MaliceX wrote:

EDIT: OK so I have attempted to install the kernel driver on XP where I already have the YMF724F-V legacy mode enabled with PowerYMF. Regardless of methods of installation (Add new hardware, update driver on Legacy Sound System which has port 388h used), I get the following:

An error occurred during the installation of the device.

The class installer has denied the request to install or upgrade the device.

Assistance is required. Thanks.

I got same issue, too. Tried on SoundTrack Dightal XG, DS-XG Legacy WDM Driver still hold address 220 330 and "388". No way to install Kernel Mode Adlib Drivers.

The basic Windows XP Ds1 Driver only install FM MIDI. Can hear FM-MIDI using with Media player, and work on GM setting in DOSBOX.
But Sound horrible because no program reset FM as DOS driver. And installing any DS-XG WDM driver remove FM synth.

Need to figure out how to enable the FM driver on XP since it's present on the .inf file. I've already tried to no avail. 🙁

MaliceX,
I fixed it. See main post.

khokh2001 wrote:

MaliceX,
I fixed it. See main post.

Excellent stuff. Kernel driver under x86 XP confirmed working! Thanks for promptly fixing. (EDIT: one minor issue, it appears when restarting after installation, device cannot start with code 10, but reinstalling driver makes it work again). Also there appears to be no way to get kernel driver working under x64 with code 10 error too, seeing as there is no x64 drivers for the card itself either )

Given that this driver works as-designed, Is there any opportunity for possible additions of extending MIDI CC support? (ie: pitch bend range extension via RPN commands, modulation, expression etc. ?)

Also I'm currently in the process of getting the patches replaced with ones converted from .ROL BNKs. (need to make this more user-accessible like a patch editor or something. Example of this using MidiPlay: here and here)

EDIT: Seems custom patches also work when injecting them directly into the binary DLL. If anyone's interested in fiddling with it say so. (note- not mapped to GM)

Is it possible to setup a patch set for this driver that emulated the superior (IMHO) 4-operator instruments used by the Voyetra SuperSAPI! FM driver?

NJRoadfan wrote:

Is it possible to setup a patch set for this driver that emulated the superior (IMHO) 4-operator instruments used by the Voyetra SuperSAPI! FM driver?

Maybe. Need to reverse-engineer the stored patch format. Do you have a link to the driver?

The stored patches are actually available as MID files (likely sysex). The default 4-operator patch set should be stored in OPL34OP.MID, and it also appears embedded in the driver itself.

http://cd.textfiles.com/audio11000/MSDOS/SNDB … ROAUDIO/PASDSA/

The SAPI!OPL.DRV file is the Windows 3.1x driver. Be warned, its handling of percussion is weird, I have to map channel 10 to channel 16 in the MIDI Mapper to get drums to play correctly as that system.ini setting doesn't seem to work in Windows 3.1.

Well, I managed to get the DOS driver version of the SAPI drivers from the defunct SOUNDSHOCK forums here since I needed a way to both play back MIDI's, log the output, then compare register data from the log to the stored binary operator data to determine whether they're raw OPL3 registers or otherwise.

Hopefully its easy to reverse engineer. Canyon.MID sounds much better with the Voyetra 4-operator "palette" then the Microsoft driver ever did. I used to go out of my way and install Windows 3.1x on machines just to use that driver for FM music!

I informed korea retro game forum about khokh2001's driver.

A volunteer reported, he used Audiotrak's maya 5.1 gold (cmi8738 mx). works on on Asus P5E-VM HDMI, but failed on Asrock z77 Extreme 6. (Both windows 7 x64)
Anyone tried this driver in Intel 7 series M/B? and need working mainboard list for this driver. (I don't know this is chipset problem or UEFI issue.)

Whatever if socram8888's method didn't work, this driver also doesn't work at all.

Feature request.
Ability to configure any port besides 0x388 in kernel-mode driver.
For all those silly one-chip PCI cards.

ik777 wrote:

I informed korea retro game forum about khokh2001's driver. […]
Show full quote

I informed korea retro game forum about khokh2001's driver.

A volunteer reported, he used Audiotrak's maya 5.1 gold (cmi8738 mx). works on on Asus P5E-VM HDMI, but failed on Asrock z77 Extreme 6. (Both windows 7 x64)
Anyone tried this driver in Intel 7 series M/B? and need working mainboard list for this driver. (I don't know this is chipset problem or UEFI issue.)

