Premiere to effects text graphics

[sauliusd66103322]

So I'm doing a text heavy project right now. Music video with a sort of typography thing. Words jumping on the screen as they are sung in the song. 

Started making it premiere, but eventually figured, premiere was too small for how big I want video eventually turn into so, "to save some effort" I got the timings of the text graphics down onto premiere and exported it to after effects and whoa... so many things happened I barely know where to start. 

1. Every single text clip is now precomped with the mosaic effect(was in premiere export), all precomps are the legth of the project. That's  over a 100 clips who should be on for less than 20 frames each howging memory and power. And they appear correctly in the timeline, the timing seems right, but:

2 I've center-center possitioned all the clips before exporting. Now they are center-top of the screen. This could be my fault as in premiere I've punched in a lot of it while the preview screen was under high scaling of an adjustment layer. They looked in place before exporting though. 

3 Halfway through the video there's a font change from Roboto to Arial Bold. Roboto is fine, however every Arial bold text has this error messege attatched to it with a syntax error untuotched by me and transferred directly from premiere to effects.

There's an adjustment layer encomposing most of the project with a varied zoom going from scale 700 to 100. So.

1. Is there a way to trim those precomps down to where thye are the leght and position of the clips within them? I really do not one to remap every single word again.

2. I'm thinking of grouping all precomps into one big one and adjust possitioning then so all the clips stay where I wanted them initially.

3. Will the error messeges affect my work flow later on?

[Rick Gerard]

Your workflow is dramatically flawed. The expression you are showing is not only poorly written and formatted making it nearly impossible to read but it makes no sense at all;

x = transform.position[0];
y = transform.position[1];
pl = 0; // zero values do nothing so this is not needed
pt = 0; // zero values do nothing so this is not needed
pr = thisComp.width; // not used
pb = thisComp.height;  // not used
pw = thisComp.width;  // duplicate of pr - not needed
ph = thisComp.height;  // duplicate of pb - not needed
pos = [x + pl, y + pt]; // redundant, because pl and pt are zero pos = position would do
wd = pw - 2048 / thisComp.pixelAspect; // does nothing if the comp is square pixels
anchorOffset = 0.5; 
apx = transform.anchorPoint[0] / sourceRectAtTime(time, true).width - anchorOffset;
pos = [pos[0] + (apx + wd), pos[1]];
hd = - 858; 
anchorOffset = ; // duplicate and states anchorOffset = nothing
apy = transform.anchorPoint[1] / sourceRectAtTime(time, true).height + anchorOffset;
pos = [pos[0], pos[1] + (apy + hd)];

the last line says that the variable pos is equal to an array of pos[0] and pos[1] plus apy and hd. The variable hd is defined as -858. The variable apy is defined as the Y value of the anchor point divided by the height of the text  + an anchor point offset. 

pos is defined earlier as just the position of the layer plus 0. Remove all the variables that are never reffered to and clean up the expression and all it does is move the text layer a little more than 800 pixels above the top of the comp. I don't know where you got it but it's garbage.

If you are trying to animate text to music the best thing to do is divide the audio track up into phrases. Each comp should be only as long as a single phrase of music or lyrics. You then animate the text. There are a bunch of simple expressions that automate this process. The process is even easier if you add layer markers to the audio track with the words you want to appear as the marker names. 

I have no idea what your design goal is. If you are just animating word position I would create a MOGRT using the essential graphics panel that automates the move, then just add these graphics in Premiere Pro. If you want to customize the animation of text layers in AE then trying to do the entire 3+ minute song in a single comp is going to be very frustrating and difficult to manage. 

If this is a project created by Dynamic Link using the entire sequence in Premiere Pro, that was your first mistake. Here again, only a single phrase at a time should be used to create an AE comp. The only thing I see rendered in the screenshot is a layer somewhere with what looks like CC Ball Action applied. If you want help with the expression we need to know what it's supposed to do. If you are having problems with a single layer, select it, press 'uu' to show all modified properties, and take a screenshot and tell us what is supposed to be happening to that layer. 

Sorry, there is not any good news. 

[sauliusd66103322]

I did not write the expression. Dynamic link lifted it from the premiere project. 

On previuos projects breaking up the clips was my way of doing things, but that was for individual effects I wanted to get to. This project, I believe, requires a diffirent kind of approach.

The rendered layer is not CC Ball Action. That's layered and shifted collored matts with a tiny Grids applied (Black, Cyan, Magenta, Yellow and White for remaining gaps), simulating close up on computer screen. Project idea is thus: start very macro on flashing words and slowly pull out, showing these words to be within comments on a youtube video. As we pull out even more we realise we are on a computer screen of somebody and then I show that person's character through what they do on the computer. What they watch, what they say and do. 

I was hoping to create with AE a big open space for that screen, since I want to be panning and jumping place to place throughout the video. Overly ambitiuos and big? Very likely. But seems possible within AE limits at least to me.

[sauliusd66103322]

Expression simply should not be there. The those layers are simple text boxes. I do not know how or why that expression box appreared.

[Rick Gerard · Accepted solution]

It's never a good idea to load up a hundred layers in a 3-minute comp and try and edit the whole thing in AE. I've been using AE for about 25 years and I wouldn't consider it. 

From your description, I visualize the project with several cuts or transitions that can easily be handled with an edit. If you want the thing to look like a single 3-minute pull from pixels to the wide shot I would still break it down into pieces. 

I'm assuming you have a shot of your actor sitting at the compute and you have some kind of a camera move in that shot. That shot is one comp. If the camera doesn't pull out then you'll need to live with scaling up the footage as you push into the screen. 

The replacement screen could easily be multiple comps, the first one starting at the pixel level, then more comps as you pull out. It's really easy to match up perspective and position if you are working with 3D layers. Then you just make an invisible cut between the comps where the action matches. I do this kind of stuff all the time. The pullout from pixels to the YouTube channel on a computer screen would be easy to do with 2 or 3 comps and it will take less time than trying to pull the whole thing off in a single shot. When you have the entire screen showing you start a new comp picking up where the last comp left off and just keep that comp the same size. Come to think of it, the text could be created and edited as a comp for each YouTube comment, then nested in the comp with the computer screen. 

To add the screen to the shot of your actor all you have to do is replace the screen and use the comp you made of the full screen as the source. Mocha AE would be a great tool for this.

There is no need to create a huge stage that you can pull back from indefinitely. Trying to do everything in a single comp is just going to over-complicate everything and make it a lot more difficult to do. 

And for that expression, I have no idea where it comes from and even if it was fixed it does nothing but move a text layer off-screen and keep it there. The only way to bring it back would be to animate the Anchor Point or make it a 3D layer and find it with a comp camera.