badmojo wrote on 2024-12-15, 21:52:

Good to know thankyou. I did try DanielK's pack with my XtremeMusic but it was problematic to install (windows XP) and some games were crashing behaving strangely, so I ended up using the drivers on Creative's site. I'd prefer to use original drivers - no particular reason I can name for this. I'll have a look for the 2010 version.

I just uploaded the 2010 driver here on Vogonsdrivers. Also, I made a guide for installing the drivers and other useful utilities in order to ensure the best possible EAX experience on X-Fi Titanium cards.

P.S.

I still have the original driver DVD which came with my X-Fi Titanium Fatal1ty. I can upload that as well, if anyone's interested.

Regarding the model numbers, I have an SB0880 and an SB0886. Both have the black shroud/shield and have 64MB of ram.

My i7 system sports an Auzentech x-fi Forte. If you’re trying to track down an Auzentech card, this is the one to go for. The SNR is crazy, and the card is significantly louder than the Creative x-fi cards. If you’re buying used and want to use analog out, be aware that it’s a low profile card, and requires a dongle to provide the connections. Digital out works without the dongle as the s/pdif is built directly into the card.

Joseph_Joestar wrote on 2024-10-07, 19:50:

So I took a look at the two games from The Suffering series (GOG versions). Apparently, GOG ships the first one with the Indirec […]
Show full quote

KainXVIII wrote on 2024-10-04, 08:35:

According to this https://www.gog.com/forum/the_suffering_serie … nofficial_patch there is no EAX support in The Suffering series (just a placeholder?).

So I took a look at the two games from The Suffering series (GOG versions). Apparently, GOG ships the first one with the IndirectSound wrapper, which I deleted, as I'm testing it on WinXP using real hardware. The game's configuration utility correctly detected the 128 hardware voices on my X-Fi Titanium, and it allowed me to tick the "Enable EAX" checkbox:

The attachment The_Suffering_Config.png is no longer available

There's also an "Environmental Audio" setting under in-game audio options, which was turned on by default on my system. Now, I can't tell if/how the game actually uses the capabilities of EAX hardware, but the developers did include the proper checks at the very least. The second game uses the same configuration utility as shown above, but it lacks the in-game "Environmental Audio" setting.

I'm the guy who did some hacking on that game years ago. Admittedly, I never got the chance to try it on XP with Creative card, but Surreal Software's first game (Drakan) has EAX implemented, even Level Editor is available where you can set EAX properties for particular area in the level. Drakan also has a checkbox to enable A3D instead (other two checkboxes are irrelevant in that case since they refer to DirectSound3D and EAX extensions).

I asked about The Suffering many years ago on this forum as I was confused about supposed EAX support, I realized later what I thought was EAX was just buggy Creative's ALchemy + Host OpenAL (Sensaura OpenAL implementation apparently, the DLL is called Sens_oal.dll by default, if renamed to ct_oal.dll, ALchemy will use it instead of its internal implementation), which in some games just makes ALL sounds echoey, even when zero EAX calls are made on the API level. Maybe Sens_oal.dll has hardcoded default reverb enabled.

So nothing conclusive came out of that thread. I failed to find evidence of working EAX in that game. Only certain thing I did find, it uses DirectSound3D for positional audio. The whole thing is odd as EAX would be fitting for this kind of game when used smartly.

Also, in the picture above, you have confidently chosen your X-Fi card. But guess what, the game always uses primary device regardless of what is selected unless it's patched, the bug was there since Drakan. They store the setting, but don't use it. So probably they didn't test that setting on a system with multiple output devices.

And it seems no once since went through the game on period correct software / hardware and reported "Yes, EAX setting in this game is definitely effective in that area or that area".

Surreal Software's games were more on the buggy side in general.

Joseph_Joestar wrote on 2024-12-15, 22:10:

I just uploaded...

Thankyou very much for this, I'll try it out when my card arrives.

badmojo wrote on 2024-12-16, 12:33:

Thankyou very much for this, I'll try it out when my card arrives.

Enjoy! Meanwhile, I've also uploaded the installation CD which came with my X-Fi Titanium Fatal1ty (SB0886) here on Vogonsdrivers.

