[SQ4CD] Ending with some missing lines...

[L@Zar0]

Hi,

I found a post "SQ4 Intro (Missing Subtitles)" in "General and Everything Else" subforum:
https://sciprogramming.com/community/index.php?topic=2031.0

As I'm working in the translation for SQ4 CD, I've reached the ending and seen that some of the ending lines are missing (in fact, they show too fast, or even near don't show, with NRS Update 3 or not). I seen that @doomlazer has fixed this deleting the lines because presumably, the voices does not exists. But, in fact, the SQ4 CD has that lines, but I don't understand how it finds (and play them).

This is part of the code showing the text or the speech depending on what we have set (Text or Speech):



And here is the Prop when using Speech in normal vanilla game:



Ok, here it says, talkerNum 34, but I can not find this talker in messages file Message.557. But it seems that plays the audio (Roger and Roger Jr. talking at the same time). Maybe it is a sound, I don't know. Apart, in Message.557, the text lines exist, but the voices not (although they seem to play in speech mode).

I would like to know how the game finds that voice AND I would like to know if it is possible to do some type of fix/patch when using Text and Speech settings at the same time (like in ScummVM).

[L@Zar0]

Ok, I finally achieved something, altough I don't like the scripting method:

https://i.imgur.com/WjfZjbp.mp4

This would be part of the code:



For the second lines, I use: "(DoAudio 2 557 34 0 2 1)"

In any case, I don't understand how the hell the game access that audios. They are not in Message.557, but I can see they are in 557.map file, at the end.

[doomlazer]

In any case, I don't understand how the hell the game access that audios. They are not in Message.557, but I can see they are in 557.map file, at the end.

Open 557.map in SCIViewer and it shows the noun, verb, condition, sequence along with its audio offset, which is the offset in AUDIO.VOL where the voice wavs are stored.

[L@Zar0]

That's what I seen, yes.

This is a bit different than the method used in KQ5CD, that's why I'm a bit confused. I understand that the compiler uses some kernel display function (kDisplay/kPrint?) for message texts strings given a noun/verb/cond/seq (in message resources tab) and the speech uses some kernel audio function (kDoAudio) for voices given a noun/verb/cond/seq (in map resources tab, and include the mouth sprites?).

I wish I had this in KQ5CD, it had been a bit more easy to add subtitles.

The problem I see in this case is that I should extract that voices and add to message resource to use them as normal method used in SQ4. So, I think I will leave the code with the little solution I found.

[doomlazer]

Yes, the KQ5 talkie is different from all the others. The voice lines in KQ5 can be patched with .aud files. Everything else requires rebuilding the entire audio.vol and/or using a diff patch for distributing audio changes - at least AFAIK.