First post, by digger
- Rank
- Oldbie
With all the wonderful new parallel audio devices that @dreamblaster has been conjuring up and cranking out lately, to what extent would it be possible to share a single parallel port among some of them?
The reason I'm asking this is because many computer models that these devices would be perfect for, such as Microchannel-based PS/2 machines and older laptops lacking sound cards, have only one parallel port and it is not easy to add an additional parallel port to those machines. Microchannel LPT cards are probably just as rare and expensive as Microchannel sound cards. As for older laptops, perhaps there are PCMCIA cards out there that would add an additional parallel port to such machines, but even if that would work, that might be too big of a hassle.
So assuming a non-ISA system with only one parallel port, that gives us the problem of having to choose between music and digital effects, when ideally, we'd like to have both in our DOS games, just like a fully-fledged sound card would provide.
Now combining a Covox Speech Thing with an OPL2LPT on a single parallel port would probably be tricky methinks, since a Covox requires high precision timing when sending samples to it and sending OPL2 control signals to it in between would probably disrupt that. But I was thinking, since the Disney Sound Source has a FIFO buffer, perhaps that would allow OPL2 signals to be mixed with DSS output data more easily, since a hypothetical combined device could simply tell the computer that the FIFO buffer is currently full whenever it's in the middle of processing OPL2 data. Would that work? And could it be done in a combined device, even while preserving native DSS compatibility?
Additionally, perhaps a "DSS Plus" or "DSS 2.0" standard could be developed for such a combined device, which would allow the output bitrate to be adjusted from 7 kHz to 11 kHz (and perhaps also 22 kHz), which would perhaps make Sound Blaster emulation in software easier for games that do not support DSS out of the box. The adlipt TSR could then perhaps be extended to provide such Sound Blaster emulation in addition to rerouting OPL2 signals, all on a single parallel port device. And perhaps the FIFO buffer could be made a bit larger to allow for better interleaving of OPL2LPT signals with DSS output streams and possibly less strain on slower CPUs?
Alternatively, perhaps an "OPL2COM" device that would plug into a serial port could provide a solution here, since pretty much any older machine that has a parallel port usually also has at least one and often two serial ports. Such a device could then be used together with a CVX4 or some kind of DSS-clone in the parallel port.
Of course, I'm just thinking out loud here. Perhaps none of this makes much sense. 🤣 What do you all think?