Background: I am trying to add civil unrest to Skyrim. I managed to get a sort of demonstration going by having protesters use an Idle marker, and creating various protest chants as (Idle) topic types (with being in a protest faction as a condition). But I have several problems.

First, when I go to record a protest slogan, I see as voice types for that slogan only the voice types of the unique actors I have already assigned to that faction. But I want the faction to grow over time and have new voices added. If I remove the GetInFaction condition, I see all voice types for the slogan, and can see they all have different folders on my hard drive. I thought that by selecting one in the list and recording, I would be able to make a different recording of the slogan for each voice type - but when I record and save, the same recording is saved for all the voice types! How can I get lots of protesters to say 'Come see the violence inherent in the system!' in their own voice types? (I realize I could create a different protest topic, with the same text, for each voicetype, but I hope there is a better way.)

Secondly, since it relies on Idles, the protests are all a bit laid back. I don't want the complete scriptedness of using repeating scenes for every demo (a hundred different Heimskrs might be just a little bit annoying), but I'd like to be able to put some occasional structure into what the protesters say - for example to make them chant one particular random slogan a few times, with urgency, before moving on to another. AI packages have this helpful looking procedure called Say - but it doesn't seem to do what I want. Does it?

Finally, it's not so much of a problem, but I wondered what some things in the generic dialogue were. For example, what does [E] mean, for example in DialogueGeneric\DialogueGenericHello(Hello): [E] Hmph.? And what does <> mean, in DialogueGeneric\DialogueGenericStandOnFurniture <>What are you doing that for?

Sorry if I have missed previous answers to these questions - I have puzzled myself into a stupor

OK, I can answer my own question about voice types, at least. It is only possible to use the Creation Kit's Record button for one voice type, but you can then copy the resulting wav file into any other voice type's folder and edit it there with a wave editor, and it will be used for NPCs with that voice type. Ugly, but at least it works, and means I don't have to assign someone of every voice type to my faction straight away. (Incidentally, the Creation Kit's preview button seems to play only the last thing you recorded, not the response you're actually looking at - making it pretty useless, as far as I can see.)

But I still haven't got my protesters to chant

E means the dialogue info is marked as a random End, or the end of a stack of random info's.

Recording in the CK is there for the most primitive of purposes. Basically just so the devs can read the lines and make placeholder audio files for testing. All actual dialogue stuff is done in external programs. The preview button is there to listen to what you just recorded in the event that you wish to do it again. It's not to listen to the actual audio.

The CK doesn't even recognize FUZ files (If they're there or not) which I find a little odd, but it's all designed like this for a reason!

Ah, thank you very much, that's one mystery solved.

To answer my other question about Say: it won't do what I want as part of a package, they're just too slow. But if I register a bunch of protesters for update and call Say in an OnUpdate event, I can get them each to chant the same thing, quite quickly. Not in time with each other, sadly, but real demos have that problem too.