VOGONS


PCI-E soundcards with FM chip?

Topic actions

First post, by Duffman

User metadata
Rank Member
Rank
Member

I've purchased a PCI-E version of the CMI8738 for it's FM chip, but I know about it's poor DOS compatibility
I'm just wondering is there any other PCI-E soundcard out there with better compatibility?, a PCI-E version of yamaha XG?

MB: ASRock B550 Steel Legend
CPU: Ryzen 9 5950X
RAM: Corsair 64GB Kit (4x16GB) DDR4 Veng LPX C18 4000MHz
SSDs: 2x Crucial MX500 1TB SATA + 1x Samsung 980 (non-pro) 1TB NVMe SSD
OSs: Win 11 Pro (NVMe) + WinXP Pro SP3 (SATA)
GPU: RTX2070 (11) GT730 (XP)

Reply 1 of 25, by Jo22

User metadata
Rank l33t++
Rank
l33t++

Cool, never heard of these PCIe versions before! Thanks for this thread! 😀

By the way, does your card come with any drivers for Win9x or XP ?

Just asking, because chances are good that if these older OSes are supported,
the card is perhaps able to map Gameport and FM in an ISA compatible address space.

I think I heard that this is a common issue today.
Newer PCI/PCIe bridges don't support these memory regions anymore.

Perhaps this is also caused by the drop of ISA PnP since Windows Vista ?
Only a few core devices are still supported (floppy, serial, parallel) today.

So support for ISA replacements like Super I/O and Low-Pin-Count (LPC) is in the decline, too.
The only reason for ISA to still to exists is because of the BIOS, which is also slowly replaced by UEFI.
And because of special applications, of course (servers, industry, research). But I heard that their ISA slots are not
DMA-capable and essentially a modified version of LPC.

Oh well, I'm afraid I'm talking too much again.. 😅

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 2 of 25, by nforce4max

User metadata
Rank l33t
Rank
l33t

You are lucky that you even found one that had some level of FM functionality but this is why so many of us just stick to ISA cards.

On a far away planet reading your posts in the year 10,191.

Reply 4 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
mr_bigmouth_502 wrote:

There are PCI-E soundcards with FM support? 😳 Mind. Blown.

Same here!

I don't even understand how it could be used...

Can we get pictures of one of these cards, or a model number at least?

Now for some blitting from the back buffer.

Reply 5 of 25, by Super_Relay

User metadata
Rank Newbie
Rank
Newbie

this one came up in a google, it claims sb16/sbpro and xp/2000 drivers

http://www.newegg.com/Product/Product.aspx?It … =9SIA80M2VC3229

"- Software compatibility: compatible with Sound Blaster, Sound Blaster Pro and DirectSound.
Operating Systems: Windows 2000, XP, Vista, Windows 7, Windows 8, Linux."

that is interesting for sure

Reply 6 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Super_Relay wrote:
this one came up in a google, it claims sb16/sbpro and xp/2000 drivers […]
Show full quote

this one came up in a google, it claims sb16/sbpro and xp/2000 drivers

http://www.newegg.com/Product/Product.aspx?It … =9SIA80M2VC3229

"- Software compatibility: compatible with Sound Blaster, Sound Blaster Pro and DirectSound.
Operating Systems: Windows 2000, XP, Vista, Windows 7, Windows 8, Linux."

that is interesting for sure

Yeah, I don't know much about the intricacies of the PCI-E, PCI and ISA BUSes or how games work with those... but I'd lean toward that information being incorrect. Soundblaster\Pro compatibility would imply that you could install that into a PCI-E slot, install DOS and use the card with a setblaster command of some kind and no intervention from Windows.

Now for some blitting from the back buffer.

Reply 7 of 25, by Jo22

User metadata
Rank l33t++
Rank
l33t++

Found something! 😁

http://www.cmedia.com.tw/support/download_center

http://www.asmedia.com.tw/eng/e_show_products.php?item=114
(@Super_Relay Thanks for the newegg link - one pic had a close up view of that chip!)

So the audio chip really is SB Pro compatible.

Perhapss drivers for the PCI version may work, too ?
I can't check them myself, but here's a list of the filenames of the older drivers..

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 8 of 25, by Scali

User metadata
Rank l33t
Rank
l33t
Ozzuneoj wrote:

Yeah, I don't know much about the intricacies of the PCI-E, PCI and ISA BUSes or how games work with those... but I'd lean toward that information being incorrect. Soundblaster\Pro compatibility would imply that you could install that into a PCI-E slot, install DOS and use the card with a setblaster command of some kind and no intervention from Windows.

