Personally I think this card is priced fairly with respect to the design costs. That being said it is more expensive than it is worth to me. No disrespect intended, just a personal value decision.

keropi wrote on 2021-01-28, 20:04:

Yes it should - furan looked at the driver but sadly none of what was discussed worked for me: Re: Orpheus Soundcard: a new DOS soundcard with SPDIF/OPL3/MPU support
interesting info about the S32A , I wonder what happens under windows - is it sure that the 9233 still responsible for FM? do you have a link to it's drivers?

Actually, I think it's a wavetable synthesis chip, but the datasheet specifically mentions it along with the OPL in the external FM section.

Here's the best driver I could find for it, I haven't tested it:

the CS9233 *does* emulate an OPL using wavetable synthesis - it sounds crap as you might have expected 🤣 (if I remember where I've heard some samples I'll link it to you)
I kinda have the feeling though that the S23A will just use the internal CS4237 FM block... will investigate it

edit:
checked the driver, this is based on the older "driver generation" from Crystal so no spdif control or the later driver fixes and features for directsound 🙁

But does it use the External FM?
If so, you might be able to check the driver for differences and Frankenstein something together.
Also, did Furan check CWDFM.DRV?
It may be that the FM section of the driver controls it.
Another avenue is to look at the PNP configuration data, the datasheet mentions it helps configure internal/external FM choices

kolderman wrote on 2021-01-28, 19:04:

fitzpatr wrote on 2021-01-28, 19:02:

Intel486dx33 wrote on 2021-01-28, 18:49:

That’s $279 for an ISA sound card.
More expensive that 1993 prices.

The Sound Blaster 16 ASP was $349 on release...accounting for inflation, that is approximately $625 today.

And that's WITH economies of scale...shit in the 90s was as expensive as heck.

no wonder I didn't have a sound card until 1995/96 x3

DNSDies wrote on 2021-01-28, 20:41:

But does it use the External FM? If so, you might be able to check the driver for differences and Frankenstein something togethe […]
Show full quote

But does it use the External FM?
If so, you might be able to check the driver for differences and Frankenstein something together.
Also, did Furan check CWDFM.DRV?
It may be that the FM section of the driver controls it.
Another avenue is to look at the PNP configuration data, the datasheet mentions it helps configure internal/external FM choices

I will try to verify it but I kinda doubt FM will originate from 9233
The PnP configuration data on the card's eeprom are correctly set for external FM - the thing is that the windows driver does not respect that and just does it's thing
Furan lost interest after the registry keys that should work had no effect ... no idea if he looked at it or not.

So if you don’t have an Orpheus for DOS you are just listing to “Looney tunes” because the audio is NOT heard as the game developer intended ? Your music tones maybe off or you may hear allot of noise.

keropi wrote on 2021-01-28, 20:57:

I will try to verify it but I kinda doubt FM will originate from 9233 The PnP configuration data on the card's eeprom are correc […]
Show full quote

DNSDies wrote on 2021-01-28, 20:41:

But does it use the External FM? If so, you might be able to check the driver for differences and Frankenstein something togethe […]
Show full quote

But does it use the External FM?
If so, you might be able to check the driver for differences and Frankenstein something together.
Also, did Furan check CWDFM.DRV?
It may be that the FM section of the driver controls it.
Another avenue is to look at the PNP configuration data, the datasheet mentions it helps configure internal/external FM choices

I will try to verify it but I kinda doubt FM will originate from 9233
The PnP configuration data on the card's eeprom are correctly set for external FM - the thing is that the windows driver does not respect that and just does it's thing
Furan lost interest after the registry keys that should work had no effect ... no idea if he looked at it or not.

I had been experimenting with alternative Windows 9x drivers (like the WSS one included with Windows 98 SE and drivers meant for older Crystal chips), but when using S/PDIF out, AFAICR, I typically only got one channel (left or right, I forget which), so I had given up . Anyway, I personally don't really see a reason to use DOS OPL3 games under Windows, so it is not an issue for me .

EDIT: Over analogue out, AFAICR (it was a while back, but I may have posted about it) using Windows 98 SE's WSS driver was giving me weird aliasing for PCM audio .

