Great to see this topic up and running again! Thanks to the guy who told me.

Thanks to tikalat for the drivers and NRS for a lot of the legwork and info. If there was a position for 'guy who knows the most about Sierra coding', it would be you.

OK, some info from various posts.

Also: If the "Making of" video on the Gabriel Knight 1 CD-ROM is any indication, the target device was not any Sound Canvas, but some Proteus-like module. SC-55 will definitely have polyphony dropouts during some songs.

The CD digital pieces (Intro, Closing, Nightmare introduction) use a EMU Proteus 2000 and possibly some other synths. The General MIDI score was not composed using a Sound Canvas, but was translated to Sound Canvas for the in-game soundtrack by other Sierra composers at the time (Robert Holmes told me he had no idea about MIDI pretty much up until GK1).
So it is a SC-55 score, but it was never composed *for that*, which is why some tracks sound really tailored for a SC-55 and others don't.

I don't think there are any polyphony issues on the GK1 score, maybe a track or two. It's also possible the score was done using a SC-55 mk II, like QFG4 at the same time.

1) Does it matter if I change Bank LSB (cc #32) before Bank MSB (cc #0)?

CC 32 is used to change instrument patch on an XG device, CC 0 is used to change an instrument patch on a GS device. Neither should be used in a Sierra GM score.

What are the factory reset values for reverb, chorus?
- VSC uses 40,0 (decimal)

This is what Sierra GM games use (40,0), there's a GM reset that is sent at the loading of the game before the Sierra logo.
Personally, I use what the enhanced soundtracks use, usually 90,33 or 90,44, but that's a matter of taste.

I'm wondering if that's why games sound somewhat funny to me using BassMidi. VSC is more natural to my ears but some users don't have VSC and I'm curious if the GM / GS reset values are causing a difference between the 2 soft synths (more blur, more richness)

I've never understood if these softsynths take SysEx properly, although VSC *should*.

That's for a GENMIDI.DRV that translates Roland MT-32 tracks to General MIDI, because Modulation is deeper on the former. I suppose you could do without the /2 for original General MIDI tracks.

NRS: Assuming this is for games like PQ1VGA, LB2, QFG1VGA that have a "General MIDI" option but the MIDI SND files are all MT-32? Remap driver?

Wouldn't this mess up the sound in a GM-specific soundtrack, or am I missing something?

Would you recommend using an SC-55 soundfont for GK1 or something else non-hardware?

Oasis: A SC-55 or probably SC-55 mk-II is the best, but yeah, something like VSC or proper SC soundfont (not generic GM soundfont, YMMV with that).

GK1 […]
Show full quote

GK1

General MIDI (DOS)
- Sfx now use full channel 1-16.
- Send modulation to device.
- Channel volume fix for channel 1-16.

Tika, for my info, does a lot of the GM drivers not send modulation info?
Also, what does 'channel volume fix' do? 😀

Thanks again for all that you and NRS do.

Spikey wrote:

The CD digital pieces (Intro, Closing, Nightmare introduction) use a EMU Proteus 2000 and possibly some other synths.

Intro sounds completely different in 5.wav on the CD than in the Making of video. The latter almost sounds as if it had been recorded from the first Wave Blaster daughterboard (CT1900).

Spikey wrote:

This is what Sierra GM games use (40,0), there's a GM reset that is sent at the loading of the game before the Sierra logo.

Sierra games send GS reset, not GM reset.

Spikey wrote:

I don't think there are any polyphony issues on the GK1 score, maybe a track or two.

Check out Closing (990.snd) after 0'40'' or Gabriel's Shop (210.snd) after 0'30''.

Spikey wrote:

Wouldn't this mess up the sound in a GM-specific soundtrack

I suppose so, although it would not be much of a mess-up given that Sierra's original GENMIDI.DRV ignores modulation completely. My GENMIDI.DRV was written as part of a GMBLAST.DRV for older games, which is why I corrected the modulation depth between MT-32 and GM. Only later did I notice that it produced different drum sounds in Space Quest V.

Hey! Good to see you around Spikey. 😀

I've never understood if these softsynths take SysEx properly, although VSC *should*.