I don't see why it couldn't work. Modern video cards do pretty much the same: a PCI-e card which is register-compatible with VGA in DOS. So I suppose you could also make a PCI-e card that is register-compatible with an SB card.
Perhaps the reason why this is not normally done is because it requires more complex PCI-e plug&play firmware or whatnot.
VGA cards have always needed complex firmware anyway, even back in the ISA days. Sound cards didn't have any firmware at all.

http://scalibq.wordpress.com/just-keeping-it- … ro-programming/

Reply 9 of 25, by Kamerat

User metadata
Rank Oldbie
Rank
Oldbie

Ace tried to get FM working on his/her PCIe CMI8738-LX without much success: I've got the OPL3 synth from a CMI8738 working in Windows 7 x64

DOS Sound Blaster compatibility: PCI sound cards vs. PCI chipsets
YouTube channel

Reply 10 of 25, by Jo22

User metadata
Rank l33t++
Rank
l33t++
Kamerat wrote:

Ace tried to get FM working on his/her PCIe CMI8738-LX without much success: I've got the OPL3 synth from a CMI8738 working in Windows 7 x64

I've just read it, thanks for sharing that link! 😀
Perhaps it is possible to get it somehow working in DOS or 9x, though (hopefully).

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 11 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Scali wrote:
I don't see why it couldn't work. Modern video cards do pretty much the same: a PCI-e card which is register-compatible with VGA […]
Show full quote
Ozzuneoj wrote:

Yeah, I don't know much about the intricacies of the PCI-E, PCI and ISA BUSes or how games work with those... but I'd lean toward that information being incorrect. Soundblaster\Pro compatibility would imply that you could install that into a PCI-E slot, install DOS and use the card with a setblaster command of some kind and no intervention from Windows.

I don't see why it couldn't work. Modern video cards do pretty much the same: a PCI-e card which is register-compatible with VGA in DOS. So I suppose you could also make a PCI-e card that is register-compatible with an SB card.
Perhaps the reason why this is not normally done is because it requires more complex PCI-e plug&play firmware or whatnot.
VGA cards have always needed complex firmware anyway, even back in the ISA days. Sound cards didn't have any firmware at all.

The difference I think is that VGA compatibility has been standard for decades, where as SB DOS compatibility on motherboards died off several years ago as far as I know. If anything you'd have to have a motherboard that allowed a PCI-E card to have direct access to all of the resources required for DOS sound compatibility. I don't think this is standard.

Even the Yamaha PCI cards everyone talks about using are only useful for DOS if the motherboard allows it to be. I could be completely off base but I don't think you can drop one of those in a brand new motherboard with a PCI slot and expect it to work in DOS.

Now for some blitting from the back buffer.

Reply 12 of 25, by Jo22

User metadata
Rank l33t++
Rank
l33t++

I guess I'm just a layman when it comes to these things, but perhaps we can't make a generalization here.
A few years ago I tried to use a PCI ViRGE in a more recent PC with PCI-E slots, but it didn't work.
The card itself was fine and worked in an older PC, but the BIOS didn't accept it as a primary VGA card.
So the issue must have been somehow related to the BIOS (PnP issue ?) or the mapping didn't work for some reason (faulty PCI bridge ?)

Speaking of VGA cards, modern cards aren't any better in this regard. Their VGA support is abysmal and can be described as synthetic at best.
Comparable to VMWare's implementation, it is only intended as an interim solution until native drivers are loaded.
I know, some people will now disagree for sure: The VGA registers are still there, and so on. Perhaps they're kinda right,
but I don't think more of the sophisticated things like early-to-mid 90's demos will still work the way as intended.
And any bare metal stuff will likely fail or be dog slow. And VESA support ? Better not talk about it.
VBE 3.0 deprecated certain things VBE 1.2 or VBE 2.x aware programs expected. And VGA fonts are another thing to consider.
The 8x14 font is likely missing to make the VBEBIOS fit in memory and perhaps can't be restored because
there's no utility for it (S3 cards got a fix utility for this long ago).

And in a few years, VGA support on the graphics card will be perhaps removed entirely because UEFI doesn't need it.
It uses UGA (Universal Graphic Adapter) or GOP (Graphics Output Protocol) and doesn't even care for text mode anymore.
So if we're lucky, its Compatibility Support Module (CSM) will still be there and include some kind of VGA/VBE emulation.
The emulation will likely only cover high-level support if that's enough for the installation routines found in
Windows 7/XP to work (Windows 8 already supported GOP).

