thanks for the feedback Shponglefan , it is great to have a direct comparison with the original card from someone else
Goldlib is as much of a clone of the original as it is possible - so the same audio quality issues will exist in both , noise included 😀
I also have the feeling the original Adlib drivers/programs are not so great , but that's another issue alltogether 🤣

Also from my side many thanks for the test and sharing your results. It confirms that the card and drivers with the GoldLib behave like the original. And that's exactly what we intended.

The problem with the jukebox, as far as I remember, only occurs when the mixer application is open and a specific song from the playlist is playing. At least that was the case for me.

640K!enough wrote on 2022-11-12, 07:31:

I mentioned this on the previous page, but for those testing this card on faster/more recent systems, it will likely be helpful to try all of the four combinations between Delayed Transaction and Passive Release. I can't say if/how it will help with this specific design, but it definitely makes a difference with the genuine 1992 original.

Okay, on the ASUS CUBX-L "Delayed Transcation" and "Passive Release" is controlled by an option called "PCI 2.1 Support". Unfortunately disabling doesn't affect the Adlib Gold, what hasn't worked before still doesn't work (Mixer.EXE, recording audio in tests). I will have to assemble a less powerful system, but this will take some time.

keropi wrote on 2022-11-12, 18:34:

thanks for the feedback Shponglefan , it is great to have a direct comparison with the original card from someone else
Goldlib is as much of a clone of the original as it is possible - so the same audio quality issues will exist in both , noise included 😀
I also have the feeling the original Adlib drivers/programs are not so great , but that's another issue alltogether 🤣

JimWest wrote on 2022-11-12, 20:15:

Also from my side many thanks for the test and sharing your results. It confirms that the card and drivers with the GoldLib behave like the original. And that's exactly what we intended.

The problem with the jukebox, as far as I remember, only occurs when the mixer application is open and a specific song from the playlist is playing. At least that was the case for me.

Happy to help! Thank you for creating such a cool sound card!

Not too fussed about the official Adlib software. The important thing is that games appear to work and work well. 😁

If someone wants to use the MCS MusicRack, that was available from Adlib, it's Y2K incompatible. After launching, it will hang at the splash screen. You have to look inside MediaRacks directory for 2 files: INLIST.DBF and PLAYLIST.DBF and edit the second byte, it's the year, and it should be no greater than 99 (it's a dBase Y2K timestamp bug). Setting it to 0x5D (93) will let the program start. The same procedure works for the "enhanced" version called MCS Stereo.

JimWest wrote on 2022-10-26, 12:14:

Oh very cool! […]
Show full quote

Beegle wrote on 2022-10-25, 20:39:

Lennart wrote on 2022-10-25, 20:36:

Yeah, that should do the trick 😀

I'll report back when get a minute to do this. Thanks for your patience!

Oh very cool!

If you could also provide the software etc. that would be great.

Thank you!

Back to report my findings. Hope this helps anybody who wants to make a MIDI cable for their GoldLib in the future 😀

- User's guide scanned

- Floppies archived (Midisoft Recording Session for AdLib Gold)

- Pinout established (to the best of my knowledge)

See animated image below to see the pinout more visually.

Beegle wrote on 2022-11-23, 04:17:

Back to report my findings. Hope this helps anybody who wants to make a MIDI cable for their GoldLib in the future :) […]
Show full quote

JimWest wrote on 2022-10-26, 12:14:

Oh very cool! […]
Show full quote

Beegle wrote on 2022-10-25, 20:39:

I'll report back when get a minute to do this. Thanks for your patience!

Oh very cool!

If you could also provide the software etc. that would be great.

Thank you!

Back to report my findings. Hope this helps anybody who wants to make a MIDI cable for their GoldLib in the future 😀

---

- User's guide scanned AdLib Gold MIDI PAK - User's Guide.pdf

---

- Floppies archived (Midisoft Recording Session for AdLib Gold)Midisoft Recording Session (for AdLib Gold) - Disk 1.zipMidisoft Recording Session (for AdLib Gold) - Disk 2.zip

---

- Pinout established (to the best of my knowledge)AdLib Gold MIDI PAK - Pinout.txt

---

See animated image below to see the pinout more visually.
AdLib Gold MIDI PAK - Pinout.gif

Thanks for the pin-out and software. The pin-out looks very different from the "standard" Sound Blaster one, so it's no surprise it requires a custom cable. The question is, is this cable passive, like someone mentioned? Does the Ad Lib Gold/GoldLib integrate the circuitry that is normally inside the MIDI cable?