I've discovered that:
- Roland VSC player = sysex okay
- Roland VST / DXi = no sysex (none!! needs manual reset / switch)
- Edirol VST = no sysex (none!! needs manual reset / switch)
- CoolSoft = GM reset puts reverb to 0, otherwise seems okay
- BASSMIDI = sysex okay (seems normal)

The current SC-55 sf2 soundfonts have some volume problems with sfx in particular. Or the ADSR is too long somewhere (Doom2 Map1). Or embedded too much default minimum reverb (imo). So cc91 (40) adds even -more- reverb to some of them. Makes it all soup mixy + over-pungent (enough) sometimes.

CC 32 is used to change instrument patch on an XG device, CC 0 is used to change an instrument patch on a GS device. Neither should be used in a Sierra GM score.

Until someone writes a custom Sierra GS score with cc0 usage (although the driver support would need to be added).

Thanks, I'd better be careful of XG MIDI in GS mode.

There's issues with SC-88 and capital tone fallback with Might and Magic IV + V. TURRICAN0 solution is to write cc0 (0) + cc32 (0). Wasn't sure if writing them in reverse order would matter (save 2-3 precious asm bytes for rewritten driver). Then again, I could just cut out cc32 also. Hmm...
Xeen and Roland synths

Personally, I use what the enhanced soundtracks use, usually 90,33 or 90,44, but that's a matter of taste.

On real hw, do you have to apply the global reverb, chorus override setting after the GS reset is sent?

What would happen though when a game like TIM2 tweaks the reverb, chorus mid-track? Does it add / subtract to the global value?

Tika, for my info, does a lot of the GM drivers not send modulation info?
Also, what does 'channel volume fix' do? Happy

I haven't seen 1 Sierra GM DOS driver send modulation. All the WIN ones do though by default.

When you change the volume sliders in options, it only changes ch2-9 volume. ch1 + ch10-16 aren't affected by default driver. Which bothered me so I needed that change. 😉

Thanks to tikalat for the drivers and NRS for a lot of the legwork and info. If there was a position for 'guy who knows the most about Sierra coding', it would be you.

NRS did all the tough detective work - channel remap, driver structure, compression, drum fallback, sierra snd format, misc. And I'm still trying to sharpen my skills.

Sometimes I don't know why I care enough to be annoyed by some minor design sound flaws. Now I find the Miles Audio drivers only remap to ch2-9 also and cut out bgm notes for sfx. Ugh. 😐

Great to see this topic up and running again! Thanks to the guy who told me.

I'm not sure there's much else to do with Sierra / Dynamix. Misc games like Twinion are MT-32. Outpost + 3D Ultra Pinball use SMF mid. I'm open to suggestions of course. :haha:

But someone like NRS + Spikey manage to pull out new expert-level trivia though. 😁

I'm looking to start a new thread soon that covers the timeline of games that use GS or 55-composed (with very brief notations). There's more 1992 releases than I thought. 😮

NRS:

Intro sounds completely different in 5.wav on the CD than in the Making of video. The latter almost sounds as if it had been recorded from the first Wave Blaster daughterboard (CT1900).

Oh, OK. It's been a LONG time since I saw that video..

I'll ask RH and see if he remembers.

Wow, it's just the MIDI file, and there's other MIDI's in the video also. Straight from the game. But not a Sound Canvas.. Weird.

I doubt that indicates the game was composed for a Proteus/etc though, I've had Robert and other-era Sierra composers tell me the game was translated from Robert's music to Sound Canvas GM (and Robert hated that, let me tell you).

EDIT: Robert says, first "probably a sound canvas", then when I mention this is not the case,
"we didn't have anything in the shop but Sound canvas, sound Blasters, and a few other Roland boxes.The final may not have been done yet."

I theories maybe it's a wavetable General MIDI recording from a Sound Blaster AWE sound card. The pizzicato from St. George's Book Shop makes me think that.

Either way, Robert adds "sounds like crap 😀". Thanks Robert 😉

He finally adds "Seems like we had some Roland 1680 racks around, I wonder if it was one of those.". Not familiar with that synth myself.
"Seibert had rows of those", apparently.

Sierra games send GS reset, not GM reset.

