Thanks for spotting these, I've updated the docs and it'll be fixed with the next release.
- Status
Thanks for spotting these, I've updated the docs and it'll be fixed with the next release.
Interesting one. The information is in EDDI and available to responders in general but doesn't make it to VoiceAttack at the moment because VoiceAttack doesn't handle arrays. There are workarounds, but none of them are very clean.
Yes, the default speech responder will read out player-sent messages for which you are the recipient (direct, local or wing). It also understands o7!
You need at least version 1.5.12.22 of VoiceAttack, which requires you to download the beta.
If you see the 'plugin EDDI 2.0.4 initialised' message in VoiceAttack when you start then EDDI is up and running. There is no startup command any more, which is a good thing because it caused many problems with profile editing, clashes with HCS and other stuff.
As I say EDDI has both and the data makes it to the responders (so the speech responder, for example, uses it), the problem is presenting the information as VoiceAttack variables because of the lack of an array type.
jgm, could you elaborate on this?
From my uneducated perspective, it doesn't seem any less clean than the numerous sets of ship compartment variables created.
FTR, not trying to be snarky... I followed the 'vaproxy' variable discussion in the VA forum and am sincerely interested in the issues that you deal with in coding around the API & Voice Attack.
Those will need to be installed through Windows. EDDI just takes from what is already installed on the system. Depending on your OS version, Win7, Win8, Win10, it can either be tricky, or easy to do
ETA: Looking closer, it appears you have Win10? You will need to install additional windows language packs. I don't have my Windows 10 machine in front of me, but here is a fairly decent guide...
http://answers.microsoft.com/en-us/...ndows-10/f920fad9-a64e-4510-b0e5-31e14688fa49
Last edited:
No, the voice that you heard in the preview was from the speech responder. I really must put together a little architecture diagram to explain how EDDI works now, but in short:
- Monitors obtain information and send them to EDDI as events. Current monitors include Elite's journal and netlog, and the Galnet RSS feed
- EDDI receives the events from the monitors and processes them to update its internal data structures. For example, if a 'jump' event is received then EDDI will update the current system
- EDDI sends the events to responders. Current responders include speech, EDSM, EDDN, and VoiceAttack
- Responders act on the events, for example the speech responder reads a script, the EDSM responder sends jump information to EDSM, and the VoiceAttack responder populates VoiceAttack variables and calls scripts
In this case the speech responder was responding and had access to all of the variables. A VoiceAttack script wouldn't have access to all those variables (for reasons explained below).
The difference between the ship information and event information is that the ship information is a hard-coded special case, whereas the event information is populated using introspection. There's nothing to stop the VoiceAttack responder being able to handle the generic info and putting it in to VoiceAttack in some form, but I haven't coded anything up to do so yet. Once 2.0 stabilises I'll start to think about this (and with any luck VoiceAttack will have implemented arrays themselves by then, which will make life easier).
Other than the Microsoft provided voices, which are sub-par IMO, you have to buy additional ones on your own. Some kind soul up-thread provide a link to Ivona which provides TTS voices. Most commonly used is 'British Amy'.
Best guess is that you have installed 32-bit voices. They show up in the TTS list but aren't actually available when it comes to speaking through the Windows TTS synthesiser. Which voices in particular are they?
- Status