JazeFox this is amazing. Thank you for putting in the effort to create this so someone as computer illiterate as me can enjoy this era of gaming. Using this got music to work in my dos games without any hitch. I had been pulling my hair out trying and failing to fix the issue. Just one file and it’s completely working.

Did anyone try to attach a CD-ROM to the PnP IDE interface on the sondcard (initialized with Unisound) ?

I have an issue with the ESS1868 cards. However, the Сrystal CX4237 card works fine (with the same software and CD-ROM).

1) Working card: Сrystal CX4237, CD-ROM DRIVE, UNISOND, DEVLOAD, VIDE-CDD.SYS 2.14, MSCDEX

2) Non-workng cards: ESS1868 (all cards give the same effect),CD-ROM DRIVE, UNISOND, DEVLOAD, VIDE-CDD.SYS 2.14

(I've made a long image since forum won't attach more then 5 files)

@alec_v did you also try address 168 and/or IRQ 10 on the ESS?

The IDE interface can be configured to run at I/O address 168h or 1E8h, and use IRQ 10, 11 or 12.

https://dosdays.co.uk/topics/Manufacturers/ess.php

Hi,

with current version of Unisound Terratec Base-1 does not longer initialze. (0.80, 0.77 doesnt work)
with Version 0.65 it works just fine.

Something is broken here.

https://www.philscomputerlab.com/terratec-pro … dia-base-1.html

Old Version attached, in case something has the same Problem.
It seems there is no version archive ?

1541 wrote on 2023-02-06, 19:37:

@alec_v did you also try address 168 and/or IRQ 10 on the ESS?

You are right, 3'rd IDE address works for the ESS1868!
I can't use 4'th IDE since I have a 3C509 card on the IRQ10

alec_v wrote on 2023-02-13, 20:11:

I can't use 4'th IDE since I have a 3C509 card on the IRQ10

If you are not using serial COM2 port, you can disable it and set 3C509 card IRQ to IRQ3.

matze79 wrote on 2023-02-07, 21:30:

Hi, […]
Show full quote

Hi,

with current version of Unisound Terratec Base-1 does not longer initialze. (0.80, 0.77 doesnt work)
with Version 0.65 it works just fine.

Something is broken here.

https://www.philscomputerlab.com/terratec-pro … dia-base-1.html

Hi!
My Terratec Base-1 is working perfectly with v0.80 and v0.77 on different computers and configurations.
As you don't provide details, it's difficult to help. It would be very useful if you detail how it's failing for you, and some other details, to be able to help.

It's noted in the first post and in the Mini-FAQ point 2, but anyway I'll copy-paste here:

Q- I found a bug or my card does not work. What can I do?
A- First, check in UNISOUND.TXT if you're using the correct options with your card/system and your card is supported. If it's still not working, please, try adding this options to UNISOUND: /FS /XXF , also try the card in diferent systems if you have them, and if it's still not working or having problems, you can post here your problem, but please, do it with as much detail as possible: card brand/model/chips, system specs (cpu/motherboard/memory/other cards installed/OS, UNISOUND version and options used to initialize, if other cards work in the same system and a detailed explanation (with steps if possible, since the system boot) and tests done. Write also the exact output of UNISOUND and some photos could help too. Without details, it's more difficult to help.

And, please, v0.65 is a very very old version with many broken things. may I ask you to remove the attachment?

It's better to find the problem and solve it with newer versions.

Thanks!

(Quote from Ranger85 / dosreloaded.de / Terratec Base 1)

I have now made screenshots. In each case with the current version 0.80 at the top, followed by 0.76 and 0.65.

I tested in three different systems:

- 486DX4 100 MHz, SIS 471
- 386DX 40, Symphony 82c461
- Pentium 90, Intel Triton

In all three systems the exact same behavior, no matter if with auxiliary parameters from the readme or without.
I have the Base-1 twice. Makes no difference which one I use.