Well, either way 😜

Check out Closing (990.snd) after 0'40'' or Gabriel's Shop (210.snd) after 0'30''.

Yeah, I would guess the SC-55mkII with QFG4 being the same at the same time, and GK1 was pretty heavy on the tech side. Thanks for the examples. I've recorded the soundtrack but I didn't remember those specifics.

I suppose so, although it would not be much of a mess-up given that Sierra's original GENMIDI.DRV ignores modulation completely. My GENMIDI.DRV was written as part of a GMBLAST.DRV for older games, which is why I corrected the modulation depth between MT-32 and GM. Only later did I notice that it produced different drum sounds in Space Quest V.

Point well made Sir 😀

Great find, all these extra notes and modulation and such.

Tikalat: Good to be seen!

Quick question, you mentioned a few pages back that Dynamix games, Rise, Willy, China had possible CM-32L percussion? Seems unlikely to me, but did we figure it out?

The current SC-55 sf2 soundfonts have some volume problems with sfx in particular. Or the ADSR is too long somewhere (Doom2 Map1). Or embedded too much default minimum reverb (imo). So cc91 (40) adds even -more- reverb to some of them. Makes it all soup mixy + over-pungent (enough) sometimes.

*I* would say this is a trait of most soundfonts in general, although there are obviously some great ones these days.

Until someone writes a custom Sierra GS score with cc0 usage (although the driver support would need to be added).

Thanks, I'd better be careful of XG MIDI in GS mode.

Well, neither of those two are a big concern, there's no CC0 bank changes in any Sierra score that I've ever seen. Or was there one, once? Darn, I forget. I feel like there may have been some weird exception somewhere..

On real hw, do you have to apply the global reverb, chorus override setting after the GS reset is sent?

Yeah, otherwise it gets reset! 😉

I'm lazy, what I do on my SC-55mkII which has a soft on/off setting, it's either on or on standby, is I send a MIDI with reverb/chorus settings I want, and then turn it 'off'(standby), then load the game, then when the logo is up turn it on. Wouldn't work on my other modules 😀

What would happen though when a game like TIM2 tweaks the reverb, chorus mid-track? Does it add / subtract to the global value?

There's no 'global value', only individual levels modified by the two controllers. I mean, there is a global value, but it's never modified in-game.
If the reverb,chorus is 40, 0 and then the controller is defined 60,33 that's what it becomes, it's not 40+60 and 0+33.

I haven't seen 1 Sierra GM DOS driver send modulation. All the WIN ones do though by default.
When you change the volume sliders in options, it only changes ch2-9 volume. ch1 + ch10-16 aren't affected by default driver. Which bothered me so I needed that change.

Great info, did not know this! I always wondered when I saw modulation in the resources. Great reason to use resource files for soundtracks and not in-game recordings, for GM games.

Good to see about the sliders too, that always annoyed me as well!

NRS did all the tough detective work - channel remap, driver structure, compression, drum fallback, sierra snd format, misc. And I'm still trying to sharpen my skills.

Yeah, sorry tika, but my comment was aimed at NRS. 😉 He's number 1 on the home team, although he will never want that title. But you certainly are number 2 now!

Sometimes I don't know why I care enough to be annoyed by some minor design sound flaws. Now I find the Miles Audio drivers only remap to ch2-9 also and cut out bgm notes for sfx. Ugh.

You're revitalised music in Sierra GM scores! Now they can play as composer intended, which is a massive deal! Great job.

The Miles Audio thing sounds like the SQ5 score. Ick!

I'm not sure there's much else to do with Sierra / Dynamix. Misc games like Twinion are MT-32. Outpost + 3D Ultra Pinball use SMF mid. I'm open to suggestions of course. :haha:

Heh! I'm sure I can find something one of these days. I seem to remember something off about QFG4 and other Windows Sierra MIDI games, and maybe KQ5 WIN also.

But someone like NRS + Spikey manage to pull out new expert-level trivia though.

Yeah, we're hardcore.. nerds 😜 But we do have fun with our little debates. I always learn something, so it's worth it just for that..

Great find, all these extra notes and modulation and such.

Now I've got a question. Other than SQ5, how many others' modulation are MT-32 specific? There's enough of these hacked up drivers with modulation fixes... but are they right? Or need the dreaded cc1 / 2?