But still, I have no complaints to speak of.
Considering the age of DOS or XP and its software, things work still astoundingly well. 😀
Perhaps we can even fix things in the future. Adlib and SVGA replicas have already been made
and XT-IDE kind of backported a newer interface to an older platform.
Perhaps someone will do similar things for Sound Blaster by using a microcontroller and some bridge chips.
If not, we could also aim for WSS, GUS or PAS compatibility. Or something else which doesn't rely so much on ISA-DMA.
But FM support is something I'm worried about the least. Re-mapping/re-directing port addresses already works, just think of SoftMPU.
Perhaps we could even use external PCI or ISA adapters, driven by Arduino or similar devices.
I think someone already did this with a SB 2.0 or Adlib card. Wait, here's the link:

-= AdLib sound card on a parallel port =-
http://www.raphnet.net/electronique/adlib/adlib_en.php

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 13 of 25, by Scali

User metadata
Rank l33t
Rank
l33t
Ozzuneoj wrote:

The difference I think is that VGA compatibility has been standard for decades, where as SB DOS compatibility on motherboards died off several years ago as far as I know. If anything you'd have to have a motherboard that allowed a PCI-E card to have direct access to all of the resources required for DOS sound compatibility. I don't think this is standard.

Well, that sounds weird.
My point is that as far as I know, PCI and PCI-e should be able to act completely transparently to the system. As in, it shouldn't matter to the software whether a card is ISA, VLB, PCI or PCI-e. The card just presents resources to the system.
So I doubt that PCI-e is designed to not be backwards compatible, but that they hacked in some exceptions to make it work specifically for VGA. I think it's far more logical that it's a generic solution.
Of course, there could always be buggy chipsets, but that's another story.

http://scalibq.wordpress.com/just-keeping-it- … ro-programming/

Reply 14 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Scali wrote:
Well, that sounds weird. My point is that as far as I know, PCI and PCI-e should be able to act completely transparently to the […]
Show full quote
Ozzuneoj wrote:

The difference I think is that VGA compatibility has been standard for decades, where as SB DOS compatibility on motherboards died off several years ago as far as I know. If anything you'd have to have a motherboard that allowed a PCI-E card to have direct access to all of the resources required for DOS sound compatibility. I don't think this is standard.

Well, that sounds weird.
My point is that as far as I know, PCI and PCI-e should be able to act completely transparently to the system. As in, it shouldn't matter to the software whether a card is ISA, VLB, PCI or PCI-e. The card just presents resources to the system.
So I doubt that PCI-e is designed to not be backwards compatible, but that they hacked in some exceptions to make it work specifically for VGA. I think it's far more logical that it's a generic solution.
Of course, there could always be buggy chipsets, but that's another story.

*shrug*

Computers are weird. Especially when we're talking about compatibility with 25 year old standards that have been mostly replaced since Windows 98 and DirectSound. VGA hasn't changed much and is still included on most modern PCs and displays, so it'd be quite detrimental to sacrifice VGA support for something else, where as DOS sound compatibility is likely near the bottom of the list of things that hardware designers are concerned about these days.

PCI and PCI-E are not the same interface and aren't treated the same by the chipset... similar to how ISA and PCI are treated totally differently. There may be some very basic level of compatibility for common legacy devices, but for most things a PCI to PCI-E bridge chip is needed between any legacy PCI devices and PCI-E interfaces. Either you'll have a sound card with a bridge chip on it, or in the case of newer boards that do not have any built in PCI support on the chipset, you'll have it directly on the motherboard board (connected to the PCI slots). Either way, the old-school sound chip that can handle SBPro and FM music has to run through one of these chips (the card at the newegg link above has an asmedia PCI to PCI-E bridge chip on it). So what you need is for the PCI-Express interface to support legacy sound specifications (like sending FM music to port 388h or 220h)... which seems possible, but there isn't much on this... on top of that you'll need either a bridge chip that is built to also allow access to these and other ports, or a modern PCI-E sound card that is built to access them directly (doesn't seem to exist, since even the card this thread is about requires a bridge chip).

If someone is brave, another thing to try would be one of these PCI to PCI-E adapters (with a bridge chip):
http://www.newegg.com/Product/Product.aspx?It … N82E16815158165

Or to simply try using a PCI sound card in a board that has PCI slots but no native PCI support (Intel Ivy Bridge and newer architectures require PCI-E bridge chips for PCI support).