Intel486dx33 wrote on 2021-01-28, 21:47:

So if you don’t have an Orpheus for DOS you are just listing to “Looney tunes”

It doesn't seem like you have any idea of the amount of effort that went into the design of this card and the numerous revisions that led to the final product. If its price is higher than you are willing to pay for a quality sound card, that is your decision, but it doesn't give you the right to disrespect their work. If you've decided you don't want one, that's fine, but kindly stop polluting the thread with unfounded criticisms and nonsense.

Do you write to Tim Cook to tell him that the latest iPad is too expensive, given that it's only a small iterative improvement over an existing product? I didn't think so. These are two people who spend their free time designing quality products for the community, then assemble and test each one. That isn't an effort that takes just a few seconds. This isn't Apple, where much of it is done by sub-contracted automated assembly and the rest in near-slave-labour conditions in Asian countries (while the executives count the billions). It's a small operation dealing with low-volume prices for quality parts, some of which are no longer made.

fitzpatr wrote on 2021-01-28, 17:35:

keropi wrote on 2021-01-28, 17:21:

It is the PC Speaker input - but there are cases than when using it it could degrade the audio quality - so it is left unpopulated. It can be used if you solder a pin header.

I assumed that, but wanted confirmation. For whatever reason, mine is populated with pins, but I'll test the use of it to make sure that it isn't a problem.

It is a simple analogue input, and isn't always enabled in software. When it is enabled and not muted, the cable may allow noise to be picked up and introduced into the audio. Since it will pass through the ADC, that may be present in the S/PDIF output as well (depending on configuration), and will be full-time, rather than only when the PC speaker is in use.

Also, a plain straight-through cable may not be enough. Some in-line components may be needed to ensure that the voltage does not exceed 1V p-pRMS.

This is probably a dumb question, but your eeprom has the 55h/BBh validation bytes at the beginning, right?

Also, did you read the section of the datasheet detailing how to bypass PnP using sending a 32 byte key to 0279h to force the chip into configuration mode?
It's on page 19 of the datasheet. The windows drivers might be doing this. If so, you could likely overwrite its configuration and send code to DISABLE any future PnP control requests and prevent Windows from commandeering it, at least until after rebooting.

640K!enough wrote on 2021-01-28, 22:28:

It is a simple analogue input, and isn't always enabled in software. When it is enabled and not muted, the cable may allow noise to be picked up and introduced into the audio. Since it will pass through the ADC, that may be present in the S/PDIF output as well (depending on configuration), and will be full-time, rather than only when the PC speaker is in use.

Also, a plain straight-through cable may not be enough. Some in-line components may be needed to ensure that the voltage does not exceed 1V p-p.

Thanks Stefano for chiming in on the speaker thing.
I must mention that the needed components for Speaker-INPUT to work safely are present in all cards that are sold so connecting the speaker cable will not cause any damage or other electrical issue.

DNSDies wrote on 2021-01-28, 22:35:

This is probably a dumb question, but your eeprom has the 55h/BBh validation bytes at the beginning, right?

Also, did you read the section of the datasheet detailing how to bypass PnP using sending a 32 byte key to 0279h to force the chip into configuration mode?
It's on page 19 of the datasheet. The windows drivers might be doing this. If so, you could likely overwrite its configuration and send code to DISABLE any future PnP control requests and prevent Windows from commandeering it, at least until after rebooting.

yes the eeprom data is fine
The windows CS4237 driver does just that, it configures the chip to whatever setup is baked into the driver on compile time (my guess) - it would actually not surprise me if the option for external OPL3 is bugged (or even untested/not-working) since at the time the chip was released (1997) OPL was already a thing of the past and to my knowledge no other CS4237-based card is paired with an actual OPL chip. So everything was hardcoded for the internal FM block and that was the end of it... it makes no difference at what state the card is before windows driver loads, it will still de-configure it and upload the new config it wants.
Maybe someone will come up with a way to alter this behavior in the future but it's certainly nothing marmes or me can do ...

640K!enough wrote on 2021-01-28, 22:28:

Intel486dx33 wrote on 2021-01-28, 21:47:

So if you don’t have an Orpheus for DOS you are just listing to “Looney tunes”

It doesn't seem like you have any idea of the amount of effort that went into the design of this card and the numerous revisions that led to the final product. If its price is higher than you are willing to pay for a quality sound card, that is your decision, but it doesn't give you the right to disrespect their work. If you've decided you don't want one, that's fine, but kindly stop polluting the thread with unfounded criticisms and nonsense.