The drivers on that CD are slightly older (January of 2009) and they are specifically for the Fatal1ty card. Not sure if they will work with other X-Fi Titanium models.

I was looking at Creative's archived Sound Blaster Gaming page, and noticed this article about Prey that I hadn't seen before. If you ever wanted to know how much X-RAM Prey uses at any given time, this is how:

Creative Labs wrote:

Open the console again and type "listSounds" to see memory and X-RAM usage (only when using the "OpenAL" sound system)

And here's the result from my WinXP system, using an X-Fi Titanium Fatal1ty Pro while playing the first level of the game:

Onboard memory = X-RAM, per the aforementioned article. Looks like Prey, Quake 4 and Doom 3 can utilize it fairly well.

I found a way to reliably reproduce EAX occlusion in Baldur's Gate 2, so I thought I'd document it here. Basically, if you place a spell caster party member at the edge of the screen while keeping your protagonist in the center, the casting "chant" will have occlusion. Here's a simple example:

In the screenshot above, Imoen is standing near the left edge of the screen, while the protagonist is at the center. When I have her cast the spell "Armor" here's how that sounds when using VxD drivers:

And this is how it sounds when using WDM drivers:

The difference should be obvious. I've included a savegame if anyone else wants to test this. Be careful not to accidentally pan the viewport while performing the test. Imoen needs to remain at the left screen edge for the occlusion to occur. Of course, Environmental Audio needs to be enabled under Sound Options for EAX to work. Lastly, the occlusion effect seems to work best when the game is played at its default resolution of 800x600.

For reference, both tests were conducted on an Audigy 1 (SB0090) using VxD or WDM drivers under Windows ME. The latest official patch v23037 for Baldur's Gate 2 was applied, but the Throne of Bhaal expansion wasn't installed. As mentioned elsewhere, occlusion applies to a lot of sounds in this game, including enemy battle cries as well as certain environment effects like water dripping or a fireplace burning. Reverb is applied in caves and dungeons.

I feel a bit bad to circlestop, but I cannot avoid to underline how that all of this retro-dreaming seems a bit pointless (if not for testing purposes) when dsoal exists.

Joseph_Joestar wrote on 2024-10-03, 10:32:

FYI, I've added the "ALchemy Support" column to the EAX games list on the Vogons Wiki. This should make it easier to discern games which are officially supported by Creative's ALchemy utility.

Previously, this support was only indicated in the notes section, which led to needless clutter. Now, you can filter ALchemy supported games by simply clicking on the arrows in that column.

Even putting aside the kcat proselytizing, I don't know why it would be that worthwhile to mention. I mean, with a specific column at least.
I don't think it's that often that it doesn't work (not even sure on why or how it wouldn't, except the famous case of games ran as administrator and loading dsound through COM) and if any I'd be way more worried with cases of it sounding like crap.
Just last week I was reflecting on how funny it is that people always call it a day when in-game toggles stop to be grayed out, despite the fact that nobody but those using ct_oal (I mean, the real one, not renamed sens_oal) ever had any hope for it to sound right.

Joseph_Joestar wrote on 2023-07-27, 14:54:

Theoretically, OpenAL games should work and sound the same on Win7 and up. In practice, you need to check the PC Gaming Wiki and see if any sound issues are referenced there.

Specifically, Bioshock has missing sounds on Win8 and Win10, so you need to run it in WinXP compatibility mode on those operating systems.

I'm pretty sure that bug is just about the WASAPI renderer breaking when the system has no microphones. I don't remember if forcing openal could be a workaround, but nonetheless even if it didn't it would be just because the game audio thread completely dies.

Joseph_Joestar wrote on 2024-06-29, 14:18:

I wish the CoD devs had used EAX 5.0 via OpenAL, which works fine on Vista+ too.

BioShock did just that, and it was released the same year. For me, it has more precise positional audio compared to the EAX-less CoD4. I think this could be (at least in part) due to Creative's "Pure Path" tech, which is exclusive to EAX 5.0.

That's very funny to argue considering the game MAY or MAY NOT actually be the one where openal is inferior to directsound, due to FMOD possibly only using it for output (i.e. not mixing).

p.s. EAX5 is supposedly openal-only, but openal can also be just eax 4 or 2.