And try an old sound card in it that functions in DOS (or even 9x with DOS compatibility). I have my doubts... simply because features are cut all the time. If that doesn't even work, then I highly doubt that a modern C-Media chip will.

As was posted earlier, others have tried with the C-Media chips and the results are spotty:
I've got the OPL3 synth from a CMI8738 working in Windows 7 x64

One user posted that they had a PCI version of a similar card working with OPL3 (FM) music on a modern chipset (Z77) that would implement a bridge chip, so it may be possible with some bridge chips... but this is only one positive result at this point, and its only one bridge chip on one motherboard. If the sound cards themselves are never paired with a bridge chip that allows this functionality, then we won't have any PCI-E cards with real, functional SB compatibility, even if the sound chip is capable of it.

Now for some blitting from the back buffer.

Reply 15 of 25, by Scali

User metadata
Rank l33t
Rank
l33t
Ozzuneoj wrote:

Computers are weird. Especially when we're talking about compatibility with 25 year old standards that have been mostly replaced since Windows 98 and DirectSound.

Compatibility is the only redeeming value of the PC platform.

Ozzuneoj wrote:

VGA hasn't changed much

Are you kidding me?
Modern GPUs have absolutely nothing to do with the original VGA standard. VGA has been 'emulated' on modern hardware since the mid-90s at least (and VGA itself obviously emulates CGA and EGA in hardware, since CGA, EGA and VGA work quite differently internally).
Point remains: the compatibility is there, so apparently it is possible.
I think the only reason why it's not being done for sound chips is that it's not worth the cost. VGA compatibility is required during the 'bootstrap' phase of your system, and during OS installation. You have to start somewhere, before you can install drivers for your actual OS and hardware.
But your computer will boot fine without a sound card, so there's no need for it to function before the OS has loaded its drivers.
So they can shave off a few cents from the price of every sound chip by not including any SB-compatibility.

Ozzuneoj wrote:

PCI and PCI-E are not the same interface and aren't treated the same by the chipset... similar to how ISA and PCI are treated totally differently.

Obviously not, but that's not what I said.
What I said is that the hardware can act transparently to the rest of the system.
Eg, regardless of whether your video card is ISA, PCI, PCI-e or whatever other interface, the CPU can just access certain memory areas for the video ram, and certain IO ports for the hardware registers. So you don't need to write specific code for ISA, PCI, PCI-e or whatever. You write it for 'VGA'. You write it once, and it runs everywhere.
The exact same could work for sound cards, if they would expose the same IO ports.
I think the main problem here is that SB-compatibility went out of style decades ago. Even the Gravis UltraSound didn't have any SB-compatibility in hardware. Later SB cards didn't really have it either, some only when booting into a DOS session from Windows.

Bottom line is: I don't see why *in theory* you couldn't take the basic components from an ISA Sound Blaster card, put them onto a PCI-e card, add the required glue logic to make sure the PCI-e plug&play initializes everything properly so that the IO-ports are mapped, and the IRQ and DMA channels are set up, and end up with a card that works exactly the same as the original ISA card would.

The only thing that is fundamentally different with modern cards is that they let the plug&play decide the resources for them, and the driver then uses those, where an SB was just jumpered to fixed addresses, and most software was hardcoded to these addresses.
But since VGA is hardcoded to addresses as well, apparently PCI-e still supports this. So it should also be possible for a sound card.

http://scalibq.wordpress.com/just-keeping-it- … ro-programming/

Reply 16 of 25, by Jo22

User metadata
Rank l33t++
Rank
l33t++
Ozzuneoj wrote:

As was posted earlier, others have tried with the C-Media chips and the results are spotty:
I've got the OPL3 synth from a CMI8738 working in Windows 7 x64

I should probably keep my mouth closed now, but that user did not try to use the card on DOS or Win9x it seems.
He/she tried to pass through OPL commands to the CMI chip directly. From Windows x64, I think.
I tried something similar with a TT-Solo and DOSBox, but had no success so far. So perhaps something is
masking these ports.. Btw, we had another discussion about soundcards just recently Pure DOS gaming system with 100% digital audio output

I'm also a bit shocked about the fact that people have such huge issues with these CMI chips in general.
I mean, the audio chip's whole legacy block was designed by intelligent engineers.
And if you look at the data sheet, you'll notice how much effort they spend on it.
It has a huge number of control registers just for this darn thing.
So it is really hard to believe it is as bad as people make it.
In fact, setting-up such a thing in DOS was easier for me than
installing an original PnP card from Creative.

