13 replies to this topic

[Kwiat_W]

So I have done the Pack Mule mod and noticed something annoying. The mule walking animation has the same walking sound as dragon animations- you hear slow loud stomps. Is there any way to change it? It seems to me that it is the default sound for all new animations. Does anybody know if changing the animation slot type would help? (currently it's IWD type 5+3E)

[Miloch]

In this post, devSin said:

Walking is the same boat (most of the hardcode IWD slots are set to use that loud-ass stomp LAZY, but I guess the IAs slots let you pick your own?), although if you wanted to have a little remote-control car dealy, you could probably use the slot to great effect and have the little thing buzz around buzz buzz buzz (i.e., the walk frames are rarely one complete step only, so you're never going to get anything that actually sounds like footsteps).

But I have never heard that sound, and I'd think I would've, or someone else playing Aurora's Shoes, because many release ago it added NPCs with the IWD svirneblin animations. Maybe I have my sounds turned down fairly low, but I doubt everyone does...

You could try overriding it with a custom 2da file for your animation sounds (which you probably want anyway) by putting something in the WALK slot. Possibly even a nonexisting sound clip. Having said that, the svirfneblin don't use 2da files (their sounds have text strings associated, so they're coded on the CRE files). Changing the animation type might help (such as to BG2) but then you'd have to reorganise all your BAM sequences.

If walking sounds are indeed hardcoded in the .exe, someone who knows about hacking that would have to look at it (like Erephine, Ascension64 or Taimon, all of whom have been pretty scarce lately). But if you think you could figure it out, be my guest. The easiest way perhaps would be to remove the .exe sound references and/or externalise them to the .2da files. Ascension64 has already done something like this with his ToB Extender, maybe someone could get him to look at it.

[Kwiat_W]

Thanks for your replay.
The 2da file isn't helpful here, changing reference in the sound slot only adds another sound that is played additionally to the current walking sound and it's not played continuously but only from time to time.

Messing with exe is definitely beyond me. I'll have to check what sounds are associated to the bg1 5+3e animation slots.

[Kwiat_W]

"BG1 simple monster" could be used, it has a different sound, but not necessary better, I would describe it as "walking on sand" or "quiet dogs snorting" (this is the same sound that werewolf animation use). Now I'm thinking about replacing sound files that are responsible for the "dragon stomps" with blank files.

[Miloch]

Now I'm thinking about replacing sound files that are responsible for the "dragon stomps" with blank files.

Hmm, didn't think of that, but that's a possibility perhaps. What are the sounds called? I suppose it might be possible to replace what few IWD dragons there are (like... none?) with those sounds in their WALK 2da slots. Wouldn't matter much if it didn't match exactly with their footsteps, since it's such a big sound (though like I've said, I've never actually heard it with my own IWD animations...).

[Kwiat_W]

WAL_09(a-d). These sounds are played not only for the IWD dragons but for those from BG2 too and for other animations: glabrezu, wailing virgin, yuan-ti, lizardmen, salamanders, myconids, orcs and some more. During my tests I have also noticed that I can never hear footsteps sounds during combat. I know there is a "Disable Footsteps During Combat" option in baldur.ini, but setting it to 0 or 1 has no effect, footstep sounds always stop playing when I give order to attack or when hostile creature shows on screen. Can someone confirm that this is game bug and not fault of my soundcard/drivers?

Edited by Kwiat_W, 24 July 2010 - 01:04 AM.

[Ascension64]

I can confirm that all walking sounds are hard-coded. Additionally, they are animation dependent. Some use fixed animation sounds, while others use animation sounds depending on the terrain type indicated in the area search bitmap.

What would be the preferred way of correcting this for IA?

[Miloch]

I can confirm that all walking sounds are hard-coded. Additionally, they are animation dependent. Some use fixed animation sounds, while others use animation sounds depending on the terrain type indicated in the area search bitmap.

What would be the preferred way of correcting this for IA?

Maybe it'd be best to have hardcoded sounds based on selection circle (creature size)? That way dragons would get dragon stomps, humanoid creatures would get regular-sounding footsteps. Then also have some means of overriding them, possibly via the existing WALK slot in 2da files and/or maybe an unused .cre sound slot (there are many of these after the biography). That way if you want your marilith to slither instead of clop around, you could override whatever's hardcoded (though in this case, you'd need to do it on your CRE file since the vanilla marilith shares a slot with the balor tanar'ri MTAN and you wouldn't want him slithering).

Is that doable with something like ToB Extender? I'm all for whatever's easiest, implementation-wise.

[Ascension64]

You can muck around with a completely externalised walk sounds by animation ID.

http://www.shsforums...post__p__497469

[Miloch]

You can muck around with a completely externalised walk sounds by animation ID.

Sweet. Next thing you'll tell me is you've managed to externalise the animation IDs themselves (or "internalise" them to actual functional versions of animate.ids and anisnd.ids) so we don't have to hack the .exe anymore .

(Though I don't know if that would make it OS X compatible, since TobEx probably isn't either...)

[Ascension64]

Externalising anim IDs is a lot more work because animation properties are hard-coded into specific classes in the EXE. I assume this is why certain animations slots can only be used for certain animation types with IA. (unless I'm wrong here)
Because I haven't been able to get hold of the proper technical specs of how the EXE patching is done, let alone reversing the entire set of changes, supporting this is a bit far-reaching at the moment, but not impossible.

TobEx is definitely not Mac OSX compatible at this time. I am uncertain as to whether it ever will be. I personally have no experience in programming for Mac.

[Miloch]

Externalising anim IDs is a lot more work because animation properties are hard-coded into specific classes in the EXE. I assume this is why certain animations slots can only be used for certain animation types with IA. (unless I'm wrong here)

Yeah. The selection circles are also hardcoded to slots.

Because I haven't been able to get hold of the proper technical specs of how the EXE patching is done

I guess infinityanimations/documentation/devnotes.txt didn't help? There's not much detail, but from "Reference offsets in the executable" onward might help somewhat.

[Ascension64]

Yeah. The selection circles are also hardcoded to slots.

There is a lot more to it than that. There are hard-coded limitations on what animations are static, how many parts an animation has (big monsters use 2x2 grid, dragons 3x3, etc.), number of orientations, and how BAM files are segregated for each animation. Probably more. This is what makes animations so complex.

I guess infinityanimations/documentation/devnotes.txt didn't help? There's not much detail, but from "Reference offsets in the executable" onward might help somewhat.

I am after exactly how they changed the code. A quick survey of the modified EXE reveals quite significant changes to the way animations are constructed, but I cannot get any detail at this time unless I reverse all the changed code.

Edited by Ascension64, 24 September 2010 - 05:17 AM.

[Miloch]

There is a lot more to it than that. There are hard-coded limitations on what animations are static, how many parts an animation has (big monsters use 2x2 grid, dragons 3x3, etc.), number of orientations, and how BAM files are segregated for each animation. Probably more. This is what makes animations so complex.

Yeah, all that too. If you launch the ia_referencepicker.exe and then open the reference chart it describes all the animation types in a fair amount of detail.

I am after exactly how they changed the code. A quick survey of the modified EXE reveals quite significant changes to the way animations are constructed, but I cannot get any detail at this time unless I reverse all the changed code.

Well only Erephine knows and she's been MIA for some time. All I can say is she used bsdiff and bspatch to generate the original patch but as to how she made the determinations for changing the .exe, I've no clue.