For comparison I took my Highscreen Sound-Boostar 16, with which it works in the new version.
This card is based on exactly the same AD1816 chip and does not differ at all from the Base1.
The old version 0.65 detects the card, but can't do anything with it.

My guess: The bug in the new version could be related to the PnP ID.

It doesn't seem to reveal source codes, so it's not possible to check if it's just a typo etc..

(Quote / end )
---

used Deepl to translate it for you.

I also can reproduce the same with my Intel Triton based Pentium.

--

i don`t see any point to remove something old that works.
It would be much better to keep each release around, for exactly such purposes.
imagine cutemouse would remove all old versions.. some newer don`t work proper with dos + windows 3.x in some systems.

@matze79: This actually did help me! I was playing around with the exactly same models of cards a while ago, one with a corrupted EEPROM at first, that I was able to reflash with the correct data, thanks to the nice help of pc-sound-legacy: Terratec Base-1 not working - looking for eeprom content. (the eeprom content can be found there, if anyone might need it).

But, even after that, it wasn't detected by unisound (and neither the drivers listed on phil's site).
So, I bought another one, finding it to be behaving completely identical. Without any quick idea of what might be causing this common issue (thinking it might maybe be some decaying AD1816), I put them aside after a while.

With the latest unisound still not detecting them, I now tried this older version. And: they indeed both get detected (as you posted) and do work perfectly fine with that. So, thank you!

Glad It did help you too,
So there already 4 people with this problem 😀

I also have the radio for the base 1 but didn’t yet try it 😉

matze79 wrote on 2023-03-06, 20:08:

[...]

@mazte79 @MJay99

Hi!

Thank you for all the testing, photos and details!! This will help to find the problem for sure!

It's not a PnP ID problem (In such case, the message Unknown card [XXXXXXX] .... should appear).
I suspect of some EEPROM brand differences that causes this problem in newer Unisound versions with ADI chips (and some very old cards of different brands). Please, can you test v0.80 this way?:

UNISOUND /FS

This should help a bit if that's the case. Let me know if it works. If not, I will release a new test version with some changes in speed for the detection phase in those cases.

About keeping the old version, no problem with that, it's OK 😀 I thought that the best way is always to have the latest version the best possible, with all bugs found and fixed, and the point in my request was only about that, trying to avoid confusion to new users with a mix of old versions that have bugs that should be avoided. As I try to give help and support regularly in order to provide a good program for everyone in the latest version, in theory, old versions are not needed. But I could be wrong, of course.

Let me know the results and we'll continue with the investigation with the details provided. Ah! I forgot to ask... could you check the EEPROM chips in the tested Base-1 cards? (Shoud be a 24C04) , and the brand? ATMEL, ST, ISSI, ATC.... ?

Thanks.

On both of my cards it's an ATMEL 24C04 and starting unisound 0.80 with /FS does detect and configure it properly, also.
Thanks for this hint, too!

Atmel 24c04 EEPROM

🙁

@matze79: Now that's really interesting... I went back, just to verify and having proof that it really is working with 0.80a:

This is on a 486DX2-66 VLB system, with nothing much of a PnP BIOS support otherwise.
Both cards I have (one repaired, one in original condition) do initialize this way.

Did you also try the /XXF option JazeFox mentioned a few posts back (Force detection of PnP cards with a partially corrupted EEPROM data (bad header checksum))?
Unless the older version did this implicitly, it probably might not help much... but still doesn't hurt to give it a try 😀

And maybe as a hail mary : /XNET (It's a workaround for Realtek (and compatible) ISA PnP network cards when paired with other ISA sound card(s).)

It works on 386DX40 with Parameter but does not work on P90 / Triton.

Maybe it’s timing related?

I have a Crystal sound card that is faking the Yamaha YMF719. The chip was sanded and relabeled, and the EEPROM strings were altered.
The UNISOUND successfully initialized the card, showing the faked name in "PnP card found:" row, and the only indication it's really the Crystal - is the name reported in mixer line "Crystal Mixer [..."
It would be good to have the detected hardware family reported explicitly in its own report row. Not only the PnP name which could be altered, but explicitly stating the name of the code branch that successfully brought card to life. This would help in identification of noname cards with forged identity.

i tried on Pentium 180Mhz ODP MMX with FX Chipset and it does also only work with 0.65 hm

Hi!