GL1zdA wrote on 2022-11-23, 08:37:

[...]
Does the Ad Lib Gold/GoldLib integrate the circuitry that is normally inside the MIDI cable?

nope, it is exactly like the original board
actual differences are very few other than the adlib asic replacement and the support ICs that were required
I can only thing the CD-ROM Audio and speaker connectors that got replaced with more common ones

btw, how is the midi functionality in games? is it usable at all? does playback work under windows for DOS titles?

keropi wrote on 2022-11-23, 08:57:

GL1zdA wrote on 2022-11-23, 08:37:

[...]
Does the Ad Lib Gold/GoldLib integrate the circuitry that is normally inside the MIDI cable?

nope, it is exactly like the original board

Yes, I know, and it seems the original had everything on board and so does GoldLib. This is the opto-isolator for the MIDI in, that normally sits inside the cable in a Sound Blaster compatible one. Adlib Gold/GoldLib have one on-bard, similar to cards like the Turtle Beach MultiSound Classic/Tahiti/Maui. If the opto-isolator is on-board, then I would assume the other parts are also present, so a passive cable might be enough.

GL1zdA wrote on 2022-11-23, 09:15:

If the opto-isolator is on-board, then I would assume the other parts are also present, so a passive cable might be enough.

Correct. Furthermore, this same design element was re-used with the Audiotrix Pro, so the cable intended for that card can also be used with the Ad Lib Gold. (Not that this helps from an availability perspective...)

Attached is the full MIDI cable pinout provided by Mediatrix, in case it's of any interest at this point. The MIDI Thru connector was apparently optional - none of the Audiotrix Pro MIDI cables that I've encountered include it.

Cloudschatze wrote on 2022-11-23, 17:13:

GL1zdA wrote on 2022-11-23, 09:15:

If the opto-isolator is on-board, then I would assume the other parts are also present, so a passive cable might be enough.

Correct. Furthermore, this same design element was re-used with the Audiotrix Pro, so the cable intended for that card can also be used with the Ad Lib Gold. (Not that this helps from an availability perspective...)

Attached is the full MIDI cable pinout provided by Mediatrix, in case it's of any interest at this point. The MIDI Thru connector was apparently optional - none of the Audiotrix Pro MIDI cables that I've encountered include it.

Yeah, I've also noticed the similarity, even asked a friend today to make a photo of his Auditrix to compare it with the Ad Lib, since I couldn't find a photo with the part number readable. Indeed, the pinout provided by Mediatrix looks like it matches Beegle's findings.

hi folks!

thought i'd pass on a brief experience report of running the Goldlib alongside
a YMF719 based ISA (PNP) sound card for the CDROM version of Dune. (with apologies to anyone who finds this obvious!)

i started initial with a sb pro 2 (non-pnp) that has hardwired ADLIB io ports.
with line-out from goldlib to line-in on sbpro2 with sbmixer line-in set to 80 & fm synth set to 0.

this appeared to work OK in Dune, but as FM Synth commands were being played by the sbpro2 then muted,
I decided switched to a YMF719 based card.

setting the variable
BLASTER=A220 I5 D1 Y0 F3A8 T4
and running unisound with line-in mixer settings configured the YMF719's adlib port to a non standard port:
unisound /VL80

here, even with the FM Synth of the YMF719 volume at normal levels, only the goldlib renderes the sound track.
at all times, Dune seems to respect the line-in mixer config

the DUNE.bat file with config for goldlib + soundblaster digital audio contains:
DNCDPRG ENG ADG388 SBP2205 HIM 386 SAF WRIC:\GAMES\DUNE\

thanks to @keropi and all involved. never thought i'd hear this!

slartibardfast0 wrote on 2022-11-26, 18:47:

hi folks! […]
Show full quote

hi folks!

thought i'd pass on a brief experience report of running the Goldlib alongside
a YMF719 based ISA (PNP) sound card for the CDROM version of Dune. (with apologies to anyone who finds this obvious!)

i started initial with a sb pro 2 (non-pnp) that has hardwired ADLIB io ports.
with line-out from goldlib to line-in on sbpro2 with sbmixer line-in set to 80 & fm synth set to 0.

this appeared to work OK in Dune, but as FM Synth commands were being played by the sbpro2 then muted,
I decided switched to a YMF719 based card.

setting the variable
BLASTER=A220 I5 D1 Y0 F3A8 T4
and running unisound with line-in mixer settings configured the YMF719's adlib port to a non standard port:
unisound /VL80

here, even with the FM Synth of the YMF719 volume at normal levels, only the goldlib renderes the sound track.
at all times, Dune seems to respect the line-in mixer config

the DUNE.bat file with config for goldlib + soundblaster digital audio contains:
DNCDPRG ENG ADG388 SBP2205 HIM 386 SAF WRIC:\GAMES\DUNE\

thanks to @keropi and all involved. never thought i'd hear this!