Joseph_Joestar wrote on 2024-09-27, 07:47:

[…]
Show full quote

1. If a game ships with eax.dll that means it's using EAX Unified. Check the version number of that file via right click > properties. For example, if it's version 4.0.0.1 that means EAX 4.0 and everything below is supported. Example game: Street Racing Syndicate.
2. If a game ships with OpenAL.dll or wrap_oal.dll open those files in a hex editor and search for the string "EAX". The versions of EAX that you find there are likely supported by the game. Example game: Pariah.
3. If all else fails, open the game's executable in a hex editor and search for the string "EAX". The versions of EAX that you find there are likely supported by the game. Example game: Halo CE.

Nothing of these is correct, because there's plenty of boilerplate code that games could have linked from various SDKs despite the fact they don't use the newer and shinier features.
This is especially true for openal if you even understand what the router and the wrapper are supposed to do (I guess that at least with Unified, you could kinda lowkey infer that there has to be at least some higher "advanced" version to downgrade from, even though you cannot be sure on the specifics).

The only 99% certain way is checking what calls the game does (dsoal should log that for you) or trying to see with ghidra or whatnot if the game is even "using" those strings to actually check for anything.

Joseph_Joestar wrote on 2024-11-29, 19:19:

Lastly, note how the game actually detects EAX 5.0 capabilities on the X-Fi, while not finding that on the Audigy 2 ZS (see the lines right below Init: AL_EXTENSIONS : EAX).

That seems quite literally just a call to alGetString(AL_EXTENSIONS).
(and I haven't checked UT3 in ghidra, but GoW source code is evidently just that)

Joseph_Joestar wrote on 2023-07-13, 15:36:

It depends on the game, and how it was coded.

Generally speaking, games made before 2001 were likely coded with Win98 in mind, and expect VxD drivers for EAX support. Some of those games (but not all) can be buggy when WDM drivers are used. NFS4 and Baldur's Gate 2 are two examples that I personally know of. There could be more, but it's probably not a widespread issue.

I don't think (and I don't see how) games were coded to discriminate between NT and 9x.
It's far far far more likely instead, that the drivers themselves were quirky in different ways (in fact they literally rewrote the entire architecture once they could drop vxd)

Joseph_Joestar wrote on 2024-12-28, 11:24:

I found a way to reliably reproduce EAX occlusion in Baldur's Gate 2, so I thought I'd document it here.

Joseph_Joestar wrote on 2023-07-13, 13:27:

Not sure, I've never played NFS4 under WinXP. Maybe someone else can help you with that. It should be noted that WinXP cannot use VxD drivers, only WDM.

And while I haven't tested NFS4 under WinXP, I did try Baldur's Gate 2. That game loses some occlusion effects when WDM drivers are used, and therefore sounds slightly worse under WinXP. It's not a huge deal, but those occlusion effects are kinda neat when enemies are off-screen or behind closed doors.

It would be really really neat if you could compare that with dsoal.
Or discuss that with the GemRB guys, which surely knows a thing or two about the game internals.

mirh wrote on 2024-12-29, 16:52:

I feel a bit bad to circlestop, but I cannot avoid to underline how that all of this retro-dreaming seems a bit pointless (if not for testing purposes) when dsoal exists.

I'm not that familiar with dsoal. I did use it for restoring EAX to a few games that I was running on Win7, but that was maybe 6-7 years ago, i.e. before I got back into retro hardware. Is there a way for dsoal to detect which EAX version(s) a game uses? If so, that would greatly simplify testing.

Even putting aside the kcat proselytizing, I don't know why it would be that worthwhile to mention. I mean, with a specific column at least.

Honestly, my main reason for splitting this into its own column was to clean up the notes section. All of the "Supported by ALchemy" notes were inherited from the previous version of the list (i.e. from Wikipedia) it's not something that was added by me or anyone else on Vogons. The only thing I changed is this: what was previously a "Supported by ALchemy" entry in the "Notes" column is now an "X" in the "ALchemy Support" column.

Nothing of these is correct, because there's plenty of boilerplate code that games could have linked from various SDKs despite the fact they don't use the newer and shinier features.
This is especially true for openal if you even understand what the router and the wrapper are supposed to do (I guess that at least with Unified, you could kinda lowkey infer that there has to be at least some higher "advanced" version to downgrade from, even though you cannot be sure on the specifics).

