Swapped the sound card again, put back Audigy Rx. No bugs so far, having Audigy Control Panel open and launching anything invoking Creative's OpenAL doesn't break EAX settings, that happened on Win11 23H2, but not on Win10 20H2.

Played the old new 64-bit CoD:MW2 for little over 1,5 hours, sound still works (tends to break with Sound BlasterX AE-5 Plus when XAudio2 is involved).

No idea if this again one of those things that work OK on one OS version, but not another.

The old builds of CoD:MW2 use Miles Sound System. They actually updated that game so many years later (presumably due to infamous remote code execution vulnerabilities) and the new 64-bit binaries use XAudio2. But Black Ops 1 hasn't been changed since 2013, this was the first CoD that dropped Miles in favor of XAudio2, where I first encountered that problem with Sound BlasterX AE-5 Plus. If I had anything else installed using XAudio2, I would have tried it, but I only know about Wolfenstein: The New Order that uses it and this one is kinda big disk space wise.

Finally, game changing discovery on Audigy Rx. If you keep speaker configuration in Audigy Control Panel at 2 / 2.1, you can turn on CMSS and it won't mess up the stereo playback. Well, it's actually default configuration. So playing plain stereo music files sounds exactly the same on headphones with CMSS on or off. At the same time, you can still pick surround configuration in Windows and you'll hear 3D effect in those newer games that do their own mixing and are multi-speaker aware.

OK (not actually OK), EAX can still be killed if game doesn't exit cleanly...

I wonder if this is fixed in newer drivers. Apparently, the latest driver for Audigy Rx is from September 2026, fixes some security issue.

Did some testing of Audigy Rx on 32-bit Windows XP using drivers from Daniel K.'s Audigy Pack 8.1.

The behavior of settings related to speaker configuration (Windows and Creative) and CMSS appear to be the same as on later systems.

EAX dies and remains unusable until reboot if you launch any OpenAL game, let it initialize the sound and then kill its process. But the same doesn't happen with DirectSound games.

So something is broken when it comes to interaction between OpenAL and the driver.

And for some reason, on 64-bit OS (at least Vista+, didn't check x64 XP), driver installation doesn't install 64-bit ct_oal.dll, but the file is there in the package. That applies to the driver in Daniel K.'s pack and after inspecting the installer of the latest driver, it applies to it as well.

But if you copy the file manually to C:\Windows\System32, then Creative's OpenAL seems to work for 64-bit programs too. I only quickly tried latest ioquake3. Though you have to change the setting in config file referencing OpenAL library, it's set to OpenAL64.dll, while the 64-bit router library (which enumerates *_oal.dll files) is still called OpenAL32.dll.

I admit I don't really understand what's being discussed above but I see Audigy Rx mentioned so I thought I'd add my experience with it on Win10. The Rx seemed to work fine for the most part and sounded great but when using ALchemy with FarCry 2 I got crashes, where it worked perfectly with my X-Fi Titanium, so I remove the Rx and haven't used it since. I consider it a nice backup card for now.

UCyborg wrote on Today, 01:52:

And for some reason, on 64-bit OS (at least Vista+, didn't check x64 XP), driver installation doesn't install 64-bit ct_oal.dll, but the file is there in the package. That applies to the driver in Daniel K.'s pack and after inspecting the installer of the latest driver, it applies to it as well.

I don't have an Audigy Rx, but according to this post, it never officially supported ALchemy. And while DanielK's drivers and some other unofficial packs can make it work, it's unlikely that Creative did any QA testing with ALchemy on that card. So it's unsurprising that certain games aren't working properly with it.

On that note, I'm not entirely sure why people keep bringing up the Audigy Rx as some modern solution to EAX, when they could buy a used X-Fi Titanium for less money, and get a much better retro gaming experience.

64-bit ct_oal.dll not being installed has nothing to do with Audigy Rx or Alchemy. It's not an Alchemy file at all, but part of OpenAL. And since Audigy Rx has an official support for the EAX effects, you'd think, the drivers would install it, haha.

I'm getting exactly the same behavior with X-Fi PCI/PCIe cards, including Titanium/HD and Onkyo Wavio. That's just a problem with the installer script. I remember Daniel_K some time ago documenting the issue and it was widely known on the creative's forums, but nothing was ever done about it, AFAIK. You simply drop the proper file into the System32 folder yourself.

