ih8registrations wrote:

GUS Pnp does that list but has a cleaner DAC, 8192 panning steps, 16bit samples vs emu8k which clips everything to 13bit, 32 channels where emu8k is limited to 30.

GUS PnP can NOT 'do that list' because in EMU8k almost everything is patented. It would be cheaper for Gravis to buy EMU8k chip than paying for every patent it has. AMD Interwave chip can do only lineral interpolation (as other high quality synths) - EMU8k has patented multipoint interpolation - that's why sounds cleaner than the rest when low quality samples are used (almost all old DOS games, Amiga mods).

The only company which broke Creative monopoly on EMU multipoint interpolation was Aureal with Vortex2 chip (27 point interpolation). This was the reason for Creative to sue r.i.p Aureal.

In GUS PnP you can put max 8 MB RAM in AWE32 up to 28 MB.

GUS PnP uses AMD Interwave chip which is completly different from EMU8k.
Some revisions of Interwave chips used to hiss (Gravis reaplced these cards in some, more client rights protected, parts of the World)
EMU8k has digital SPDIF output ( 2 pin connector available on all AWE cards) so clean DAC does not matter.

EMU8k has 32 channels. The 30 channels you mentioned are because Creative to cut down costs of another DAC wired OPL3 FM stereo to EMU8k stealing EMU 2 voices. In pure DOS without Creative driver (when EMU8k mixer is not yet programmed to process OPL3 stereo output) you have all 32ch for use. However this greedy Creative behaviour has advatange: with EMU8K you can apply reverb, chorus (or other effect) on OPL3 sound making cooler sound than in original (or on other cards) 😀

On EMU8k you can upload your own sound processing algorithms - some of AWE32 drivers offered Qsound stereo - which was such algorithm uploaded on EMU8k during driver init. Creative never released EMU8K DSP assembler so users could never write their own effects. However some limited effects creators appeared made by skilled programmers. Better situation is with EMU10k1 - we have such assembler which comes with linux driver tools and Prologic Surround Decoder algorithm which you can upload on EMU10k1 if you wish.

Targaff wrote:

Didn't turn up any for me :>

whoops...made a mistake. It's actually aweromgm...without the 32 😜

This debate is somewhat pointless. Yes, EMU8K might have supported all that, but none of the games really took advantage of it (ok, maybe Dungeon Keeper did). I remember games playing exactly the same on my awe32 and sb16, with the midi quality difference of course. Most games have the same setting for SB16 and AWE32 for digital audio, only midi setting differed...

zbiggy, that sort of stuff is trivial to write today. As the docs say, full SW mod tracker with all these features would perhaps need a P120MHz, now what was the slowest machine you can buy today? (Don't answer... it's a 1.2GHz VIA C3 laptop that was advertised as "super-fast CPU" on TV yesterday 😳 )

Now, it may of course be possible that the Emu8k did improve anything, but games rarely need interpolation for SFX: They play their samples at a fixed rate.

My point being: If it's for music, use something else (we already agreed on this). If it's for SFX, Emu8k is overkill and wouldn't noticeably improver things - use something else. If it's your MOD tracker: it's way easier to write a modern MOD tracker than to write Emu8k emulation just for this one software.

Feel free to prove me wrong, however, with my game SFX assumptions. I for my part didn't hear a difference with my AWE64 (basically AWE32 plus softsynth for more polyphony...).

GUS PnP can do hardware pitch shifting, mixing at 44k, volume change without clicks, chorus & reverb, and smooth panning. Touche, Creative is unique in implementing said features the Creative labs way. Your meaculpa, point interpolation should be better than linear, the trick being to realise the advantage(I have a stack of awe32/64s & a GUS. I don't hear it).

One could get the wrong impression from what you wrote that they have patents on the mere ideas rather than their implementation; they don't. If you were familier with Creative vs Aureal, you'd know it had nothing to do with interpolation, but lawsuit abuse on Creative's part. They suck, the better company & tech lost.

I'll throw some pedanticary at you as well, GUS PnP max memory was 16MB.

The first revision of interwave was recalled/replaced. You're really going to try to use that as a bullet point for yay/nay of awe vs gus? reeeaching.
Clean DAC did matter when not using SPDIF output, as was the norm.
It can have an advantage when using FM, it has the limitation when you want to use the synth, no matter how sugar coated.