I get that certain games may have just included the relevant .DLLs by copy/pasting from some SDK, but if such a game is also on Creative's "EAX Featured Games" list, I think it's worth taking into consideration.

It's far far far more likely instead, that the drivers themselves were quirky in different ways (in fact they literally rewrote the entire architecture once they could drop vxd)

This is possible of course. The end result is the same though. Certain games sound different depending on which driver type (VxD or WDM) is used. And since VxD drivers cannot be used on Windows 2000/XP/Vista+ such games will sound differently on those operating systems when using real hardware.

It would be really really neat if you could compare that with dsoal.
Or discuss that with the GemRB guys, which surely knows a thing or two about the game internals.

I think this should be pretty easy for anyone to check on their own, given the test case that I've posted earlier. Use whatever means to restore EAX support to the game, then compare the output to the recording that I've made on real hardware. As to what causes Baldur's Gate 2 occlusion effects to behave differently depending on whether VxD or WDM drivers are used, I'll happily leave that to someone with more knowledge of the game code. I'm not a programmer, just a hobbyist and fan.

Joseph_Joestar wrote on 2024-12-29, 17:16:

I did use it for restoring EAX to a few games that I was running on Win7, but that was maybe 6-7 years ago, i.e. before I got back into retro hardware. Is there a way for dsoal to detect which EAX version(s) a game uses? If so, that would greatly simplify testing.

3 years ago it didn't even support EAX.
Now you can just use DSOAL_LOGLEVEL (not sure the minimum required value, but still) and it should be there.

To be super duper fair, just one or two appearances (especially at the beginning) MIGHT just be part of whatever lousy enumeration a game has, but if you see the effects themselves getting called multiple times that's as much evidence as you may ever want.

Joseph_Joestar wrote on 2024-12-29, 17:16:

This is possible of course. The end result is the same though. Certain games sound different depending on which driver type (VxD or WDM) is used. And since VxD drivers cannot be used on Windows 2000/XP/Vista+ such games will sound differently on those operating systems when using real hardware.

That's the thing though.
You are confusing sblfx.dll with e10kx2k.sys and ct_oal. If it's not the game to be doing anything stupid, then it's just a matter of the driver behaving like it should or not.

mirh wrote on 2024-12-30, 00:19:

3 years ago it didn't even support EAX.

It may have just been DirectSound back then. As I recall, the first time I heard about dsoal was from some guide on getting Fallout: New Vegas running optimally when using Windows Vista and newer. I tried it with that game and maybe two others, but that's about it.

Now you can just use DSOAL_LOGLEVEL (not sure the minimum required value, but still) and it should be there.

Sounds good, I'll give that a look. During my testing, I found a few games that had an "Enable EAX" checkbox in their options menu, but no other data to indicate which EAX version(s) were supported. If dsoal could help clear that up, it would be great.

Does this logging require the game to actually play an EAX effect? Or is it enough to simply start the game with the EAX checkbox ticked in its options menu while using dsoal?

You are confusing sblfx.dll with e10kx2k.sys and ct_oal.

Not sure where that was implied, as I never mentioned any of those files. To be clear, I'm simply reporting the differences that I've heard while playing Baldur's Gate 2 when using VxD vs. WDM drivers on real retro hardware. The recordings I made were meant to showcase that. Phil did something similar with Need for Speed 4: High Stakes in this video (around the 7:21 mark). Again, I'll happily leave investigating how and why this happens to someone who is more qualified to do so, as I'm not a programmer nor an audio expert.

Joseph_Joestar wrote on 2024-12-30, 03:22:

It may have just been DirectSound back then.

I just meant that today it's far ahead (if still not imperfect) so any experience is probably going to be different.

Joseph_Joestar wrote on 2024-12-30, 03:22:

Does this logging require the game to actually play an EAX effect? Or is it enough to simply start the game with the EAX checkbox ticked in its options menu while using dsoal?

I mean, there has to be a call?
Some games may do some wacky shit (like for instance we discovered that GameCODA only uses openal if it can also detect EAX3.. or maybe it was that it can use EAX3 with openal only if detects that the vendor is creative) but generally directsound should always go through IKsPropertySet::QuerySupport for any checks.