I don't think, Creative ever had QA-testing.

X-Fi Titanium has an unfortunate glitch with OpenAL on modern systems where the sound streams aren't closed properly, which leads to the long looping sounds becoming dropped out when the max voices limit is reached. Same goes for the shorter but more frequent samples, like footsteps, horse canter or gun burst. In fact, those kind of sounds is what's exactly triggering the issue. So, silence instead of ambient environment sound or music missing, uneven footsteps, missing gunshots.

This is most noticeable in native OAL games as they tend to use the channels way more aggressively. Lord of the Rings Online was the most egregious example (though they've switched to OpenAL Soft recently), but other games, including dsound ones through Alchemy suffered as well.

This applies only to Titanium SKUs and there's a workaround for that by Bibendovsky guy (who also worked on the implementing EAX emulation into OpenOAL Soft), but it's reportedly not perfect and doesn't fix all the problems.

Which is indeed very unfortunate because that's still a cheap and readily available card, has PCIe bus, full EAX features support, 7.1 analog output, and requires no activation of DDL/DTS Connect encoders. In other words, everything one can desire from the hardware accelerated legacy Creative soundcard.

Yet, I would agree that Rx is not the perfect alternative. For one, it doesn't even support the full feature set (even though not a lot of games used it, but still).

The main alternatives are PCI X-Fi through PCI-PCIe bridge, which can be flakey and you'll have to hunt for the encoder activation crack since the activation services went down, or Titanium HD, which doesn't have analog surround output, though encoders are available out of the box, or Onkyo Wavio SE300PCIE, which was Japan only, hard to find and tends to be pretty expensive. However, the last one has the 7.1 output and, I think, has encoders activated.

Zoomer wrote on Today, 08:08:

X-Fi Titanium has an unfortunate glitch with OpenAL on modern systems where the sound streams aren't closed properly, which leads to the long looping sounds becoming dropped out when the max voices limit is reached. Same goes for the shorter but more frequent samples, like footsteps, horse canter or gun burst. In fact, those kind of sounds is what's exactly triggering the issue. So, silence instead of ambient environment sound or music missing, uneven footsteps, missing gunshots.

DanielK's final X-Fi support pack claims to have fixed the OpenAL issues. While I haven't tested that extensively, I can at least say that Doom 3, Quake 4 and Battlefield 2 sounded fine to me under Win10 when using those drivers.

On a retro rig running WinXP, this can be circumvented by simply using an older driver version. The last good one without the OpenAL bug seems to be 2.17.0008. Of course, anything older than that should work fine too.

No, it didn't and I don't see where Daniel claims it did. There was a crash related to SRC and OpenAL, but that's totally different.

You probably haven't noticed it. But when you do, it'll become so annoying to the point of being unplayable. Specifically, I remember Quake 4 sometimes dropping ambient sounds during an intense fight. You won't notice that until the fight is over, but after that you'll hear dead silence and footsteps, voices, other effects, but nothing else. It's very characteristic.

I'll leave the link to the fix just in case (I haven't tested it myself): https://github.com/bibendovsky/xfitssfix/discussions/1

Here the bug is discussed in more detail: https://github.com/bibendovsky/eaxefx/issues/ … ment-1161557119

On retro rig it doesn't matter at all since you can use another card.

2.17.0008 is a bad idea due to potential BSODs (depending on your config and X-Ram presence).

I wouldn't recommend Titanium at all in its current state.

Sure you may be less susceptible to the bug, or you won't have x-ram, or your have some exotic config that won't bsod under Windows XP and 2.17.0008, but honestly it's like walking on a minefield.

Zoomer wrote on Today, 09:40:

No, it didn't and I don't see where Daniel claims it did. There was a crash related to SRC and OpenAL, but that's totally different.

In the release notes for his X-Fi Support Pack 8.0 (Refresh 3) he states that "OpenAL works fine, tested on many games (Bioshock, Killing Floor)." This is right below the SRC issue, so I'm unsure if it's related to that or not.

Zoomer wrote on Today, 09:40:

You probably haven't noticed it. But when you do, it'll become so annoying to the point of being unplayable. Specifically, I remember Quake 4 sometimes dropping ambient sounds during an intense fight. You won't notice that until the fight is over, but after that you'll hear dead silence and footsteps, voices, other effects, but nothing else. It's very characteristic.