Unaccessable unrealised feature; features of another chip; I'm bowled over.

🤣 I thought the "Creative labs way" was to buy out smaller competitors and relabel their products as Creative products and then put them on the shelves with poor support, only to abandon them six months later after realizing that they put out crap and need to buy out someone else's product to sell as a replacement.

I guess that was kind of a troll post, but I really haven't cared much for Creative since they entered the PCI sound card market by buying out Ensoniq.

gulikoza wrote:

whoops...made a mistake. It's actually aweromgm...without the 32 😜

😁

Here's two places to get the AWERomGM soundfont (which was made by hand and isn't quite perfect, but it sounds right most of the time):

http://seermusic.com/sounds/aweromgm.zip
http://www.maz-sound.com/archives/aweromgm.rar

If you run DOSBox under Linux, it's possible to use Timidity (a software MIDI synthesizer) as a back-end to play the MIDI music from it (with the game set to use General MIDI) - and with AWEROMGM or the official GUS patches, you can make it sound just like an AWE32 or a GUS. See my post in the "Where to get GUS patches" thread.

However, Timidity does eat up a lot of CPU time, so the 'load a Soundfont' solution is probably best if it works for you.

All the games I ever saw (admittedly that will be less than many people here) only used the AWE32 to play MIDI music and just treated it as a SB16 for sound effects - for music they would have a "Sound Blaster AWE32" option, and for sound effects they would have a "Sound Blaster 16 / AWE32" option. Since all these games also had a "General MIDI" (MPU-401) option for music, there's probably no point in specifically emulating the AWE32.

I know I'm bumping an old topic, but this is still relevant in some cases.

For instance, someone is asking me about support for the PC version of Road Rash 32 bit, which apparently utilizes the EMU8000 for MIDI playback. It also has a unique SBK file for each song it plays, and they contain overlapping instrument numbers, so they cannot easily be merged into a single bank to load on startup.

Also, if you may recall, the original Gravis Ultrasound was an expensive card when it was first launched. That didn't stop lots of random developers from utilizing it for more than just General MIDI playback. Nor did it stop anyone from developing a reasonable simulation of its features, improved by the fact that it is capable of always mixing at 44100Hz or higher, regardless of how many channels are enabled, where the original device would be limited to way less if all 32 channels were enabled.

The MT-32 was also a very expensive investment, but that didn't stop numerous game developers from utilizing it. Or developers from attempting to simulate it in software.

EDIT 2: Hmm, I see it's a Windows game. A wrapper against aweman32.dll could be made, but it would have to talk to either sfman32.dll, or talk to a software synthesizer driver.

I too am very interested in this!

kode54 wrote on 2014-04-20, 23:52:

I know I'm bumping an old topic, but this is still relevant in some cases. […]
Show full quote

I know I'm bumping an old topic, but this is still relevant in some cases.

For instance, someone is asking me about support for the PC version of Road Rash 32 bit, which apparently utilizes the EMU8000 for MIDI playback. It also has a unique SBK file for each song it plays, and they contain overlapping instrument numbers, so they cannot easily be merged into a single bank to load on startup.

Also, if you may recall, the original Gravis Ultrasound was an expensive card when it was first launched. That didn't stop lots of random developers from utilizing it for more than just General MIDI playback. Nor did it stop anyone from developing a reasonable simulation of its features, improved by the fact that it is capable of always mixing at 44100Hz or higher, regardless of how many channels are enabled, where the original device would be limited to way less if all 32 channels were enabled.

The MT-32 was also a very expensive investment, but that didn't stop numerous game developers from utilizing it. Or developers from attempting to simulate it in software.

EDIT 2: Hmm, I see it's a Windows game. A wrapper against aweman32.dll could be made, but it would have to talk to either sfman32.dll, or talk to a software synthesizer driver.

I'd implement that as a VSTI library for the VSTI driver, with a shim version of sfman32 and aweman32.

Basically, the applications that need uploadable soundfont banks make use of private api in those libraries. Just provide surrogates for the interfaces, and have your vsti library accept or load the appropriate .sf2 / .sbk files.

Queries for available RAM or EMU hardware version are just software definble responses.

It could be DLS based softsynth internally.
https://en.m.wikipedia.org/wiki/DLS_format