Whatever if socram8888's method didn't work, this driver also doesn't work at all.

ASUS P8Z77-V PRO w/ Intel i7-3770 4x8GB G.Skill Trident-X, dual-boot running Windows XP SP3 x86, and Windows 7 x64, with a PCI YAMAHA YMF724F-V card. Works fine under XP, not on 7 x64 due to lack of normal drivers. XP kernel driver requires "OPL3 or compatible" driver to be disabled and reenabled to get past Code 10 error on startup.

Progress update on SAPI! patch retrieval, here's a rundown (using SAPI.COM):
- need to work out correspondence of C0, E0 and maybe BD registers
- melodic patches begin at 0x0A
- All patches are 2x2op (no modulation between channel pairs, only additive)
- Instrument register bye order: 20, 40, 60, 80, ??, 23, 43, 63, 83, ?? ??. 20*, 40*, 60*, 80*, ??*, 23*, 43*, 63*, 83*, ??*, ??*, ??*, ??*, ??*, ??* (those denoted * are second port)
- 0x19 (25 byte) instruments
- Percussion is all 2op with a slightly different byte format order (TBD)

Interestingly, the SAPI driver supports pitch bend range extension via RPN/Data Entry commands, as well as a number of other MIDI CC's. It would be nice if the Windows driver could use this.

MaliceX wrote:

Interestingly, the SAPI driver supports pitch bend range extension via RPN/Data Entry commands, as well as a number of other MIDI CC's. It would be nice if the Windows driver could use this.

Neat. Its interesting that Voyetra says its a "4 operator" driver despite the patches not using all 4 at once. They seemed to have pushed the OPL3 to the limit with that driver, its also obvious they designed it with content creation in mind (the DOS drivers were commonly bundled with their sequencer software). None of the emulators I've found replicate its output perfectly (using raw OPL dumps) although DOSBox seems to to come the closest. The Vintage Tone Java OPL3 emulator drops the percussion although the author says this is normal in the release notes. Only one of the emulation engines that comes with the AdPlug Winamp plug-in comes close to working with its output.

Don't get me wrong. It still sounds good and technically they are still 4-operator patches, just that the overall algorithm used is about as good as layering two 2op patches together.

Thank You!

It works in my Windows 7 Ultimate 64-bit.

I just clicked on the install.bat file in the user folder. The device management doesn't show it, but the Coolsoft Virtual MIDI Synth - the program I use to choose Soundfonts and to select the default MIDI output device in Windows 7, can see it.

One odd thing is that, the YMF262, Microsoft GS Synth and the CoolSoft Virtual MIDI Synth are now switched. Meaning that at the default WMP's choice of output, I need to select the Microsoft GS Wavetable Synth to play the OPL3 sounds, and if I select the YMF262 Synth Emulator, it plays my custom Soundfont using the CoolSoft VirtualMIDI Synth instead. But I guess this is due to the Coolsoft Configurator itself. The USB to MIDI adapter works fine, which is connected to my MT-32 currently.

But anyway, it works. Thanks again.

Now I wonder if I can somehow route this YMF262 Emulator to VMWare Player...

Tried on MSI P45 and bought two 8738-mx soundcards. succeed.

My problem was both bad card(don't have fm part, won't sounds in dos or win98) and m/b won't support kernel mode synth.

Is there any possible way for the emulation driver to read the port?

Like for say the rare Windows games that write to it, like Warpath (which is 16-bit btw)

Just thought I'd let people know that I have forked the source code of khokh2001's user mode driver, just to add some improvements to the driver.

http://vgm.mdscene.net/forum/viewtopic.php?t=1167#4266

I'll need to ask Sergey about procuring the kernel mode driver sources at some point, but in the meantime I'll be making my own updates on the user mode driver only.

That forked driver doesn't seem to show up in Windows 7 x64...

Would be nice if there were two more INFs for those synths to be added as separate drivers

leileilol wrote:

That forked driver doesn't seem to show up in Windows 7 x64...

Would be nice if there were two more INFs for those synths to be added as separate drivers

Odd, because I'm using the same OS (You cannot use the INF for installing under x64, and don't forget to rename either DLL to opl3emu.dll) and have no issues using the driver (This WILL overwrite the original one from Sergey). Mind you there may be changes from time to time. Any changes noted on the repository, the binary link will always be updated simultaneously.

I'll look into the options of having alternative names. At the moment I'm looking at options on implementing switchable patch banks, which will need a configuration utility among other things.