I have experienced that behavior with OldUnreal and its OpenAL sound renderer a few years back. In fact, that's how I first became aware of this bug. It's well documented on their forums, and they suggested using an older driver version as a workaround. So I know what you mean, I just haven't personally seen it on my Win10 system so far. But again, I haven't done any extensive EAX gaming there.

Zoomer wrote on Today, 09:40:

2.17.0008 is a bad idea due to potential BSODs (depending on your config and X-Ram presence).

Source on that? I've used this driver version (and a slightly older one) on two of my WinXP systems for years, and haven't had any BSOD issues. This is with an X-Fi Titanium Fatal1ty Pro (SB0886) which has 64MB X-RAM.

"works fine" is related to the post Win10 build 1903 problems. The changes made by Microsoft in the audio stack rendered the X-Fi products inoperable. The card was detected, kinda working, but no sound was being output, no audio settings were available. OpenAL complained that no hardware or software sinks were detected - or something along these lines. Latest builds fixed that, hence that "works fine" line.

Yeah, I've read the old unreal forum and it's exactly that, and they're also referring to my old post on the LOTRO forum. I even should have this video somewhere, which is referenced in the topic title, haha. And this is not simply an older version - that's the whole other branch where they have changed the whole hardware acceleration part of the driver. And they did that only for Titanium and Titanium HD. And that's when the bug was introduced - but only on Titanium.

As to the source for Windows XP, well, uuh, you see, the source is me 😜 Well, I mean, it's difficult to find anything now with that new fancy Google search, and X-Fi Titanium + Windows XP wasn't the most popular configuration. However, I was very active EAX proponent during those days, filed lots of reports (going often under the name of Jeph) and tested lots of system configurations. From what I can remember, this whole Titanium BSOD story was due to the way they implemented X-Ram on the PCIe SKUs, and that the older way of addressing it (the way it was done during the PCI era SKUs) was causing the system crash. Hence the new driver branch and the aforementioned bug.

Yes, Windows Vista/7 was more at risk (more like, the BSOD was guaranteed) and in Windows XP it only happened under very specific circumstances (the game must had been using the native hardware OAL implementation) - and even then it had to be a very demanding OpenAL load (oh, and also either 2+ core CPU or HT CPU). And as you know, even the games that have such an implementation, often fall back to DS3D under Windows XP - again, D&D Online/LOTRO engine is the prime example - hence little reports of those. If you're still up for sources, try looking into Unreal 2 + titanium + BSOD + Windows XP, as this title was using way more sound sources than u1 and ut, had native implementation and also defaulted to it (I think).

Again, all this is esoteric because you don't need Titanium in a retro rig. For one, it would definitely have a PCI slot (and most likely won't have any PCIe).

For another, when you go for OpenAL games, you'll most likely jump into the Windows Vista/7 era or very close to it. When you go there, there's no reason to run these dx9/dx10 games on a retro rig. They're perfectly serviced by a more modern systems + dgVoodoo (and 4k resolutions, 120fps+ modes, etc, or a CRT monitor). At least that's how I do it.

The only remaining stone is EAX. And for that you'll need a real creative product that's still working on modern systems. And that's not Titanium, or at least not for now - is what I'm on about 😀

Edit: and I mean, that's great if a Titanium already works in your retro build. The crash is very rare and very specific - and most of the times you can simply switch to DS3D and often maintain EAX compatibility. So it doesn't really matter. It's just, knowing all that and keeping in mind the PCI slot availability in a retro rig, it's difficult to recommend this specific card to someone, who's going for a new retro build. The PCI variants also have way more fancier options with way more functional front bay panels, remote controls and even a cool breakout box. It was the Creative's heyday.

Zoomer wrote on Today, 11:47:

"works fine" is related to the post Win10 build 1903 problems. The changes made by Microsoft in the audio stack rendered the X-Fi products inoperable. The card was detected, kinda working, but no sound was being output, no audio settings were available. OpenAL complained that no hardware or software sinks were detected - or something along these lines. Latest builds fixed that, hence that "works fine" line.

I see. That makes sense then, and it's unrelated to the OpenAL bug with the disappearing sounds. Thanks for the clarification! Also, there was another driver mod that attempted to tackle this, but I'm not sure if it's still being actively developed.