I'd been wondering about trying to run the Adlib Gold / Goldlib alongside other sound cards. Thanks for posting this, sounds like something worth trying. 😀

Searching for the fastest config running the GoldLib reliably, I've put it into a Pentium MMX PC. At 233 MHz the mixer, that didn't want to start (hung) on the VIA C3 (even at 200 MHz) would start, but playing with the surround module would still cause hard hangs. Downclocked to 100 MHz (2x 50 MHz), the surround module starts to work reliably. There's still a possibility, it will hang, but mashing keys as fast as possible to switch it on/off or change the effect will most of the time work OK. The motherboard is the ASUS XP55T2P4 with the 430HX/PIIX3 chipset, 8-bit I/O recovery time is set to 8 BUSCLK and Passive Release is disabled. There's no setting for Delayed Transactions.

The thing that still doesn't work on a Pentium 100 MHz is recording - it tries to record, but will just play back silence.

After some initial issues making Goldlib and Orpheus coexist in a build while using UNISOUND, I've been able to make it work after some pointers/ideas from Keropi.

I switched from UNISOUND to ORPHINIT, first of all.
Then created two .INI files for ORPHINIT:

GOLD.INI:

1OPLBase=398
2FMMode=off

ORPHEUS.INI:

1OPLBase=388
2FMMode=OPL3

Then created two menu entries in autoexec.bat:

GOLD:

1ORPHINIT /F:GOLD.INI
2SET GOLD=388

ORPHEUS:

1ORPHINIT /F:ORPHEUS.INI
2SET GOLD=398

Probably some of those things have no effect or are not necessary (such as setting OPLBase when FMMode is off) but it works perfectly so I couldn't be bothered making more tests 😁

Yesterday I didn't know this card existed, today I don't know how I'll live without it. Hoping my position in the ordering queue is not too big 🤞

Maybe it's a dumb question, but for the recording test I just have to plug-in a typical mic from a headset into the "last" jack, and it should work? Because it's the one I'm still struggling with.

The jack is correct. But I don’t think a „modern“ mic will work. I tried several one and only one worked for me. It was this one https://www.ebay.de/itm/133877655249

JimWest wrote on 2022-12-01, 15:39:

The jack is correct. But I don’t think a „modern“ mic will work. I tried several one and only one worked for me. It was this one https://www.ebay.de/itm/133877655249

Okay, I educated myself about mics, and it seems most of these “small”, cheap mics used in headsets and these “stick” multimedia mics from the 90s are electret mics, and they need some power to work (not the 48V phantom power condenser mics need, but still some power), since they have a small amplifier inside. The one from eBay is a dynamic mic and will work without any power.

Boohyaka wrote on 2022-11-29, 10:03:

After some initial issues making Goldlib and Orpheus coexist in a build while using UNISOUND, I've been able to make it work aft […]
Show full quote

After some initial issues making Goldlib and Orpheus coexist in a build while using UNISOUND, I've been able to make it work after some pointers/ideas from Keropi.

ORPHEUS:

Copy code to clipboard

1ORPHINIT /F:ORPHEUS.INI
2SET GOLD=398

There is one potentially important point to be made here: setting the GOLD environment variable has no effect on the hardware. If the clone implements this the same way as the genuine card, unless the I/O port change is written to the configuration registers of the control chip (or written as part of its stored configuration), I think you will find that the Gold is still, in fact, at 388H.

Changing the environment variable may prevent some software from trying to probe the control chip at 388H, but there are no guarantees. A more aggressive detection routine may still try to fall back to 388H, at which point it may upset the YMF289 on Orpheus. If that happens, a re-boot or just running ORPHINIT again should silence any noise produced and restore it to a usable/valid state. One option you may want to try, if you have problems with your current solution, is to re-locate the Gold to 398H and write that to its persistent configuration. Then, if you actually want to use the Gold with software that doesn't know how to find it at 398H, you can simply move it back to 388H without writing that to the permanent configuration. That way, a re-boot will bring it back up at 398H.