Joseph_Joestar wrote on 2024-12-30, 03:22:

Not sure where that was implied, as I never mentioned any of those files. To be clear, I'm simply reporting the differences that I've heard while playing Baldur's Gate 2 when using VxD drivers vs. WDM drivers on real retro hardware.

If it's a game thing.. Then there might or might not be a doable workaround that isn't patching the exe - but regardless it's still going to be non-trivial to find.
If it's a driver thing, then for what you know any single version could have been a fix. (to be super fair it might even be an OS thing.. but you would guess that regardless there's nothing that the exclusive driver couldn't still take care of)

EDIT: though now that I think to it even better.. AFAIU sblfx.dll was supposed to be the common driver between both VxD and the 9x and NT WDM versions. It would seem certainly strange if the same driver version with the same OS and everything, made a difference just by changing that. Could you or anybody else reproduce it with another card?

mirh wrote on 2024-12-31, 00:37:

Could you or anybody else reproduce it with another card?

Sure, I've reproduced it on an Audigy 2 ZS (SB0350) with the same results: occlusion is present with VxD drivers, while there is no occlusion with WDM drivers. As far as I know, that's the last Creative card which had VxD driver support. Also, I made a few more recordings with the ambient sounds and music muted in the game's sound options to make the voice lines and the occlusion effect stand out a bit more:

As previously, VxD drivers provide occlusion. Meaning, the "chanting" voice and the spell completion effect are muffled.

While there is still no occlusion with WDM drivers.

I did a test with dsoal as well. For reference, I downloaded dsoal.zip from here which I think is the latest stable version (currently it lists DSOAL v0.9.6 + OpenAL Soft v1.24.0 dated November 16th from this year). There is no occlusion with this version of dsoal, though I did notice something else. Looking at the waveforms in Audacity, both VxD drivers and dsoal pan hard to the left, which is correct, as the character is standing on the left edge of the screen. However, WDM drivers don't seem to pan that hard, with some sounds still coming from the right speaker.

EDIT - I double checked and it looks like CMSS was accidentally turned on during my WDM recording, which messed up the panning. Apparently, it automatically turns itself on when you switch drivers or sometimes even when you change your speaker configuration. Therefore, I re-did and re-uploaded all three recordings while making absolutely sure that CMSS was turned off.

I think you're probably right that the occlusion functionality in Baldur's Gate 2 is driver related. It's possible that Creative changed their minds on how this should work at some point, but the game was likely coded with the old driver behavior in mind.

In other news, I was testing my recently acquired X-Fi XtremeGamer Fatali1y Pro and noticed that there's an X-Fi Demo Disc on Vogonsdrivers. I downloaded it out of curiosity, even though it originally came with the X-Fi Elite Pro, wondering if it would work on my card. And sure enough, it does.

Lots of interesting info there. For me, the absolute highlight is the ability to read up on any EAX feature (e.g. MacroFX, PurePath, Occlusion etc.) and then run a demo which showcases it in action. In particular, the MacroFX demo sounded great on my 5.1 surround speakers. Also, if you're using headphones, be sure to give the CMSS-3D demo a try.

Another thing of interest might be the game promo section. Basically, it's a local copy of Creative's "EAX Featured Games" website from mid 2005. Pretty neat.

All in all, if you have an X-Fi sound card, I highly recommend checking out this demo disc. I tried it on my X-Fi XtremeGamer Fatali1y Pro (PCI) and my X-Fi Titanium Fatal1ty (PCIe) and it worked fine on both cards.

A word of caution though. This demo disc was released during the WinXP era, and it might use some Macromedia Flash elements. I'm not sure how well that would work on a modern OS.

Sorry.. I meant with a card of another vendor (or at least something older like a Live). All those cards are EMU10K2-based, with pretty much the same exact driver.

p.s. I got notice that openal-soft had some sounds muffled in the last couple of versions, you may want to pair dsoal with a build from the last week or so (and there's also the c++-rewrite branch of interest)

mirh wrote on 2024-12-31, 13:58:

Sorry.. I meant with a card of another vendor (or at least something older like a Live). All those cards are EMU10K2-based, with pretty much the same exact driver.

