leileilol wrote:

CS4232's okay, CS4231A is terrible

In what way?

The onboard FM can be deactivated via software and then comes the FMonster 😀
Don't get me wrong the FM is good on the crystal, but some music sounds strange/off, some don't.
So for me the by far best way seems to use the excellent! sbpro compatible crystal-onboard and
deactivate the FM and use an isa-fm only board...therefore the fm-monster will be excellent for that.

cs432x are all based in cs4231,which is a reference codec. Many highend cards from 90's used crystal codecs, mainly cs4231, I think if someone says that cs4231 card sounds bad, it's because card is broken in some way.

Yes, CS4231 was considered the best codec of its time. But it is a codec and not a controller. Interfacing it to a PC without a controller, in style of the original Windows/Microsoft Sound System card, is not very user friendly. CS4232 and later are controller+codec in one, with much more interfacing functionality. Apples and Oranges.

gerwin wrote:

Yes, CS4231 was considered the best codec of its time. But it is a codec and not a controller. Interfacing it to a PC without a controller, in style of the original Windows/Microsoft Sound System card, is not very user friendly. CS4232 and later are controller+codec in one, with much more interfacing functionality. Apples and Oranges.

Yes, that's it, my point was, if the codec sounded bad is because something is broken. The codec itself or some other part of sound controller. Point is that many board manufacturers wanted it cheap so they could sell lots of cards, making 2 layer boards, with poor ground and supply, so sound was there but not as it should. If you see the high end cards with cs4231, they all have 4 layers, like some audiotrix and gravis cards. Sound is so much cleaner.

I agree. My previous post was no critisism, just a general reminder.
Though I doubt Gravis is among the good examples of CS4231 use. Their cards before the Interwave were often among the worst performers in audio quality tests.

I remember an old usenet post from a Mediatrix employee, asking around for companies selling CS4231 chips, because they were having trouble stocking enough of them. He had no interest in the Analog Devices equivalent.

In late 90's laptops, the ones with the best PCM output have been the ones with Crystal chips. PCM/WAVE output on Crystal chips is excellent in my experience.

Yup, I love Crystal chips. As many of the posts above have mentioned, I've found the FM to be close to a genuine OPL3, with a few notes that do sound off or odd, but for me, it definitely sounds a lot better than CQM.

The SB/SB Pro PCM output of the CS4232/5/6/7 (as long as the analogue circuitry of the soundcard has been well designed) is arguably the best out there. Very good S/N ratio, the SB LPF sounds really good, no perceptible aliasing at the output and certainly has one of the most compatible SB/SB Pro controllers in it. The Epic Pinball mods sound sooo good on Crystal chips, especially compared to the YMF715/8/9 OPL-SAx chips. (I also really like the OPL-SAx chips, just prefer the Crystals)

I found my unicorn a few years ago by accident, the Turtle Beach Tropez Plus/TBS-2001. It has a Crystal CS4232, 4MB ICS Wavefront Wavetable, 3 SIMM slots for RAM, genuine YMF262 OPL3 and a Yamaha YSS225 SoundFX processor. It was also very well designed, the only other soundcard that's on par to it's quietness and sound quality is the AWE-64 Gold. Pity the AWE-64 Gold was only SB compatible and not SB Pro compatible and only had CQM, otherwise it would have been pretty much the perfect ISA soundcard.

No fans necessary! The Crystal CS423 chips can be passively cooled. 😎

...Okay, I'll see myself out. 😉

No fans necessary! The Crystal CS423 chips can be passively cooled. 😎

🤣 🤣

Will need liquid nitrogen when you overclock them for 24-bit 192KHz output..

I have tested the DMX Xfire 1024 (CS4624) with a PIII Coppermine FSB133 Slot 1 i820 system under Windows ME.

At first, I tested the latest Terratec WDM 5.12.1.3057 drivers from the Terratec website. These are crap drivers, because I get sound stutting and crackling sounds in many games, like Mafia 1 or Rally Trophy. Only reducing the Sound Hardware Acceleration to Basic has helped.

Next, I tried out the latest 5.12.1.4101 WDM drivers for the Hercules Gamesurround Fortissimo II, same layout and same CS4624 chip as the Xfire 2014! The drivers are much better, Mafia 1 now with a beautiful sound with 100% Sound Hardware Acceleration, also with EAX, Rally Trophy now also runs fine with DirectSound 3D Hardware, also tried with A3D, sounds ok, but freezes after a short time and the sound hangs. Incoming also sounds fine with A3D and showing Sensaura Splash Screen on start.

I now ordered a Turtle Beach Santa Cruz with the CS4630. The latest WDM version is 5.12.1.4193 for the Cruz, should also work with CS4624 cards?

