The next step in developing your media strategy is to determine how the media you'll be importing will be used. Are you including long excerpts or short? Will your readers need to compare different media files side by side? Are you using annotations to do close reading? The answers to these questions will vary widely depending on the rhetorical tactics of your project, and, like your media sources, each should be tested early and often on your target platforms.
How large are your media files? The cap for uploads to the Scalar server at scalar.usc.edu is 2 MB, but you can of course host files of any size on your own server. If you're working with existing media files from an archive, you'll most likely have to accommodate how the archive has decided to serve their media, but if you're working with files from other sources, then you need to consider your target audience, their likely connection speed, and the specific usage scenarios surrounding each type of file to determine the optimal size. Just because you have a high-resolution image doesn't necessarily mean you should include it at that resolution in your Scalar book; think about the minimum level of detail and depth you need to make your point and resize your media accordingly.
Downloadable vs. Streaming
Media like audio and video which unspool continuously over time are the most demanding for both network connections and computer processors. In general: the shorter and smaller your media clips, and the fewer clips you embed on a single page, the better your book will perform.
Temporal media generally come in two flavors: downloadable and streaming. Downloadable media files work like any other files; in general they must be downloaded completely before becoming usable. This can cause problems if you're using annotations; for example, if you've annotated the last five seconds of a 20 minute video that is delivered by the downloadable method, your readers are going to have to wait for the entire 20 minutes to download before they can watch the five seconds you're interested in.
Streaming media have no such limitations; they are "random access," meaning you can jump to any position in the media and start downloading data from that point. Many media archives like Critical Commons, YouTube, and Vimeo support streaming and will perform well with Scalar annotations no matter where they're placed.
Number of Media Players Per Page
One of the great things about Scalar is the way it enables you to place multiple media files in juxtaposition on a page. There are practical limits, however, to the usefulness of this approach, especially when it comes to video and audio. A page with three or four video players should work fine; one with six or more will exhibit noticeable slowdown as the browser attempts to initialize each player simultaneously. If you have a relatively powerful computer, you may not personally experience these slowdowns, but you should keep in mind the experience of your potential readers with less powerful machines.
Since short chunks of content tend to work best on the web anyway, designing your media references to keep the maximum number of players per page to a reasonable amount is a good approach. Achieving this may be as simple as dividing media-rich sections of your book into multiple pages.
Annotations can be approached in a number of different ways in Scalar. If your goal is to highlight one specific portion of a video or audio clip, it might sense to embed a link to the annotation itself in the text; this will cause the media to cue up to the correct position as the page loads.
If, however, you want to highlight multiple annotations within a single clip, this approach will lead to redundancies, as each annotation link will spawn a new media player, even though all are referring to the same video. Instead, try embedding a link to the entire clip instead: then, as readers play the video, the annotations will each be displayed in turn. Since each annotation is itself a page, another approach would be to place all of the annotations on a path: then, as the user steps through the path, they will be directed to the appropriate portions of the annotated media in the sequence you determine.
For more on creating annotations, visit the Annotating Media section.