Zoomer wrote on Today, 11:47:

As to the source for Windows XP, well, uuh, you see, the source is me 😜 Well, I mean, it's difficult to find anything now with that new fancy Google search, and X-Fi Titanium + Windows XP wasn't the most popular configuration. However, I was very active EAX proponent during those days, filed lots of reports (going often under the name of Jeph) and tested lots of system configurations. From what I can remember, this whole Titanium BSOD story was due to the way they implemented X-Ram on the PCIe SKUs, and that the older way of addressing it (the way it was done during the PCI era SKUs) was causing the system crash. Hence the new driver branch and the aforementioned bug.

For what it's worth, I've been using the aforementioned X-Fi Titanium Fatal1ty on an i5 3570K under Windows XP for nearly 5 years with no BSOD issues. I did a full playthrough of several native OpenAL games like Quake 4 and Prey on that system, and they worked just fine. I've also used a regular X-Fi Titanium (SB0880) on a Core 2 Duo E8600 rig for some time, with no problems either. So while I don't doubt the issue that you mentioned exists, it didn't affect either of my WinXP systems. This is with the 2.17.0008 driver.

Joseph_Joestar wrote on Today, 12:12:

Also, there was another driver mod that attempted to tackle this, but I'm not sure if it's still being actively developed.

Yeah, I saw this and it reminded me of the wild west times where there were multiple of those mods. The "PAX" something or other mod was especially adorable with the change log including lines like "IMPROVED bASs!!11", or "sound is now more CLEAR!!". I'm exaggerating of course (but just a little), but these were the times 😀

For starters, I don't even think (correct me, if I'm wrong) that the Creative driver is using any APO at all. Maybe only in the later drivers of x5 era for DDL on the fly encoding, or something?

Other improvements aren't less perplexing. I mean, we're in the THD enhancements territory already 😁

And again, I managed to get to the BSOD in Windows XP only by stress testing. I remember it being BF2 with lots of spawned bots, then the same in UT2004, and after the long playing session in LOTRO. Why I keep bringing up LOTRO is because the game is kinda janky in the sound department. It did the steps and hooves samples in such a way that the BSOD was almost instantaneous (some advanced special sound effects, I tell ya). However under Windows XP the game could have been running for 30 minutes or even more, before jumping to a BSOD.

All of this was only on Titanium. Back then I was able to compare it to X-Fi gamer and Audigy 2 ZS.

We then even got the special driver version with an experimental "disable XRAM" registry key from the Creative staff on the forums. You can find the reference to this key on the web, but I doubt it works anywhere but this very specific beta package. And I was the only one who tested it in Windows XP and reported success. Then the new branch is what followed a couple of months later.

Zoomer wrote on Today, 13:02:

Yeah, I saw this and it reminded me of the wild west times where there were multiple of those mods. The "PAX" something or other mod was especially adorable with the change log including lines like "IMPROVED bASs!!11", or "sound is now more CLEAR!!". I'm exaggerating of course (but just a little), but these were the times 😀

Heh, I vaguely remember reading about those. Fun times indeed! 😁 For the record, I never tried the modded driver that I linked to earlier. That's because I almost exclusively play EAX games under WinXP, and that driver doesn't work there. So I can't say how good/bad it might be.

Zoomer wrote on Today, 13:02:

And again, I managed to get to the BSOD in Windows XP only by stress testing. I remember it being BF2 with lots of spawned bots, then the same in UT2004, and after the long playing session in LOTRO. Why I keep bringing up LOTRO is because the game is kinda janky in the sound department. It did the steps and hooves samples in such a way that the BSOD was almost instantaneous (some advanced special sound effects, I tell ya). However under Windows XP the game could have been running for 30 minutes or even more, before jumping to a BSOD.

From my testing, Battlefield 2/2142 can almost reach the 128 hardware voice limit of X-Fi cards during heavy firefights. It's similar with Quake 4 and Prey, though not quite to that extent.

To clarify, those tests were conducted under Win10 using DanielK's drivers. But back when I was actually playing through those games (from start to finish) it was under WinXP on the i5 3570K system with the 2.17.0008 driver. I've likely encountered similarly resource heavy situations during my playthrough, but I had no BSOD issues. Also, I played a fair bit of UT2004 on that setup (with the EAX 5.0 patch) and it was fine too.