if they're both PNP maybe that can cause a headache with the CTCM tool. if one is PNP and another is plain isa then i'd say it's possible.

its definitely worth trying! if it's PNP and you're using win98 or 95 set all the resources there and let CTCM work it out. i'm curious to know if CTCM can initialise 2 cards at once?

and try unisound ! it's great and so easy.

Every known MIDI-related issue with the SB16's shared-MCU architecture is a consequence of concurrent, interrupt-driven digital audio playback. Move that function to another card, and you'll have problem-free MIDI playback, regardless of the card involved. Where the AWE64 resolves the additional single-cycle DMA clicking problem, I'd suggest using its PCM faculties in a dual-card scenario, rather than those of the SB16.

For what it's worth, the Sam & Max issue you describe is a result of Lucasarts' MIDI driver dropping bytes while the SB's MCU is occupied with digital audio playback operations. You can mitigate this problem in a single-card SB16/AWE setup by increasing the size of the game's playback buffer (thereby reducing the number of DMA-transfer-related interrupts).

The SB16 I have allows you to set the IRQ, etc. with jumpers, so it might work out!

(I erroneously described the SB16 as a SBAWE32 in the OP, so I corrected that... I got mixed up!)

Jumpered sb16 will do the trick. But i am not sure about games that use adlib opl port.

Cloudschatze wrote on 2020-05-28, 02:52:

Where the AWE64 resolves the additional single-cycle DMA clicking problem, I'd suggest using its PCM faculties in a dual-card scenario, rather than those of the AWE32.

For what it's worth, the Sam & Max issue you describe is a result of Lucasarts' MIDI driver dropping bytes while the SB's MCU is occupied with digital audio playback operations. You can mitigate this problem in a single-card SB16/AWE setup by increasing the size of the game's playback buffer (thereby reducing the number of DMA-transfer-related interrupts).

Very interesting, thank you for posting this! Re: using PCM from AWE32, would you then suggest using the AWE64 as the "main sound card" (i.e. IRQ 5, I/O 220, DMA 1/5, MPU disabled) and the AWE32 as the "dedicated MIDI port" (i.e. MPU 330, IRQ 2)? I do have a spare PNP AWE32 lying around I could use to test this on my system, though I'm not certain how the Creative PNP utility would handle this.

Re: the Sam & Max issue, is the playback buffer something that can be set somewhere in a config file, or would the game need to be hacked?

Why not just get an SB Pro clone card with decent MPU401 and real OPL3 (chipsets from Aztech, Opti , Yamaha or ESS are good choices; make sure it has a real OPL3, though) to go along with your AWE64 ? You will get the added benefit of a Sound Blaster Pro .

EDIT: And you won't have to play "what if" with the potential consequences of having two SB16 compatible devices, especially if you run Windows too .

I guess the major issue there is getting it here in Japan. It's relatively simple to find AWE32/64 cards here, and you sometimes find SB16s or very rarely other cards with Yamaha OPL3... but I'm concerned with pure DOS compatibility for non-SB cards.

newtmonkey wrote on 2020-05-28, 03:26:

I guess the major issue there is getting it here in Japan. It's relatively simple to find AWE32/64 cards here, and you sometimes find SB16s or very rarely other cards with Yamaha OPL3... but I'm concerned with pure DOS compatibility for non-SB cards.

Well, availability trumps everything else .

As for DOS compatibility, an ISA (important) SB Pro clone with an MPU401 will at least give you a working MPU401 and SB Pro compatibility along with an OPL3 . Creative Labs never managed to make a full bug-free MPU-401, even their own SB16/AWE32/AWE64 is not fully SB Pro compatible (stereo does not work) and, finally, OPL3 compatibility is a given with all the ISA SB Pro clone cards that have one (it's mapped to the proper I/O ports and will work exactly the same as one on a Creative Sb Pro or SB16 would ).

