I know that the MPU401 emulator is reporting that it's disabled, but is it possible to set the Base Address to a different port such as 320, 340, or 300 before then disabling the PAS16's MPU401 emulator? Might be worth a try if possible to see if it clears up whatever the issue is.

vfxsup wrote on 2024-08-15, 16:07:

I have a 486 with the following sound cards in it: […]
Show full quote

I have a 486 with the following sound cards in it:

1. PicoGUS in MPU-401 mode (330, IRQ 2/9)
2. SB 1.5 (IRQ 7)
3. Pro Audio Spectrum 16 (IRQ 11)

The SB and the PAS16 work well. Outputting anything over the PicoGUS/MPU-401 freezes the system. Removing the PAS16 and the PicoGUS/MPU-401 suddenly works perfectly.

The PicoGUS is running the latest firmware (2.0.0) and pgusinit (3.0.0). Jumpers are set to IRQ 2/9 and DMA disabled (tried it with the DMA enabled, didn't make a difference).

PAS16 is configured in the config.sys as follows:

device=c:\pas16\mvsound.sys d:5 q:11 s:0 :m0 :j0

mvsound.sys is the latest version 3.26.

Running proas.exe seems to confirm that the MPU-401 emulator is disabled on the PAS16:

MPU 401 Emulator Information: Active: 00 Port: 0330 IRQ: 00 […]
Show full quote

MPU 401 Emulator Information:
Active: 00
Port: 0330
IRQ: 00

First I've thought that this might be the same problem as Problems with Pro Audio Spectrum 16 and PC MIDI clone card but using the latest mvsound.sys didn't fix it. Anyone has an idea where to look next?

Thanks!

What is the exact version of the PAS16?

I have a PAS16 and picogus working together in a 486-100 system with a 3rd sound card (soundscape). I am using the Thunderboard chip on the pas16 for sb. Here are my resources.

Pas16: irq 12, dma 6, port 388, midi disabled
TB: irq 5, dma 1, port 220
SS: irq 10, sbirq 7, dma 0, midi port 350, sb emulation disabled
Picogus: irq 2, dma 3, gusport 240, midi 330
Joystick enabled on pg disabled on other card
Pas16 running mvsound.sys v03.23
Pas16 is a pro audio studio card PCB 3.4.3

I have not done extensive testing but all sound and music options are working in duke3d setup including ext midi via picogus

Rest of system resources:
Bus mouse on irq 4 (no ser/par ports)
Isa nic on irq 11, io base 360
Ide on irq 14
CD-ROM is scsi attached to pas16 using tslcdr.sys
Vlb vga s3 805 chip
Mobo is AIR 486sh rev 1.1, sis 85c461 chipset

Hope this helps

kaebeez wrote on 2024-08-17, 09:42:

I know that the MPU401 emulator is reporting that it's disabled, but is it possible to set the Base Address to a different port such as 320, 340, or 300 before then disabling the PAS16's MPU401 emulator? Might be worth a try if possible to see if it clears up whatever the issue is.

Good idea! Unfortunately, the second time you call mvsound.sys from config.sys, it throws an error that it's already loaded. I've tried to enable the PAS16's MPU401 with a different port, but that didn't help either...

zuldan wrote on 2024-08-17, 11:37:

What is the exact version of the PAS16?

It's a PAS16 LMSI, 650-0042-01K. I'll include a picture here:

These are (I believe) the jumper settings for it: https://theretroweb.com/expansioncard/documen … ation/53358.pdf

mistermister wrote on 2024-08-18, 15:09:

I have a PAS16 and picogus working together in a 486-100 system with a 3rd sound card (soundscape). I am using the Thunderboard […]
Show full quote

I have a PAS16 and picogus working together in a 486-100 system with a 3rd sound card (soundscape). I am using the Thunderboard chip on the pas16 for sb. Here are my resources.

Pas16: irq 12, dma 6, port 388, midi disabled
TB: irq 5, dma 1, port 220
SS: irq 10, sbirq 7, dma 0, midi port 350, sb emulation disabled
Picogus: irq 2, dma 3, gusport 240, midi 330
Joystick enabled on pg disabled on other card
Pas16 running mvsound.sys v03.23
Pas16 is a pro audio studio card PCB 3.4.3

I have not done extensive testing but all sound and music options are working in duke3d setup including ext midi via picogus

Rest of system resources:
Bus mouse on irq 4 (no ser/par ports)
Isa nic on irq 11, io base 360
Ide on irq 14
CD-ROM is scsi attached to pas16 using tslcdr.sys
Vlb vga s3 805 chip
Mobo is AIR 486sh rev 1.1, sis 85c461 chipset

Hope this helps

Thank you! This shows that it is indeed possible and is a great reference for further experimentation.

vfxsup wrote on 2024-08-19, 09:53:

It's a PAS16 LMSI, 650-0042-01K. I'll include a picture here: […]
Show full quote

zuldan wrote on 2024-08-17, 11:37:

What is the exact version of the PAS16?

It's a PAS16 LMSI, 650-0042-01K. I'll include a picture here:

These are (I believe) the jumper settings for it: https://theretroweb.com/expansioncard/documen … ation/53358.pdf

Thanks the reason I asked is that I have 2x 650-0022-01 (early revision). They both refuse to work nicely with any other sound cards. It’s like their address range overlaps (bug?). I also have 2x 650-0060-13B (later revision) and they work no problem with other sound cards. https://dosdays.co.uk/topics/retro_review_pas16_pt1.php

Question to polpo:

I noticed that the Sound Blaster mode has quite some lag, and that it is more / different than the lag of the Adlib component of the Sound Blaster mode. I noticed this while playing Scream Tracker 3 modules with both Adlib and sample instruments. If you listen any song of the Amnesia demo you clearly hear the sampled instruments lag behind. I also made a small test file if you need to confirm this for yourself.
My question is: is the lag / delay by design and unavoidable for performance reasons? I suppose it matters less in games (but there is still a delay between action and sound). Could it be fixed or at least equalised with the Adlib code, so they have the same latency?

digistorm wrote on 2024-08-19, 21:39:

Question to polpo:

I noticed that the Sound Blaster mode has quite some lag, and that it is more / different than the lag of the Adlib component of the Sound Blaster mode. I noticed this while playing Scream Tracker 3 modules with both Adlib and sample instruments. If you listen any song of the Amnesia demo you clearly hear the sampled instruments lag behind. I also made a small test file if you need to confirm this for yourself.
My question is: is the lag / delay by design and unavoidable for performance reasons? I suppose it matters less in games (but there is still a delay between action and sound). Could it be fixed or at least equalised with the Adlib code, so they have the same latency?

You are probably better off asking this in his thread PicoGUS: ISA sound card emulator with Raspberry Pi Pico (Gravis Ultrasound, AdLib, MPU-401, Tandy, CMS)

Yeah wrong topic excuse me.

zuldan wrote on 2024-08-19, 11:09:

Thanks the reason I asked is that I have 2x 650-0022-01 (early revision). They both refuse to work nicely with any other sound cards. It’s like their address range overlaps (bug?). I also have 2x 650-0060-13B (later revision) and they work no problem with other sound cards. https://dosdays.co.uk/topics/retro_review_pas16_pt1.php

So it seems there are indeed differences between the revisions. Thanks for that information. I'll be looking out for a late revision card and will try with that as well...