I just tried it on my Sound Blaster Live! 5.1 (SB0100). That's the oldest EAX 2.0 capable sound card that I have. The results were interesting. When using the original installation CD from 2001 both VxD and WDM drivers properly provided occlusion in Baldur's Gate 2.

I even tried a newer SBLive installation CD from 2004 and those WDM drivers still properly provided occlusion. At a guess, Creative changed something within the Audigy (and X-Fi) code branch of the WDM drivers. After that change was made, BG2 occlusion became broken on those cards when using WDM drivers, while still remaining functional with VxD drivers. Either that, or there's a hardware difference between SBLive and Audigy cards that WDM drivers don't account for.

One more thing to note. Examining the waveforms of the SBLive recordings in Audacity reveals that neither VxD nor WDM drivers pan fully to the left. This isn't particularly surprising to me, since SBLive cards have some kind of proto-CMSS which is always active. For example, they auto-expand pure stereo sources (like CD Audio music) to surround when 5.1 speakers are connected. I haven't found a way to disable this undesirable behavior, which is why I rarely use SBLive cards nowadays.

mirh wrote on 2024-12-31, 13:58:

I got notice that openal-soft had some sounds muffled in the last couple of versions, you may want to pair dsoal with a build from the last week or so (and there's also the c++-rewrite branch of interest)

I tried pairing DSOAL v0.9.6 with OpenAL Soft v1.24.1. There was no change to the previous test, BG2 occlusion still didn't work with dsoal. I also tried the binaries from the current OpenAL Soft Pre-release build, and there was no change.

Joseph_Joestar wrote on 2024-12-31, 19:31:

One more thing to note. Examining the waveforms of the SBLive recordings in Audacity reveals that neither VxD nor WDM drivers pan fully to the left. This isn't particularly surprising to me, since SBLive cards have some kind of proto-CMSS which is always active. For example, they auto-expand pure stereo sources (like CD Audio music) to surround when 5.1 speakers are connected. I haven't found a way to disable this undesirable behavior, which is why I rarely use SBLive cards nowadays.

There's quite some contention (registry included, I wonder if RegMon couldn't help in this regard) about how CMSS1 could be disabled.
Putting even aside that it may automatically trigger with EAX. But even if it was trying the A2 software or the Kx drivers I'm sure something could be done.

Joseph_Joestar wrote on 2024-12-31, 19:31:

I also tried the binaries from the current OpenAL Soft Pre-release build, and there was no change.

I see, then it would be cool if you could open a bug report.

mirh wrote on 2025-01-03, 15:21:

I see, then it would be cool if you could open a bug report.

I'm too lazy to create a new GitHub account just for that, but maybe someone else can do it based on my notes. To summarize the reproduction steps:

1. Install Baldur's Gate 2 and apply the latest official patch v23037
2. Download the save game I attached to this post
3. Load the save game, select Imoen and have her cast the Armor spell

Observed behavior:
Imoen's spell casting chant has no occlusion effects applied

Expected behavior:
Imoen's spell casting chant should have occlusion effects applied

For an example on how that's supposed to sound, listen to the BG2_EAX_VxD.flac audio clip attached to the same post as the save.

What do you suggest for best EAX experiment ?

option 1. modern 5.1 or 7.1 modern headphones ?
option 2. real 7.1 or 5.1 system with speakers & amplifier ?
option 3. just good stereo headphones ?

Funny that you still need these old gaming setups to experience the original intended gameplay, since there's no EAX on steam/gog.

paatudos wrote on 2025-01-20, 15:34:

What do you suggest for best EAX experiment ? […]
Show full quote

What do you suggest for best EAX experiment ?

option 1. modern 5.1 or 7.1 modern headphones ?
option 2. real 7.1 or 5.1 system with speakers & amplifier ?
option 3. just good stereo headphones ?

Avoid option 1, as that's not needed for EAX. Just get a pair of good quality stereo headphones (e.g. something from Sennheiser) and you're good to go. The EAX and CMSS-3D algorithms will take care of all the positional audio processing. But if you're like me, and find wearing headphones for a longer time period uncomfortable, go for 5.1 or 7.1 speakers instead.

This is assuming an actual Creative Labs sound card is used of course. I can't speak for on-board audio solutions from other manufacturers, or for software emulation.