Edit: Please calm down, people. The whole problem is probably just caused by something minor we don't know about yet.

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 17 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Scali wrote:
Compatibility is the only redeeming value of the PC platform. […]
Show full quote
Ozzuneoj wrote:

Computers are weird. Especially when we're talking about compatibility with 25 year old standards that have been mostly replaced since Windows 98 and DirectSound.

Compatibility is the only redeeming value of the PC platform.

Ozzuneoj wrote:

VGA hasn't changed much

Are you kidding me?
Modern GPUs have absolutely nothing to do with the original VGA standard. VGA has been 'emulated' on modern hardware since the mid-90s at least (and VGA itself obviously emulates CGA and EGA in hardware, since CGA, EGA and VGA work quite differently internally).
Point remains: the compatibility is there, so apparently it is possible.
I think the only reason why it's not being done for sound chips is that it's not worth the cost. VGA compatibility is required during the 'bootstrap' phase of your system, and during OS installation. You have to start somewhere, before you can install drivers for your actual OS and hardware.
But your computer will boot fine without a sound card, so there's no need for it to function before the OS has loaded its drivers.
So they can shave off a few cents from the price of every sound chip by not including any SB-compatibility.

Ozzuneoj wrote:

PCI and PCI-E are not the same interface and aren't treated the same by the chipset... similar to how ISA and PCI are treated totally differently.

Obviously not, but that's not what I said.
What I said is that the hardware can act transparently to the rest of the system.
Eg, regardless of whether your video card is ISA, PCI, PCI-e or whatever other interface, the CPU can just access certain memory areas for the video ram, and certain IO ports for the hardware registers. So you don't need to write specific code for ISA, PCI, PCI-e or whatever. You write it for 'VGA'. You write it once, and it runs everywhere.
The exact same could work for sound cards, if they would expose the same IO ports.
I think the main problem here is that SB-compatibility went out of style decades ago. Even the Gravis UltraSound didn't have any SB-compatibility in hardware. Later SB cards didn't really have it either, some only when booting into a DOS session from Windows.

Bottom line is: I don't see why *in theory* you couldn't take the basic components from an ISA Sound Blaster card, put them onto a PCI-e card, add the required glue logic to make sure the PCI-e plug&play initializes everything properly so that the IO-ports are mapped, and the IRQ and DMA channels are set up, and end up with a card that works exactly the same as the original ISA card would.

The only thing that is fundamentally different with modern cards is that they let the plug&play decide the resources for them, and the driver then uses those, where an SB was just jumpered to fixed addresses, and most software was hardcoded to these addresses.
But since VGA is hardcoded to addresses as well, apparently PCI-e still supports this. So it should also be possible for a sound card.

I don't see what we're disagreeing on.

Theoretical hardware with perfect ISA backward compatibility is great, but as you said, due to cost cutting (and no market for it), anything even close to that is highly unlikely to ever exist.

Also, my comment about VGA is obviously directed at the fact that VGA devices generally work together and the interface itself is still common. I don't know anything about VGA being emulated, and this isn't really the place for a discussion on that subject. I was merely pointing out that comparing a 30 year old standard that is still widely supported (emulated or not) to one that hasn't had main stream support in 15 years isn't going to shine new light on the subject. We both acknowledge that it is up to the people that make the hardware to determine if this kind of thing is possible, and in this situation (legacy sound card on modern interface) you have to have a sound chip that was designed with this compatibility, a bridge chip that was designed to allow it and a motherboard that doesn't prevent all of that from working smoothly. All for basically no return on their part... because how many people buy modern hardware to run DOS games outside of DOSBox?

We both know this, and as others have shown in other threads, it could work with the perfect setup, but there are so many factors against it that it is unlikely to be easy to get working. Anyone who wants to pursue it is more than welcome of course, and I'd be interested to see the result. I just have a feeling its going to end with "there are three motherboards on the planet that can do this with one specific sound card and it is nowhere near as good as just using DOSBox or a dedicated older PC".

Now for some blitting from the back buffer.

Reply 18 of 25, by Scali

User metadata
Rank l33t
Rank
l33t
Ozzuneoj wrote:

Also, my comment about VGA is obviously directed at the fact that VGA devices generally work together and the interface itself is still common. I don't know anything about VGA being emulated, and this isn't really the place for a discussion on that subject.

