe Clay
Workbench.tv

Some contents or functionalities here are not available due to your cookie preferences!

This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.

[Joe Clay]

Use aerender. I don’t render to AME so I’m not sure if this will help but here’s the technique I use:
https://workbench.tv/tutorials/2016-06-03_UsingAerender/

Joe Clay
Workbench.tv

[Joe Clay]

As far as I’ve always heard, AE doesn’t use the GPU too much, it’s mostly reliant upon CPU. They’ve been slowly adding in things that use it, and the next version is supposed to use it more from what I’ve heard. That said, they also tend to mostly support nvidia cards with cuda. My secondary machine is pretty much the other machine you listed (MBP mid-2015 with those same specs) and I haven’t had any issues. It’s not as fast as the hackintosh I built a few years ago, but I wouldn’t expect it to be since that monster is running at 4.3GHz on all cores. I can’t run AE without the discreet card so I can’t offer any more information, but it seems quick enough on this machine. If you have a project that would be easily transferable to me through dropbox or something I can render it and tell you how long it takes if that helps.

Joe Clay
Workbench.tv

[Joe Clay]

Without seeing it I can only guess. I’ve used a transition before that sounds like what you’re talking about. Here’s how I do it.

1. Check the 3D layer box on the two video layers you want to transition between
2. Make a 3D null and push it back in z-space to half of the comp width (960 for full HD comps)
3. Parent the first layer to the null and set a key for Y rotation
4. Move down the timeline and rotate the null 90°
5. Parent the second layer to the null, now you should have a transition between them
6. Make a new comp the same size as your main comp
7. Make a gradient and use mosaic, or use different rectangles (or venetian blinds) to make a matte using only grey tones—this tutorial might help:
https://workbench.tv/tutorials/2016-02-19_QuickDisplace/
The intro to the tutorial uses a similar technique.
8. Put that comp into your main comp and turn off it’s visibility
9. Set up time displacement on a new adjustment layer and set the Time Displacement Layer to your matte
10. Tweak to taste

Joe Clay
Workbench.tv

[Joe Clay]

Haha, that’s awesome! I’m glad you got into it! Sorry I didn’t see this sooner. I’m glad I came across it! I wish there was a way to set the forum to auto subscribe me when I comment without me having to remember to check it!

Anyway let’s see here.

1. I modified someone’s importer. Their code had an issue reading lines for some reason. So since I just needed to display data, I populated it right from the original loop. Theirs used that split() method to split the entire file by the commas, but a lot of CSVs don’t have the ending comma at the line, and it became a hassle trying to add it back it and getting it to work correctly. Anyway, you can store the data into an array. I’d probably make it into an array of objects. This is some pseudo code that should explain it. Unfortunately, PHP is a lot more convenient in a lot of things and I think of that first (especially with arrays!) so I can’t guarantee this works out of the box.

Before the loop set up a variable:
var data = new Array();

And during the file reading loop instead of adding the shape do something like this:
datapoint.x = text[0];
datapoint.y = text[1];
data[] = datapoint;


Anyway, I was thinking about that if I were to expand it further in the future. Also, set up a variable for the max outside the loop, let’s call it maxValue.

var maxValue;

Then in the loop:

if(text[0] > maxValue.x) { maxValue.x = text[0] };
if(text[y] > maxValue.y) { maxValue.y = text[0] };

You might have to initialize maxValue.x and maxValue.y as 0.

Then you’ll have the results when the loop is done. Make a new function to display the values, and call it. It’ll contain something like this:

for (i = 0, i < data.length; i++) {
textLayer = comp.layers.addShape();
textLayer.position.setValue([data[i].x*10,1080-data[i].y]);

And of course you’d need to add in your code that would deal with the max values. I might get to this one day. Haha. But I think it’s awesome how far you’ve come—it’s especially amazing that you hadn’t done any scripting before. AE can be an especially large pain in that regard.

2. That one I’m not sure of. I haven’t done much with text. I did notice that the justification isn’t actually centered for some reason, so I changed them in the comp but it still wasn’t quite right for some reason.

3. I think in your code you could define myComp earlier, probably at the top. Then you can just do:

myComp.width
and
myComp.height

If I recall correctly.

4. Haha, I try to keep mine short usually, but it’s never a bad habit to be descriptive!

Joe Clay
Workbench.tv