Right margin issue with Final Draft scripts

I’ve gotten to a point where I can combine novel and scripts into one epub, but I’m running into a right margin issue.

I’ve searched the forum and did a google search, but tried all suggestions and nothing works.

I already tried loading the script settings from a Final Draft script (I’m using Final Draft 8) and that doesn’t work.

In script settings, I’ve manually changed the indents, but the only one that won’t budge is the right margin for dialogue. It is set to 6 and no matter what I change it to, it stays there.

In the Scrivener program, the formatting looks how it should. Also, if I export to PDF, it looks fine. But when I export to an epub, it will run dialogue to the end of the screen like it is an action line. I attached a screenshot. A perfect example is the dialogue at the top. The dialogue here should end in alignment with the period after the word “agreement” in the action line below it.

Since it looks fine in Scrivener, I can’t make adjustments, so not sure what to do. Is it a Final Draft issue? Would I have better luck using another format or input method such as Movie Magic or RTF?


FDX, RTF and all of that has little bearing upon the ePub output, which is based upon HTML. You can view the formatting for each type of element in the Layout compile options pane, by clicking on the [b]Customize CSS...[/b] button. Here you can compare the Action and Dialogue CSS classes:

[code]p.action {text-align:left; text-indent:0;
margin: 12pt 0 0 0; font-family: monospace;}

p.dialogue {text-align:left; text-indent:0;
margin: 0 20% 0 20%; font-family: monospace;}[/code]

Note how dialogue is spaced as a percentile, not a fixed value, for the left and right values[size=80][1][/size]. This is because script formatting in an e-book environment can never be precisely accurate down to the millimetre, and in most cases you do not want to used fixed values for indentation. Remember that a cumulative indent of 3” on an iPhone screen would consume more space than the width of the display! So these approximate values are used to provide the reader with a familiar looking layout as best as can be done one something that isn’t actually 8.5 x 11”. Unfortunately it appears the iPad rendering engine has chosen to ignore a right-margin value. If you verify the .epub file with another program like Adobe Digital Editions or Calibre, you will note that the output looks more like a standard script. You could experiment with larger values in the custom CSS, but be aware that will probably produce undesired effects on anything other than an iPad.


  1. Margins are declared as TOP RIGHT BOTTOM LEFT, so “0 20% 0 20% o” means 0 margins above and below and 20% of the display width for left and right.

Is the iPad rendering engine choosing to ignore the right margin value a bug specific to epubs exported in Scrivener?

Before I discovered Scrivener, I was able to export an epub for a screenplay from another program and the margins were spot on when reading on the ipad. I have also downloaded epub scripts from the ibookstore which have displayed fine as well.

I wouldn’t say it is specific to Scrivener. ePub files are just HTML and CSS, much like a web site. What you put into the CSS is going to be how the book displays, and it doesn’t matter if you started with Scrivener, Word or Sigil to get to that spot. The point as I noted is that Scrivener uses percentage based spacing to automatically adapt the layout to different display sizes and the iPad seems to be ignoring percentage based margin settings on the right edge for whatever reason. That would be something worth reporting as a bug to Apple, though they probably have their reasons for ignoring it.

As suggested, you can control the custom CSS in the Layout compile option pane, and experiment with other metric models. It’s important to test on as many devices as you possibly can, when doing so, since changes that look good on one particular platform+software may break down in another combination (much as in this case, the 20% system appears to break down on the iPad+iBook combination). To be clear I haven’t actually tested this myself, but if I were to do so, I would start by adjusting the percentage number in incremental values to see if there is any response from iBooks, and if that doesn’t work I would start experimenting with other measurement models.