Quick question, you mentioned a few pages back that Dynamix games, Rise, Willy, China had possible CM-32L percussion? Seems unlikely to me, but did we figure it out?

Oh that. Probably some lazy (incorrect warning) parser log of mine - probably used custom timbres (iirc) but still needs in-game checking to strike them off.

Never got around to that (totally forgot).

there's no CC0 bank changes in any Sierra score that I've ever seen. Or was there one, once? Darn, I forget. I feel like there may have been some weird exception somewhere..

I haven't seen any. Just indirectly fishing out information sometimes from the pros. :hehe:

I'm lazy, what I do on my SC-55mkII which has a soft on/off setting, it's either on or on standby, is I send a MIDI with reverb/chorus settings I want, and then turn it 'off'(standby), then load the game, then when the logo is up turn it on. Wouldn't work on my other modules Happy

That's a cute method. 😀:)

It dawned on me that you could try getting one of those... uhm... VST thingies or MIDI message monitors (can't remember what they're called). If it sends a reset, you can edit the message or tack on to it (or so I hear).

If the reverb,chorus is 40, 0 and then the controller is defined 60,33 that's what it becomes, it's not 40+60 and 0+33.

Okay. That makes sense.

Yeah, sorry tika, but my comment was aimed at NRS.

That's good! I don't like the spotlight anyway (well.. when having others judge parts of yourself that you disapprove of). :p

The Miles Audio thing sounds like the SQ5 score. Ick!

To be fair to John Miles, it's a temporary channel swap. He uses the lowest polyphony channel for the sfx. Then restores back to the original music data when it's over (except for drum kits - POP2 sfx error).

When you hear it drop out though, it's a buggery boo sometimes. Or the Miles driver goofs up reverb or some value - it just sounds wrong somehow. I did have ~3 channels not restore when playing Blackthorne once for about 10-20 seconds. That was bad.

Heh! I'm sure I can find something one of these days. I seem to remember something off about QFG4 and other Windows Sierra MIDI games, and maybe KQ5 WIN also.

The Windows drivers use smaller channel reserve (esp. ch10) + polyphony (~16?) limits. KQ6 WIN - go to the castle front. You'll hear the lack of some percussion until you stop sending cc4B. Then it comes back. I thought I patched up some of those... can't find them though.

And the WIN drivers only have room for storing 15 MIDI channel memory-wise (DOS = 16). So I couldn't use the full 16 channels (correctly) if I tried without memory spill overwrites.

edit: Phew! That's a mouthful.

Me again. I stumbled on a curiosity in LSL6 while recording the soundtrack. I've noticed that the track named "Samba de Hooters" has two distinct versions, one for MT-32 and GM, and a different one for AdLib/SB (I'll attach the recordings)
They have two different melodies at the bridge part of the song, although they sound identical in the other parts.
So I was wondering, is this actually intentional or is it a mistake on the developer's side?
I wasn't sure where to ask, and since you guys here seem to have a hang at disembowelling the games I thought I could drop the question here ^^

Me again. I stumbled on a curiosity in LSL6 while recording the soundtrack. I've noticed that the track named "Samba de Hooters" […]
Show full quote

Me again. I stumbled on a curiosity in LSL6 while recording the soundtrack. I've noticed that the track named "Samba de Hooters" has two distinct versions, one for MT-32 and GM, and a different one for AdLib/SB (I'll attach the recordings)
They have two different melodies at the bridge part of the song, although they sound identical in the other parts.
So I was wondering, is this actually intentional or is it a mistake on the developer's side?
I wasn't sure where to ask, and since you guys here seem to have a hang at disembowelling the games I thought I could drop the question here ^^

Very interesting, thanks for posting this. The General MIDI original is designed to have a loop at about 1:20, where the song starts again, but the flute part is changed to the second melody. There are several tracks that do this in LSL6 (e.g. the Blues Bar jazz and rock tracks), I wonder if those do the same on the Adlib tracks.

My guess is that the Adlib version lacked the polyphony (simultaneous note playing ability) to play all those flute notes in the crescendo of the song (multiple notes being played often), so they had to simplify it a little bit. But it really does sound different.