I think it is relevant, since video cards have changed tremendously over the years, and apparently this has not prevented them from remaining backward-compatible, in whichever way.
Perhaps the issue with sound cards is that software could (and should?) take care of it.
Eg, the NTVDM in Windows 2000 and newer had its own SB emulation built in, therefore cards didn't require it.

Ozzuneoj wrote:

We both acknowledge that it is up to the people that make the hardware to determine if this kind of thing is possible

No we don't. I say that to the best of my knowledge, it is technically possible to do so. And probably not even very hard... I think a hobbyist might be able to concoct some sort of PCI-e contraption by using a real OPL3 chip and some FPGA or microcontroller to take care of the PCI-e interface.

Obviously it is not economically feasible, but that's another story. I merely objected to the claim that it was somehow technically impossible to drive an OPL3-based card via a PCI-e interface.

http://scalibq.wordpress.com/just-keeping-it- … ro-programming/

Reply 19 of 25, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Scali wrote:
I think it is relevant, since video cards have changed tremendously over the years, and apparently this has not prevented them f […]
Show full quote
Ozzuneoj wrote:

Also, my comment about VGA is obviously directed at the fact that VGA devices generally work together and the interface itself is still common. I don't know anything about VGA being emulated, and this isn't really the place for a discussion on that subject.

I think it is relevant, since video cards have changed tremendously over the years, and apparently this has not prevented them from remaining backward-compatible, in whichever way.
Perhaps the issue with sound cards is that software could (and should?) take care of it.
Eg, the NTVDM in Windows 2000 and newer had its own SB emulation built in, therefore cards didn't require it.

Like I said, the necessity for VGA to remain compatible is the reason that it HAS remained compatible. There is a long history of VGA being used on PCs, which is why I can put an OAK OTIVGA 16bit ISA VGA\EGA card into an 8bit slot in my IBM 5150 and have it display an image on a modern VGA LCD screen with minimal fuss. VGA has only very recently not been included on some PC graphics cards (some no longer have DVI-I). It is a bit of a gamble for mainstream hardware to not include VGA, because both "normal" users (who may still have a VGA monitor or PC they want to use) and enthusiasts (who may have more than one device, one of them being VGA) frequently make use of the interface.

Comparatively few people actually care about Soundblaster DOS compatibility though and it has been that way for over a decade, so there isn't a good source of components or information that would allow companies to create the hardware necessary to do this, and there would be an extremely small market for it.

It is a completely different situation. Its like saying, why don't cars still have cassette decks if they can still give them radios? Hardly anyone cares about cassette decks, where as radios are cheap and incredibly common and would seem crazy to leave out.

Also, software emulation already exists (DOSBox, VDMSound, This), and new projects are popping up all the time. It's a totally different thing than having a 2016 computer fully natively support 1991 features in hardware.

Scali wrote:
Ozzuneoj wrote:

We both acknowledge that it is up to the people that make the hardware to determine if this kind of thing is possible

No we don't. I say that to the best of my knowledge, it is technically possible to do so. And probably not even very hard... I think a hobbyist might be able to concoct some sort of PCI-e contraption by using a real OPL3 chip and some FPGA or microcontroller to take care of the PCI-e interface.

Obviously it is not economically feasible, but that's another story. I merely objected to the claim that it was somehow technically impossible to drive an OPL3-based card via a PCI-e interface.

*shrug*

We're splitting hairs here. I said "it is up to the people that make the hardware to determine if this kind of thing is possible" and you said "I think a hobbyist might be able to concoct some sort of PCI-e contraption by using a real OPL3 chip and some FPGA or microcontroller to take care of the PCI-e interface."

What more can I say? I'm agreeing with you. Someone would need to make it for it to work.

I never said it was technically impossible, but even if the right combination of cards\boards exists currently, it can't possibly work well enough to justify having to look for a specific motherboard and sound card combination to use Windows 9x\DOS on a system with an Ivy Bridge or faster CPU (since anything older has native PCI anyway).

If someone builds a custom card that can do what we want and goes into a PCI-E slot, then that'd be awesome. But we'd still be saddled with trying to run old operating systems and games on hardware that is several dozen times faster and more complex than the hardware it was meant for. Because of this problem, I also doubt that anyone will bother making the hardware to do this, unless someone actually develops and mass produces a genuine old-school gaming PC platform with easy to find components, from the ground up with ISA, PCI, AGP, PCI-E and full DOS compatibility for all of them... in which case, why not just use an ISA card?

Last edited by Ozzuneoj on 2016-08-26, 15:26. Edited 1 time in total.

Now for some blitting from the back buffer.