Is there a website with all released VXD/WDM reference drivers for the CS4624/4630 cards?

I think I have a couple of these, one is a spare I keep in my parts bin that I've not even used yet, the other is the WSS Compatible CS-4231-KQ in my NEC Versa M/75.

Is there a difference between the CS4231A and CS4231KQ? I keep reading on websites back and fourth that this chip has FM, and then it does not. So far have not been able to get FM working on my 4231 at all, even with sB Compatible mode. That's one reason I've been stalking E-bay for PCMCIA sound cards and toying with making my own PCMCIA adlib clone.

I'll say this, the sound quality of digital audio on the 4231 is amazing to say the least, even crammed through that headphone blipper on the Versa. Games with WSS Support and high quality samples tend to sound excellent and very clear. I'm planning to soon try out some DAW work with the M/75 to see how it takes input from guitars/basses/synths. If my DX4 desktop can do it well, no reason the DX4 laptop can't.

4231 has no internal FM at all - maybe the info assumes an external FM chip (so as to have a cs4231+fm "full" WSS compatible card)
AFAIK internal FM started with the broken implementation in the CS4235 chip, it was fixed in CS4236 and carried over to CS4237 and CS4238 - beyond these I have no knowledge what the next chips did

keropi wrote on 2021-04-09, 19:43:

4231 has no internal FM at all - maybe the info assumes an external FM chip (so as to have a cs4231+fm "full" WSS compatible card)
AFAIK internal FM started with the broken implementation in the CS4235 chip, it was fixed in CS4236 and carried over to CS4237 and CS4238 - beyond these I have no knowledge what the next chips did

Ah, okay, thx for the clarification. It's kind of nuts because I see quite a few posts here and on VCFED for that chip.

keropi wrote on 2021-04-09, 19:43:

4231 has no internal FM at all - maybe the info assumes an external FM chip (so as to have a cs4231+fm "full" WSS compatible card)
AFAIK internal FM started with the broken implementation in the CS4235 chip, it was fixed in CS4236 and carried over to CS4237 and CS4238 - beyond these I have no knowledge what the next chips did

Does anyone know when the CS4289 FM synth companion chip first came out?

Difference between 4231 and 4231A is that A adds a way to turn the chip into a dumb DAC + mixer for some external device. It is something no PC orientated thing uses, they want the controller part to be active too for WSS business.

gdjacobs wrote on 2021-04-10, 19:43:

keropi wrote on 2021-04-09, 19:43:

4231 has no internal FM at all - maybe the info assumes an external FM chip (so as to have a cs4231+fm "full" WSS compatible card)
AFAIK internal FM started with the broken implementation in the CS4235 chip, it was fixed in CS4236 and carried over to CS4237 and CS4238 - beyond these I have no knowledge what the next chips did

Does anyone know when the CS4289 FM synth companion chip first came out?

I have one card with this chip, I can check the date code for the card and chip if you want. It won't be a definitive answer but can be a baseline.

Also, is there solid proof that this chip's OPL3 is worse than CS4236 anywhere? I hear this parrotted over and over but to me it sounds pretty identical to CS4236? I think Crystal just integrated this into the CS4235, fucked it up then fixed it in CS4236..

appiah4 wrote on 2021-04-12, 09:36:

Also, is there solid proof that this chip's OPL3 is worse than CS4236 anywhere? I hear this parrotted over and over but to me it sounds pretty identical to CS4236? I think Crystal just integrated this into the CS4235, fucked it up then fixed it in CS4236..

Surely there's a way of running through the synth operations of the chips then doing a differential analysis to see how they compare.

gdjacobs wrote on 2021-04-12, 20:09:

appiah4 wrote on 2021-04-12, 09:36:

Also, is there solid proof that this chip's OPL3 is worse than CS4236 anywhere? I hear this parrotted over and over but to me it sounds pretty identical to CS4236? I think Crystal just integrated this into the CS4235, fucked it up then fixed it in CS4236..

Surely there's a way of running through the synth operations of the chips then doing a differential analysis to see how they compare.

Yeah, but no one seems to have done it despite the claims that one is inferior (which is not my experience..)

appiah4 wrote on 2021-04-13, 07:51:

Yeah, but no one seems to have done it despite the claims that one is inferior (which is not my experience..)

I only have a CS4289 paired with a CS4232 (the famous Zoltrix card), and I find the FM output harsh and objectionable. Not as bad as CQM, mind you. I never noticed any of the FM clicking the CS4235 produces, but I wasn't actively looking for that last time I had that card installed.

I suspect a lot of it is horses for courses, and I just don't like CrystalFM.