The composer, Dan Kehler, had nothing to do with the Adlib conversions of the soundtracks, those were usually given to an 'apprentice' in the music dept. at Sierra, so it wouldn't be out of the question for the music to get altered in the name of it working with the hardware.

I will reply to the rest soon. 😀

Guys,
Thanks you for your hard work and dedication.
This is a great ongoing effort.
The only downside is that your drivers are scattered all over this thread.
I started to collect them in my SkyDrive.
You can use this link to view it:
http://sdrv.ms/1aOkEpl

NRS/tikalat/Other interested parties: You can contact me in a PM for a write access, and I'd love if you could sort this out, throw out the old ones, rename appropriately. Maybe make sure all the archives has a good readme inside, etc.

sort this out, throw out the old ones, rename appropriately. Maybe make sure all the archives has a good readme inside, etc.

Most of it is my mess. So I need to clean house.

NewRisingSun
- Betrayal at Krondor (DOS)
- King's Quest I: Quest for the Crown (SCI) (DOS)
- Quarky & Quaysoo's Turbo Science (DOS)
- Quest for Glory II: Trial by Fire (DOS)
- Rise of the Dragon (DOS)
- Space Quest V: The Next Mutation (DOS)
- The Adventures of Willy Beamish (DOS)
- The Incredible Machine (DOS)

Tikalat
- Betrayal at Krondor (DOS) - NRS addendum - GM driver
- Freddy Pharkas: Frontier Pharmacist (DOS,WIN)
- Gabriel Knight: Sins of the Fathers (DOS,WIN)
- Jones in the Fast Lane (DOS CD)
- King's Quest VI: Heir Today, Gone Tomorrow (DOS,WIN)
- King's Quest VII: The Princeless Bride (DOS,WIN)
- Leisure Suit Larry 6: Shape Up or Slip Out! (DOS,WIN)
- Pepper's Adventures in Time (DOS,WIN)
- Police Quest IV: Open Season (DOS,WIN)
- Quarky & Quaysoo's Turbo Science (DOS) - NRS addendum - GM driver
- Quest for Glory III: Wages of War (DOS)
- Quest for Glory IV: Shadows of Darkness (DOS,WIN)
- Rise of the Dragon (DOS) - NRS addendum - alternate exe
- Slater & Charlie Go Camping (DOS, WIN)
- Sorcerian (DOS)
- Space Quest III: Pirates of Pestulon (DOS)
- Space Quest IV: Roger Wilco and the Time Rippers (DOS) - NRS addendum - subtitles
- Space Quest V: The Next Mutation (DOS) - NRS addendum - alternate exe
- The Even More! Incredible Machine (DOS)
- The Incredible Machine (DOS) - NRS addendum - GM driver
- The Incredible Machine 2 (DOS) + NRS addendum - pitch bend

List completed.

Thanks tikalat.
Please check your private messages.

Now I've got a question. Other than SQ5, how many others' modulation are MT-32 specific? There's enough of these hacked up drivers with modulation fixes... but are they right? Or need the dreaded cc1 / 2?

Well, I recently noticed in LSL6, which I'm secretly working on, actually has controller 78 (vibrato delay) messages, and IIRC, many Sierra GM games also do. I don't believe that controller is sent, along with the modulation. I think both my and Tom's (QuestStudios') MIDI files prove this, with none of our GM recordings ever capturing one of those cc78's (again, IIRC).

I don't know if that counts as a GS program change, I think not, but it's still stopping the gamer hearing the MIDI properly.

Oh that. Probably some lazy (incorrect warning) parser log of mine - probably used custom timbres (iirc) but still needs in-game checking to strike them off.

Never got around to that (totally forgot).

How could I easily establish that? Hear the extracted MIDI and then go in-game to a prerecorded save point and hear that? I assume I'd hear a difference in the percussion?

I haven't seen any. Just indirectly fishing out information sometimes from the pros. :hehe:

For some reason I feel like there is maybe one time it happens, but it could just be a hazy recall or wishful thinking. 😜

That's a cute method.

It dawned on me that you could try getting one of those... uhm... VST thingies or MIDI message monitors (can't remember what they're called). If it sends a reset, you can edit the message or tack on to it (or so I hear).

