ESS 1869F vs AWE32 vs YMF701 filter ON/OFF at 22kHz in DOS

ESS 1869F vs AWE32 vs YMF701 filter ON/OFF at 11kHz in DOS

AWE32 filtering at 11, 22 and 44kHz

Full bandwidth 44kHz: ESS1869 vs AWE32 vs YMF701 vs Gravis Ultrasound PnP (synth mode)

From my testing, the AWE32 has the cleanest implementation of the sampling rate-based variable lowpass filter.
The Yamaha ISA cards had the worst filtering out of the box, and even with the capacitor mod, it still has too slow a rolloff to compete.
The Ultrasnound PnP was by far the loudest and required lots of tweaking to not saturate/clip the line-in of the X-Fi soundcard I used to do recording.

Interesting stuff! You’ve listed all of the ESS chipsets as being WSS compatible - do you know how that works? I’ve never seen a driver package set up resources for it, where the OPLSAx and Crystal cards I’ve tried do it as a matter of course.

badmojo wrote:

Interesting stuff! You’ve listed all of the ESS chipsets as being WSS compatible - do you know how that works? I’ve never seen a driver package set up resources for it, where the OPLSAx and Crystal cards I’ve tried do it as a matter of course.

I'll admit that for the ESS cards, I went by their datasheets.
When I get home later, I'll try and see if it really works. What intrigues me though is that some cards don't support 48kHz, but claim WSS compatiblity, where the WSS does support 48kHz/16bit.

For FM synth subjective quality, I am very impressed with the ESS AudioDrive's implementation.

I'd say that after the genuine OPL3, the ESS AudioDrive is a close 2nd in terms of pleasing sound and similarity. The Crystal FM sound chip is a little more metallic and hollow sounding to the ESS chip with some obvious differences in some instruments. Most games and tracks I played sounded very close to their OPL3 equivalent, or at the very least, still sounded pleasing to the ears.
The CQM chip found on some SB16, AWE32 and AWE64 sounded the worst IMO. It had the most tinny and metallic sound and could easily be identified when compared to an OPL3 or even their competitors.

tpowell.ca wrote:

So far, for my 486 vintage PC, my favorite card is the ESS688 with the genuine OPL3 chip combined with a HardMPU card for midi duties.

Could you elaborate on that? According to your comparison table, ESS688 with true OPL3 only has advantage over SB Pro 2 in Windows which is not really relevant on a 486. I mean, AFAIK, the "16 bit" part comes from WSS, not SB16 compatibility (unless we count the few games that support ESS natively).

I'm curious because I kind of wanted to get a card like that because it looks cool to me, but I don't have a good enough practical reason because I already have an SB Pro 2. 😀

some cards with cs4237 have spdif out, like TB Malibu 64

jheronimus wrote:

tpowell.ca wrote:

So far, for my 486 vintage PC, my favorite card is the ESS688 with the genuine OPL3 chip combined with a HardMPU card for midi duties.

Could you elaborate on that? According to your comparison table, ESS688 with true OPL3 only has advantage over SB Pro 2 in Windows which is not really relevant on a 486. I mean, AFAIK, the "16 bit" part comes from WSS, not SB16 compatibility (unless we count the few games that support ESS natively).

I'm curious because I kind of wanted to get a card like that because it looks cool to me, but I don't have a good enough practical reason because I already have an SB Pro 2. 😀

I just checked again and the datasheet for the ESS chips all state WSS compatibility, but having just tried now to get it to work and nope. Nada and nothing. I'm guessing they never implemented the necessary initialization to enable WSS compatibility.
In Windows 9x, the built-in drivers don't use the WSS compatibility anyways. It works fine in 16bit at 44kHz stereo and sounds great. In no way any worse than an SB16.

The advantage of the ESS688 over an SBPro is the superior low pass filter implementation and it's a lot less speed sensitive to high bus speeds. Some games also natively support the ESS688, whick may give a slight quality advantage.

Nice comparison!
Could you please specify which CTxxxx model applies to your AWE32?

badmojo wrote:

Interesting stuff! You’ve listed all of the ESS chipsets as being WSS compatible - do you know how that works?

The 16-bit mode of these ESS audiodrives is similar but not the same as WSS. There are specific 16-bit ESS drivers for DOS Miles Sound System games, like Warcraft II.

The ES1688 card I have can be configured with jumpers. It is a nice card. There was an FM test here on vogons where ES-FM was tested to match exact OPL3 behaviour: Re: The way to detect OPL3 clone

gerwin wrote:

Nice comparison! Could you please specify which CTxxxx model applies to your AWE32? […]
Show full quote

Nice comparison!
Could you please specify which CTxxxx model applies to your AWE32?

badmojo wrote:

Interesting stuff! You’ve listed all of the ESS chipsets as being WSS compatible - do you know how that works?

The 16-bit mode of these ESS audiodrives is similar but not the same as WSS. There are specific 16-bit ESS drivers for DOS Miles Sound System games, like Warcraft II.

The ES1688 card I have can be configured with jumpers. It is a nice card. There was an FM test here on vogons where ES-FM was tested to match exact OPL3 behaviour: Re: The way to detect OPL3 clone

The tested AWE32 was a CT2760 rev3 (with the CT1747 with a licensed OPL3 synth).
The ESS688 I used is a jumperless Compaq part, but benefits from having a real YMF262 with a YAC51x and a proper line-out.

Honestly, if you have a late model Sound Blaster Pro 2 (with the 2nd crystal oscillator) there is probably no reason to look for an ESS-based replacement.
My Sound Blaster Pro is an early model unfortunately that really doesn't work well on anything faster than a slow Pentium 1.