I wanted to clarify about Interwave support, as a recent post on Orpheus II thread seems to claim some kind of problem or incompatibility with UNISOUND's Interwave support.

So far, no known problems have been detected with Interwave support. If something wrong is detected with it, please, report it, with details, to be able to investigate and solve it as soon as possible.

I suspect that maybe that statement may be related with a post user Itling started here last year: Re: UNISOUND - Universal ISA PnP Sound Card Driver for DOS v0.78b
reporting some problems with samples, supposedly caused by RAM not being initialized (which in fact UNISOUND does).

I was in contact with user Itling in private, we spent about 2 weeks performing extensive tests with a lot of different systems, different Interwave-based cards, different software, games, demos, module players, over 40 different modules of different formats, different adjustments... Everything worked well. After a lot of testing, in the ends, it turned out to be a defective card Itling had in his collection. So in the end it was all a false alarm.

I have reviewed subsequent posts after those tests and I noticed that we forgot to mention the testing and results, and some time later, when v0.80 was released, no fixes were included in the changelog because there was nothing to fix. It was my mistake to have forgotten to mention it. Anyway I hope this is now clear.

-----------------------------

About Orpheus II support: If users and/or authors have interest, it would be a honor to add proper support for it to UNISOUND, of course.

Interwave part: If everything is like a GUS-Pnp or know compatible cards (including EEPROM and PnP IDs), everything should work out of the box.

Crystal part: I would need info about the connections between the Crystal and other parts like the OPL, mixer ins/outs... , the EEPROM config, PnP ID and device string.

----------------------------

matze79 wrote on 2023-03-17, 21:14:

...

The problem with some Terratec Base-1 cards has been located, being a mix of AD181x + EPROM model + Chipset/Chipset timings, and a workaround will be released in the next version. Note that very old (like v0.65) init and detection routines will not be used because it would be a big step back, as it would break support for a lot of cards and software that was added with later versions. But, no worries, In the next release, the standard init (without parameters) will address 99% of the problematic "mixes" of AD181x, and the other 1% (as the rare combination of some systems with Intel's FX + some ATMEL EPROMS) will be solved using a new parameter with a slightly different and slower method.

I can confirm JazeFox's systemene about my PnP issues: they were due to hardware problems which, for reasons known only to Murphy, seemed to get worse with time thereby confusing me greatly.
(The help and support from JazeFox was stellar, as usual.)

(PS: It's Ltning (short for Lightning because Doom didn't support long enough nicks..), not "ltling" ;) )

JazeFox wrote on 2023-03-18, 09:38:

I suspect that maybe that statement may be related with a post user Itling started here last year: Re: UNISOUND - Universal ISA PnP Sound Card Driver for DOS v0.78b
reporting some problems with samples, supposedly caused by RAM not being initialized (which in fact UNISOUND does).

I stand corrected. It was in reference to that report, and I had no way of knowing that it was ever investigated, nor that there was nothing to correct to begin with. My assumption was that you skipped the "voice parking" procedure when initialising the InterWave, and there was some sloppy application that caused garbage to be summed into the output. It should be unlikely, but it's not impossible.

JazeFox wrote on 2023-03-18, 09:38:

About Orpheus II support: If users and/or authors have interest, it would be a honor to add proper support for it to UNISOUND, of course.

Interwave part: If everything is like a GUS-Pnp or know compatible cards (including EEPROM and PnP IDs), everything should work out of the box.

Crystal part: I would need info about the connections between the Crystal and other parts like the OPL, mixer ins/outs... , the EEPROM config, PnP ID and device string.

If you don't already have all of the information you need for Orpheus II support, want to discuss any aspect of the initialisation routines, and/or need an additional volunteer for testing, please feel free to contact me privately.