Titling Best Practices

Question

Hello
Working with AE and Premier I am starting to wonder what the best way of titling my videos in Premier is.
When I use AE I can get all sorts of fancy results but sometimes the quality of the edges of the fonts and stuff looks jaggy.
My project is video from a consumer camera, a Panasonic PV-GS150 to be exact, and although it is a good camera the video isnt super quality.
After I capture my footage in Premier I get it laid out and then I use AE to create Compositions that I import and drop into the video project in Premier and place it over the footage to create my titles. I have had differing success. Sometimes it is super crisp and sometimes it shows a case of the jaggies.
Could this be caused by interlacing or the type of scan that I set the render of the video too? I have been using progressive scan lately and I think the video quality of some of my latest projects have been good.
Being a perfectionist I would like to know if there is something I am doing wrong. Is there a setting that would cause a little jaggyness, or some process that I could be doing different?
I was thinking about taking the section of video where I want my titles and putting it in the AE project and then putting the AE comp into the timeline in premier in place of where I want the titles to go.
What is the best way?

Answer 1

Have you read the
"Best practices for creating text and vector graphics for video" section of After Effects CS3 Help on the Web?
If you don't feel like clicking, here's the body of that document:
"Text that looks good on your computer screen as you are creating it can sometimes look bad when viewed in a final output movie. These differences can arise from the device used to view the movie or from the compression scheme used to encode the movie. The same is true for other vector graphics, such as shapes in shape layers. Keep the following in mind as you create and animate text and vector graphics for video:
* You should always preview your movie on the same sort of device that your audience will use to view it, such as an NTSC video monitor. (See Preview on an external video monitor.)
* Avoid sharp color transitions, especially from one highly saturated color to its complementary color. Sharp color transitions are difficult for many compression schemessuch as those used by MPEG and JPEG standardsto encode. This can cause visual noise near sharp transitions. For analog television, the same sharp transitions can cause spikes outside the allowed range for the signal, also causing noise.
* When text will be over moving images, make sure that the text has a contrasting border (such as a glow or a stroke) so that the text is still readable when something the same color as the fill passes behind the text.
* Avoid thin horizontal elements, which can vanish from the frame if they happen to be on an even scan line during an odd field, or vice versa. The height of the horizontal bar in a capital H, for example, should be three pixels or greater. You can thicken horizontal elements by increasing font size, using a bold (or faux bold) style, or applying a stroke. (See Formatting characters.)
* When animating text to move verticallyfor scrolling credits, for examplemove the text vertically at a rate in pixels per second that is an even multiple of the field rate for the interlaced video format. This prevents a kind of twitter that can come from the text movement being out of phase with the scan lines. For NTSC, good values include 0, 119.88, and 239.76 pixels per second; for PAL, good values include 0, 100, and 200 pixels per second.
Apply the Autoscroll - Vertical animation preset in the Behaviors category to quickly create a vertical text crawl.
Fortunately, many problems with text in video and compressed movie formats can be solved with one simple technique: Apply Fast Blur to the text layer, with a Blurriness setting between 1 and 2. A slight blur can soften color transitions and cause thin horizontal elements to expand."