marxveix wrote on Yesterday, 11:33:There are many drivers, what driver version is good for ES1370 in Win9x? […]
Show full quote
There are many drivers, what driver version is good for ES1370 in Win9x?
https://archive.org/details/ensoniq-audiopci-drivers
https://archive.org/details/audio-pci-by-enso … 00-installation
https://archive.org/details/creative-sound-bl … io-pci-software
Mystery card here is cheap Ensoniq and it does not give great result by these bencmark results, was it es1370,es1371 or es1373, i dont know.
https://www.philscomputerlab.com/best-sound-c … r-socket-7.html
Super interesting and he is right in comments (e.g. ES1370 does processing in software) but there shouldn't be any processing in the test - the card should be just pulling pre-processed wavs. It would be interesting to compare Microsoft basic drivers.
If Phil installed extra demo software for the cards instead of just the INF, I'd ask why the **** do that?
I remember the Creative Labs control panel had binaural setup to output 3D sound from stereo inputs streams on stereo speakers - educational, but that is all software processing that would theoretically work with any sound card and would be best disabled for FPS. Creative's installer probably unpacks a load of other stuff (wave tables, mixers, etc.) to load into main system memory because the ES1370 is very basic. I always did wonder why Creative was investing in software for a product line they were shutting down, and I assumed at the time it was actually to lead people like me towards buying the SB Live that could do similar things in hardware.
Phil also mentioned A3D, but failed to distinguish between software A3D and hardware A3D. If you throw A3D or EAX at an ES1370 then that will translate into extra CPU cycles and more cache spills - hammering FPS.
If you are comparing 3D sound then Aureal Vortex (original) cards would be interesting because they are simpler than Aureal Vortex 2.0, and developers probably targeted the original - but we need to know settings to know what features were active in the tests.
I assume all the benchmark results reflect a vendor's default settings, but there is no record of what those defaults are. What Phil has really done is test random software stacks.
