SOUND BOARD CHIPSET: VARIOUS PRODUCTS
by Crystal Semiconductor

Last update: 12/26/97

Phone:
??
Fax:
??
BBS:
??
Web:
http://www.crystal.com, http://www.io.com/~timur/crystalos2.html (Timur Tabi's CS advocacy page)
e-mail:
pcaudio@crystal.cirrus.com
Type:
Varies
OS/2 Drivers:
Varies
Drivers from:
IBM, http://www.crystal.com/drivers/audio-d.htm, OS/2 Warp 4.0 box
Boards using this chipset include:
Numerous products, including Intel Atlantis motherboards, some Acer products, the Crystal Lake boards, and Crystal Computer Corporation products (but the last of these reportedly don't work in OS/2).
Comments:
Crystal Semiconductor manufactures a number of chips and chipsets, including the CS4231, CS4232, CS4236, CS4248, CS9233 (for wavetable support), and CS8905 (for reverb/chorus support). I'm unclear about precisely what most of these do, but many of them are in use in many products. CS is also now moving into the PCI arena with at least one PCI audio chipset (the 4610), but the last I'd heard, it was not yet supported under OS/2. Some boards using some CS chipsets (particularly the CS4231) may work with OS/2 3.0's "Business Audio" drivers, and others have drivers in the OS/2 Warp 4.0 box as part of the standard install process. Recent Crystal drivers reportedly support full duplex on at least some hardware, but only when the input and output sample rates are the same. I've also seen reports that these drivers allow OS/2 programs to retain access to the sound card when a sound-using Windows program is active. Truly simultaneous sound is still impossible, but these drivers negotiate access when sound requests are made rather than when the Windows program is launched. Most CS-based boards reportedly work well in OS/2, including VTD, though I've received some reports that it may be necessary to modify the OS/2 CONFIG.SYS line for CWSOUND.SYS to modify the default input volume. The parameter "/X2:9" worked well for the person making the report, but of course this may vary from system to system. Valid values for the /X2 parameter reportedly go up to 32. Some people don't seem to be able to get VTD working with their CS-based boards at all, but I don't know if this is a problem with specific boards, the configurations in question (such as failure to adjust the default input volume), other hardware (e.g., using the wrong microphones -- see the Technology section), or something else. One more tip: If you have a board that uses a CS4231 in conjunction with an OPTi 929, you may want to use the OPTi MAD16.SYS driver instead of the CS-supplied CWCONFIG.SYS to initialize the board; MAD16.SYS reportedly works better with this combination. The 1.71-level drivers do not include IBM's MPU-401 driver, so you'll need to get those separately for wavetable support. If you do this, I'm told that Warp 4.0 has a bug that prevents more than one MIDI device at a time from working, so you should not install the OPL-3 driver if you want MPU-401 MIDI. If you're looking for a CS-based board or more technical information on specific CS chipsets, Timur Tabi's CS advocacy page includes a fair amount of useful information, as well as CS cheerleading (he's now a CS employee).

Copyright © 1996, 1997, Rod Smith, rodsmith@rodsbooks.com
Return to the chipsets listing
Return to the OS/2 Soundcard Summary main page