Do you write to Tim Cook to tell him that the latest iPad is too expensive, given that it's only a small iterative improvement over an existing product? I didn't think so. These are two people who spend their free time designing quality products for the community, then assemble and test each one. That isn't an effort that takes just a few seconds. This isn't Apple, where much of it is done by sub-contracted automated assembly and the rest in near-slave-labour conditions in Asian countries (while the executives count the billions). It's a small operation dealing with low-volume prices for quality parts, some of which are no longer made.

+1

I was tempted to respond, but you likely worded it better than I would have .

I visited the website:
http://orpheus-soundcard.eu/

1) Did they ever fix the problem with the SPDIF driver ? ( LGR said there maybe a problem with the setup ? ).
2) At what state is this project ? ( Beta or near Final release ).
3) Are they going to fix known issues or leave it as is ?

Intel486dx33 wrote on 2021-01-29, 00:25:

I visited the website:
http://orpheus-soundcard.eu/

1) Did they ever fix the problem with the SPDIF driver ? ( LGR said there maybe a problem with the setup ? ).

What problem with S/PDIF ? S/PDIF works fine for me (and AFAIK, other people as well) in both DOS and Windows 98 SE . I have never heard of an issue, please provide more details .

Intel486dx33 wrote on 2021-01-29, 00:25:

2) At what state is this project ? ( Beta or near Final release ).

What kind of question is that ? The hardware has been shipping for months and works fine . As for initialization software, Unisound supported the card from launch day and the slightly later released Orphinit provided more flexibility . As for Windows 9x, the external OPL3 limitation is due to Crystal Semiconductor's driver, which is closed source, but the driver otherwise works fine, including S/PDIF output .

Intel486dx33 wrote on 2021-01-29, 00:25:

3) Are they going to fix known issues or leave it as is ?

What known issues ? The Duke Nukem II and Tyrian/Tyrian 2K ones ? Something else that you are conveniently not mentioning ?

The Duke Nukem II one is due to a limitation of the CS4237 chip and is not Orpheus specific . What are you expecting as a fix ? Crystal Semiconductor doing us a solid and redesigning the CS4237 chip, doing a new production run and sending little elves to replace the chips installed in all the cards (Orpheus and others) with a CS4237 (and all the other CS chips with the same limitation) ? By the way, there is a workaround for that Duke Nukem 2 ADPCM fix? (no elves required)

The Tyrian/Tyrian 2K one has a workaround that can be implemented by temporarily setting WSS DMA to DMA channel 0 . A 10 second effort of making a batch file can fix that .

If it is some other issue(s), please specify, nobody here is a mind reader, AFAIK .

EDIT : Corrected typos

I think it's best to just ignore that forum member guys, he's ruining an otherwise excellent thread.

I can't wait till it's my turn to give keropi money.

for what it's worth I love the whole idea behind the card and very much appriceate the work that went into it the fact that someone took the time to make this is actually a huge selling point for me I like to support projects like that.

It is true LGR mentions he had issues at the time with orphinit and he just used unisound instead.
He never contacted me about it and when I asked for more info he provided some basic one but was unable to provide more details since he had already moved to other projects.
He is actually the only person mentioning problems with orphinit - since then noone made a similar complain or submitted a bug report.
It is possible that someone else might also had an issue but since unisound also supports the card we never heard anything about it.

The Orpheus website - as simple as it is - conveys all info I could think of: from hardware to software limitations and known games that have issues.
If something else is discovered it will also be listed on the site.
Hardware-wise there are no errors or deficiencies in the design, software-wise if someone produces a valid bug report it will be looked into and fixed.
We have not "abandoned" the card and collect money from buyers - what software can be fixed will be fixed and any extra hardware limitations or incompatibilities will be listed so people can have all the info before they decide if the card is worth it for them or not.

Once again I would like to thank everyone for their incredible support and please rest assured we do everything we can to fulfil the pending requests - it just takes a considerable amount of time to build/test each card and we are truly grateful for your patience 😀

Having said that I hope the thread will return to more meaningful discussion - Orpheus (and every hardware out there for that matter) is not for everyone nor is it the panacea to DOS sound.
We did our best and came up with it - some people deem it worthy some don't, that's life.

Keep up the great work keropi. I'm really glad you made this card. I hope people like you and LABS proliferate and there are a lot more options in the market soon. The Amiga scene is alive with dozens of modern hardware projects, let's hope retro PC scene does the same soon.