Thanks for the info darry! I'll keep an eye out for ISA SB clones, though they seem quite rare here. I bid on a couple PCI ESS- and Yamaha-based cards just to play around with, they are not expensive (as of right now...)

newtmonkey wrote on 2020-05-28, 03:00:

Re: using PCM from AWE32, would you then suggest using the AWE64 as the "main sound card" (i.e. IRQ 5, I/O 220, DMA 1/5, MPU disabled) and the AWE32 as the "dedicated MIDI port" (i.e. MPU 330, IRQ 2)?

Sure. The AWE64 will additionally allow you to configure its FM at an alternate address, in case that duality is a concern.

Re: the Sam & Max issue, is the playback buffer something that can be set somewhere in a config file, or would the game need to be hacked?

I don't recall if Sam & Max has a startup configuration menu, but you can specify the buffer size as a command-line parameter.

E.g.:

samnmax m s220/5/16 g330
(Mouse, Sound Blaster at 220h, I5, 16KB buffer, General MIDI at 330h)

8KB is the default buffer size. 16KB resolves the hanging issues in one of my SB16 setups, but you can also try 32KB or 64KB otherwise, at the expense of requiring additional conventional memory.

darry wrote on 2020-05-28, 03:39:

Creative Labs never managed to make a full bug-free MPU-401...

In the dual-card scenario that's been described, the SB16's MPU is "bug-free."

Thank you for that, Cloudschatze! I'll try the Sam & Max command-line parameter tonight, and will attempt installing the AWE32 & AWE64 this weekend. 😀

Cloudschatze wrote on 2020-05-28, 03:51:

darry wrote on 2020-05-28, 03:39:

Creative Labs never managed to make a full bug-free MPU-401...

In the dual-card scenario that's been described, the SB16's MPU is "bug-free."

Except, for example, for the bug that manifests itself in Duke Nukem 3D when digital audio>11KHz is played while MPU401 is used on any DSP SB16 .

AKA "* MPU-401 Stuttering with high sampling rates."

See Sound Blaster 16 Bugs and Deficiencies Summary .

This dual card scenario is a workaround, it does not make Creative Labs' implementation bug-free .
I would call it MPU-401 undesired symptom-free, but not bug free .

EDIT : My point was that an SB16/AWE32/AWE64 with an MPU-401 implementation free of all bugs, avoidable or not, does not exist , whereas Creative Labs' competitors have been successful in not having such bugs .

Cloudschatze wrote on 2020-05-28, 03:47:

samnmax m s220/5/16 g330
(Mouse, Sound Blaster at 220h, I5, 16KB buffer, General MIDI at 330h)

This seems to have worked with a 32KB buffer! Thanks so much.

[EDIT] Spoke too soon! The carnival area plays fine, but sitting through the intro leaves a hanging note that never goes away, whether set to 16KB or 32KB.

darry wrote on 2020-05-28, 04:24:

This dual card scenario is a workaround, it does not make Creative Labs' implementation bug-free .

Not bug-free, "bug-free." 😀

Of course it's a workaround. That's the entire premise of this thread - whether a dual-card scenario involving two SB16 architectures can mitigate the issues with the SB16's MPU implementation, and it can.

Cloudschatze wrote on 2020-05-28, 04:38:

darry wrote on 2020-05-28, 04:24:

This dual card scenario is a workaround, it does not make Creative Labs' implementation bug-free .

Not bug-free, "bug-free." 😀

Of course it's a workaround. That's the entire premise of this thread - whether a dual-card scenario involving two SB16 architectures can mitigate the issues with the SB16's MPU implementation, and it can.

Agreed.

If already going through the trouble to set up dual sound cards in DOS, I wouldn't use an SB16 + AWE64 (which is essentially two SB16). I would use an AWE64 and some SBPro clone that has a real OPL3 synth and a MIDI out port. This can really cover a lot more ground.

I use a SB32 3670, basically a AW64 gold "lite"
I your case, if possible get a Audician 32. Bug free, Opl3 and the wavetable software is fine.