37 réponses à ce sujet

[Nitro9a]

Ah, I see now. [Continue] is automatic if you've got no text in the line. Thanks!

[Avaraen]

Nitro9a wrote...

Thanks, but the problem I'm having is that the .wav files do not override the individual lines of the .fsb. I'm making some progress though. I found a program (FMOD) that is able to make new .fsb files with the .wav files I've created, I'm still ironing out the kinks but it's a step in the right direction, I think. If/when I manage to work it all out I'll post what I've learned. Thanks again Avaraen for brainstorming with me.

If you extract all the .wav files from the original .fsb into your toolset directory, overwrite the .wav files as necessary, and then Generate VO for the conversation, won't it create a new .fsb for you with your changed files included? The toolset seems to automatically run FMOD when selecting that option from Tools > Generate VO, so theoretically an outside program or external FMOD application shouldn't be neccessary.

I'm working on some additional tweaks and additions which include adding some new dialogue (using existing/edited OC dialogue), and it is creating a working .fsb directly from the toolset. The downside is it seems I have to create a new .fsb including all the OC files for that dialogue whether they're changed or not. And for some reason, even with the correct line ID's, the VO generation is not picking up all the new .wav files, so some of them work, and some of them don't. Even more fun is when trying to include all the original .wav files for the dialogue, I'm getting an fmt chunk error... good times. Hope you have had better luck than I have. XD

BTW, has anyone figured out how to create a new comment? For example, in Zevran's dialogue file, there's a "[COMMENT] IF: Zevran has been recruited" line that is putting the Scripting Comment directly into the displayed line. I have a feeling I'm missing something painfully obvious, but I cannot seem to figure out how to create those comment lines myself.

[Nitro9a]

Avaraen,at what point are those files getting created where are those new .fsb files getting stored?



.wav files have to be 16bit mono 24 kz PCM. That might be why they are not getting recognized. Let me know if that's the case.

[KingEtzel]

The new FSB ( and FEV btw) get created the time you press "Generate VO" in the conversation editor (with ROOT selected in the conversation!). The program will then pick all WAVs as needed from the respective folder (see DABuilderWiki for this) and packs them into a FSB/FEV set for this specific conversation.

The directory for the temporary FSB is then this directory here:

C:\\\\...\\\\Documents\\\\BioWare\\\\Dragon Age\\\\Toolset\\\\resourcepostintermediate\\\\win32\\\\processed\\\\addins\\\\[MODULE-NAME]\\\\audio\\\\win32\\\\packages\\\\core\\\\vo\\\\en-us\\\\[DLG-NAME]

It stays there, until you produce a player package...

Modifié par KingEtzel, 18 décembre 2009 - 03:21 .

[Avaraen]

The new .fsb is put in the resourcepostintermediate subdirectory as KingEtzel mentioned, and also gets put in packages\\core\\override, so you have to delete it twice if you want to get rid of it, heh.



I'd originally been using a sound editor that couldn't save to 24k Hz (22k or 32k was as close as it got), and found out the hard way the toolset throws an error and will not generate VO at all if the files are not in the correct format (it specifies the format issue). I got a different editor and converted my new .wav files to the PCM 24k Hz 16bit format, and didn't get any error messages when generating VO, but with only my modified .wav files in the override directory, it picked up some of them, and generated entirely new VO for the rest of the dialogue file. That new .fsb overrode the original voice files with the robovoice, so I moved the extracted contents of the original .fsb into the override directory and tried to re-generate VO... and that's when I got the fmt chunk error in the toolset log. Since I did not get that error with only my files, I reckon the problem is with the extracted files. I'll continue with it tomorrow, as I've been awake about 19 hours now and my mental focus is pretty much gone now.

[Nitro9a]

KingEtzel wrote...

The new FSB ( and FEV btw) get created the time you press "Generate VO" in the conversation editor (with ROOT selected in the conversation!). The program will then pick all WAVs as needed from the respective folder (see DABuilderWiki for this) and packs them into a FSB/FEV set for this specific conversation.

The directory for the temporary FSB is then this directory here:

C:\\\\\\\\...\\\\\\\\Documents\\\\\\\\BioWare\\\\\\\\Dragon Age\\\\\\\\Toolset\\\\\\\\resourcepostintermediate\\\\\\\\win32\\\\\\\\processed\\\\\\\\addins\\\\\\\\[MODULE-NAME]\\\\\\\\audio\\\\\\\\win32\\\\\\\\packages\\\\\\\\core\\\\\\\\vo\\\\\\\\en-us\\\\\\\\[DLG-NAME]

It stays there, until you produce a player package...

Okay.  Right now I've got four new lines of dialogue, four new .wav files (in the correct format with the correct line ids), and they are in [moduleid]/module/core/override/toolsetexport.  The wavs are also in singleplayer/override/toolsetexport because I noticed that that's where the .dlb and .dlg files are getting stored since modifying a dialogue modifies the single player campaign.  So, just in case.  When I generate vo with root selected I get three .fsb/.fev files for individual lines, only one of which are mine.  It should be generating a single .fsb with all of the original (1000+) lines and my four new lines, correct?

[Avaraen]

I've finally had some success with adding new sound files. As I mentioned above, I had issues with the fmt chunk on the wav files when generating VO with the original dialogue wav files included. Experimenting with various sound editors, I found some of them would not load the original wav files; whether it's an issue with fsbext or the original packaging of the wavs, it seems the extracted wav files had no format data associated with them. Goldwave has a batch converter that allowed me to convert all the extracted wavs to the correct 24kHz format... from there, the toolset was more than happy to create a new .fsb and .fev file for me. So far, the new sound and the original dialogue sounds seem to be working fine (very minimal testing).

[Nitro9a]

Thanks Avaraen. I would have never guessed that the extracted files would be in the wrong format. I was only worried about mine. Go figure.

[Lady of Lore]

Ouch, gotta come back and read this at not such a late hour. I'm working on trying out some conversation modifications,, or really just a few small additions something like the slap morrigan mod only simpler with Alistair.

I'll come back and read through this more carefully tomorrow. I tried to right now and felt my eyes dry out.



Thanks for all the people who are posting things on here. Some of the questions were the same I had!

[Lady of Lore]

Ok, its starting to make sense now that I'm more awake. I was trying to figure out how to get fmod to open the dialog files but it looks like I'm missing a step. The player seems to only accept fev files....hmmmm....*does some more digging*

[Nitro9a]

Hey Lady of Lore. I'm not clear on what you are trying to do. Are you trying to just play the wavs, extract them, open the dialogues in the toolset, or something else entirely?

[Lady of Lore]

Nitro9a, I'm trying to get vo lines and save new copies of them to add more dialog. (From what I've figured out, you need to have the dialog line id and have the wave file with the vo that you want to have the same id for when the program searches for it to play it in game.)



So the step I'm on now is to pull the wav files from the fsb compilation and create copies of the existing vo lines such as "Now that the warm fuzzy part is over..." so I can have Alistair really responding to the lines that I'm adding.



I'm not changing anything from the existing dialog, just adding a few lines here and there sort of like the slap morrigan add on where its just tagged into the dialog options.

[Mengtzu]

The 60 characters limit is probably because that's what fits in one line of text for a standard definition tv. However, the PC really shouldn't be saying that much in one go. Don't make us read

Yeah, I had some...verbose...PC lines early in my module and got feedback that they were running off the edge even on some PC monitors.  PC lines never wrap!

I think for a mod you can push it a bit past 60 characters, but it's unwise to go much more.