A lazy gamer method!
Huh, the other idea isn't bad either.

That's good! I don't like the spotlight anyway (well.. when having others judge parts of yourself that you disapprove of). :p

I think you're overplaying this whole 'obsessed nerd' thing, you're just bugfixing, and you have skills to fix stuff relatively easily so that's a good deal in my book.

To be fair to John Miles, it's a temporary channel swap. He uses the lowest polyphony channel for the sfx. Then restores back to the original music data when it's over (except for drum kits - POP2 sfx error).

When you hear it drop out though, it's a buggery boo sometimes. Or the Miles driver goofs up reverb or some value - it just sounds wrong somehow. I did have ~3 channels not restore when playing Blackthorne once for about 10-20 seconds. That was bad.

That sounds.. interesting. Surely that would have been figured out in beta testing! Although I would have thought the same for SQ5's weird sound issues also.

The Windows drivers use smaller channel reserve (esp. ch10) + polyphony (~16?) limits. KQ6 WIN - go to the castle front. You'll hear the lack of some percussion until you stop sending cc4B. Then it comes back. I thought I patched up some of those... can't find them though.

And the WIN drivers only have room for storing 15 MIDI channel memory-wise (DOS = 16). So I couldn't use the full 16 channels (correctly) if I tried without memory spill overwrites.

Well, there shouldn't be too many examples where you HAVE to use all 16 MIDI channels for the particular area with occasional SFX to play back correctly, with no polyphony issues, etc.

Don't forget some Sierra GM games were actually composed with the Roland SC-55mkII, not the SC-55, and are composed for extra polyphony again (eg QFG4), so would be hard to get right for GM in the Windows version.

Playing QFG4 I had some other issue though, back in the day, seemed like notes cut off or something? Sorry I'm not being more descriptive, but I learned not to playback Sierra games under Windows a long time ago..

edit: Phew! That's a mouthful.

Yes it is, but I'm having fun. 😀

Well, I recently noticed in LSL6, which I'm secretly working on, actually has controller 78 (vibrato delay) messages, and IIRC, many Sierra GM games also do. I don't believe that controller is sent, along with the modulation. I think both my and Tom's (QuestStudios') MIDI files prove this, with none of our GM recordings ever capturing one of those cc78's (again, IIRC).

I don't know if that counts as a GS program change, I think not, but it's still stopping the gamer hearing the MIDI properly.

Cool. Surprised I missed that with my logger. Gotta fix that. Mwaha.

Probably isn't sent in DOS or WIN drivers. Nice catch.

How could I easily establish that? Hear the extracted MIDI and then go in-game to a prerecorded save point and hear that? I assume I'd hear a difference in the percussion?

That sounds like a good idea. Otherwise you'd need a MUNT logger, which I don't really have off-hand and doesn't really work.

For some reason I feel like there is maybe one time it happens, but it could just be a hazy recall or wishful thinking. Sticking Tongue Out

The human mind is better than the machine program (my stuff anyway). Maybe it'll happen.. didn't Sierra publish other games too? Battle Bugs, Alien Legacy?

(well.. when having others judge parts of yourself that you disapprove of). :p

Ahem. This was an indirect sly reference to my physical appearance. :cheeky:

'obsessed nerd' .. .. […]
Show full quote

'obsessed nerd'
..
..

Well, there shouldn't be too many examples where you HAVE to use all 16 MIDI channels for the particular area with occasional SFX to play back correctly, with no polyphony issues, etc.

Don't forget some Sierra GM games were actually composed with the Roland SC-55mkII, not the SC-55, and are composed for extra polyphony again (eg QFG4), so would be hard to get right for GM in the Windows version.

I easily confuse nerd, geek, twit, dweeb, dork. I'm sure they all mean something different. It doesn't help having all these titles stuck in my head over the decades (falsely too). Buffoon and fool I do understand. 🤣

You know more about music than I'll ever understand in my lifetime. Some of that polyphony stuff is out of my league.

Hmm... alright.

I'll come clean and be more genuine and stop pretending on a holistic + learned ignoramus level. Be good for myself. More reporting and less exaggerating. More natural curmudgeony that cycles with the moon.. 😀
</playful>

<mood swing>
<brooding>
Polyphony - I want a safety net. Even if it's not used.

you're just bugfixing, and you have skills to fix stuff relatively easily so that's a good deal in my book.

Okay. Understand.

Playing QFG4 I had some other issue though, back in the day, seemed like notes cut off or something? Sorry I'm not being more descriptive, but I learned not to playback Sierra games under Windows a long time ago..

Tweak what I know. Hope it works. Give up if still broken. 😀

OmerMor wrote:

I'd love if you could sort this out, throw out the old ones, rename appropriately. Maybe make sure all the archives has a good readme inside

I'm not your secretary.

Spikey wrote:

Well, I recently noticed in LSL6, which I'm secretly working on, actually has controller 78 (vibrato delay) messages,

Controller 78 is only defined as such in General MIDI Level 2, published in 1999. In Sierra MIDI files, controllers in the 70-79 range control things like voice allocation.

NRS: So controller 78 values from 0 to 1 (which is all I ever see, or 1 to 0) changes polyphony/partial reserve, or does it change what MIDI channel is being used?

EDIT: I think I see what it does, correct me if I'm wrong.
In tracks like Larry 6's blues bar tracks, where the song loops at a point but one track plays a different lead melody (same loop with different lead), controller 78 is used to control which part of the lead is playing. So it's like an on/off switch that determines whether part a or part b is used with the rest of the loop. Right?

Tika: Let's not 'fix' the controller 78 stuff just yet, as NRS has pointed out, I may have missed the point bigtime about what it was for. I think it is sent anyway.

OK, I'll check the Dynamix stuff. Should be interesting.

I don't know much about music at all.. I just know in QFG4, the SC-55 drops notes (e.g. The Castle), with a mkII it does not. NRS says GK1 does the same thing with some tracks (e.g. Closing themes, St George's Book Shop). So we know polyphony (how many simultaneous notes a device can playback at once) is an issue with the SC-55 vs mkII.

qfg4 win driver notes
- bump polyphony from 16 -> 32
- remove channel reserve (notes cutting)
- ??? volume stays at constant 127 max. see title logo dos vs win ???
- (note self) hacked driver currently Ext. MIDI only (look into adding base + ext flag hack)

see if I can do more about this before updating qfg4 dos + win combo driver pack.

other news:
checked Epyx (Sierra) Battle Bugs - FM mono only. XMIDI files.

Quest for Glory IV windows driver fixed
Re: Sierra/Dynamix sound driver hacking
--> Re: Sierra/Dynamix sound driver hacking

Game uses lots of FM pan-pot. Windows driver = buggy (send cc7 volume, fix note velocity, polyphony, volume slider).
- driver adjusts note velocity using volume slider = bad (example attached)

DOS + WIN drivers updated (KQ6 + PQ4)
- Re: Sierra/Dynamix sound driver hacking

Misc cleanups

Pepper's Adventures in Time updated
- DOS + WIN
- DOS sends drum kits (CM-64) - used at title screen, clock breaking
- WIN already sent drum kits.
- SB16 DAC DOS driver
- MIDI trivia

Re: Sierra/Dynamix sound driver hacking

Hi guys,

Im not sure if I should post this here or not but as it is related to the work you have been doing with some of these Sierra games (BTW awesome job!) it seemed the best place.

I was wondering if any of you had considered making combined MT-32 music/SB fx drivers for non-sierra games like Wing Commander 1? I use it as an example as the MT-32 music is much better than the SB music but (and I admit its just my opinion) the FX for things like shield hits, afterburners and laser blasts leave a bit to be desired when using MT-32.

I confess a complete lack of knowledge when it comes to drivers, so if this question is incredibly naive I apologize. Either way you guys are doing an amazing job on these Sierra drivers and I along with some of my friends are extremely grateful for your hard work.

As the user above, I was wondering if any of you kind fellows would like to work on the driver for Lemmings 3D?
The LAPC-1 option in setup doesn't output anything for some weird reason. All that work are the GeneralMidi/SoundCanvas and AdLib options, and that doesn't really reproduce properly how the music would sound on a MT-32 or LAPC-1.