7F20 wrote on 2024-01-12, 02:46:

Weird. I have Blood sound setup set to Sound Canvas, I expected that's what it meant. What else could it be? […]
Show full quote

Falcosoft wrote on 2024-01-12, 02:34:

7F20 wrote on 2024-01-11, 22:03:

So, I'm seeing while playing Blood that fluidsynth is missing a bunch of sounds (bank 56) using pretty much any .sf2 file I throw at it. What do I have to do to get complete sounds in Blood? Is it a situation that I would have to seek out and buy a specific SC-55 unit in order to even get the full compliment of instruments?

SC-55 (or even SC-88, SC-88Pro) does not have valid instruments at Bank 56 either. If it's not a conversion error then that Blood track is definitely not meant to be played on an SC-55.

Weird. I have Blood sound setup set to Sound Canvas, I expected that's what it meant. What else could it be?

Incidentally, I left this post here because on Page 9 of this thread, user spiffythedog notes that Blood uses "CTF for invalid variation tone bank change messages," so I thought that might be a clue to the error messages.

To be honest, I'm not sure what's it's supposedly missing at Bank 56, but it's substituting Bank 0 instead, so maybe that points to something? I wonder if it's playing the wrong stuff?

The original SC-55 plays the corresponding capital tone (Bank 0 + Program) when it receives an invalid Bank select + Program change message. This mechanism is called CTF (Capital Tone Fallback). SF2 softsynths like FluidSynth and BassMidi (and even Sound Blaster's HW SF2 synths) use the same mechanism when your SF2 soundfont is missing a preset at the requested Bank + Program combination.
So you do not miss anything. It's another question what the original intent of the composer of affected tracks was.

Falcosoft wrote on 2024-01-12, 03:17:

So you do not miss anything. It's another question what the original intent of the composer of affected tracks was.

It was suggested by user spiffythedog that it may have been composed for the AWE32. Perhaps that means that the SC version was an afterthought and they didn't really care about the CTF in bank 56?

This guy did a rip of the AWE32 version and put it on youtube: https://www.youtube.com/watch?v=tlfHoeOTO6s

As he points out in the comments, the drums really suck. I think I'll take an SC version of it, even if it's not completely accurate

Hello all
I have a sc55(no mk2) and had nearly perfect mt32 emulation.
20 days ago everything was working perfect but for the last 10 days in mt32 emulation all the music sound strange
The rom is 2.0. I tried it with different soundcards and systems but the result is the same.
Games like doom are the same but all the other lucasarts games sound strange as I mentioned before.
Is there anything I should check or do ?
I reset the gs and all the setting but nothing changed.

keropi wrote on 2023-03-23, 18:57:

IMHO you are looking too much into this for gaming... sure it is an interesting find and discussion for the sake of"retro knowledge" but in reality *any* SC-55 will be great for gaming there is no need to hunt down specific serials or revisions
when you play for example the sewer level in DN3D both breath and click instruments work fine - plus you won't really mind in the heat of the action 😁
jm2c

I like this. For years I have been banging my head against a [virtual] wall trying to determine which is the most accurate to composition intent (even started my own topic, Intended Composition Device(s), in an attempt to get further input on how this affects games of my interest), getting to the point of acquiring one of the "GS Standard" SC-55s years back only to once again be questioning whether I really needed it for Duke Nukem 3D or Master of Orion. I notice the MT-32 list on VOGONS Wiki saying certain games like Dark Forces were allegedly composed on SC-88 because of polyphony issues on older Sound Canvases and I realize there is still a lot of information out there based on anecdotal assumption rather than scientific investigation.

And yet I still cannot manage to keep from bouncing between whether I would be better served with a unit that has capital-tone fallback or higher polyphony. The substance of my question has always been the same: how much of a difference does it actually make? I also have Gravis UltraSound and Yamaha XG support, and this topic has covered how some compositions (it is confirmed elsewhere Heretic is one) were done on Sound Blaster AWE, but nothing seems to suggest it would even be noticeable in gameplay like, say, the difference with a game composed for the original MT-32 being played on a succeeding Roland LA device (as an aside: I question some of the VOGONS Wiki listings suggesting games from EA and Dynamix, other than Heart of China which IS confirmed to be affected, exhibiting issues on CM-32L or LAPC-I due to being composed for the original MT-32).

Mad_Wolf wrote on 2026-02-23, 20:00:

Hello all I have a sc55(no mk2) and had nearly perfect mt32 emulation. 20 days ago everything was working perfect but for the la […]
Show full quote

Hello all
I have a sc55(no mk2) and had nearly perfect mt32 emulation.
20 days ago everything was working perfect but for the last 10 days in mt32 emulation all the music sound strange
The rom is 2.0. I tried it with different soundcards and systems but the result is the same.
Games like doom are the same but all the other lucasarts games sound strange as I mentioned before.
Is there anything I should check or do ?
I reset the gs and all the setting but nothing changed.

My mkII has some weird behavior where the instruments are all messed up if it's booted up without the battery installed. Resetting the device from the front panel always fixes it. I would be very curious to know if your non-mkII has the same problem.

The Roland Sound Canvas SC-55mkII may be out of tune when powered on without a battery

BaronSFel001 wrote on 2026-08-06, 14:13:

I like this. For years I have been banging my head against a [virtual] wall trying to determine which is the most accurate to composition intent (even started my own topic, Intended Composition Device(s), in an attempt to get further input on how this affects games of my interest), getting to the point of acquiring one of the "GS Standard" SC-55s years back only to once again be questioning whether I really needed it for Duke Nukem 3D or Master of Orion. I notice the MT-32 list on VOGONS Wiki saying certain games like Dark Forces were allegedly composed on SC-88 because of polyphony issues on older Sound Canvases and I realize there is still a lot of information out there based on anecdotal assumption rather than scientific investigation.

Some of it comes down to personal preference but device limitations need to be considered as well. For example, the Warcraft 2 soundtrack sounds better on the early SC-55 to my ears but was composed on an SC-88 and requires more polyphony than the SC-7 and SC-55 can provide without dropping instruments.

Here is a good reference comparison for some some of the various Roland synthesizers in my testing.

Significant differences with certain MIDI files across Roland synthesizers

An analysis of Warcraft 2 MIDI music on Roland synthesizers

BaronSFel001 wrote on 2026-08-06, 14:13:

keropi wrote on 2023-03-23, 18:57:

IMHO you are looking too much into this for gaming... sure it is an interesting find and discussion for the sake of"retro knowledge" but in reality *any* SC-55 will be great for gaming there is no need to hunt down specific serials or revisions
when you play for example the sewer level in DN3D both breath and click instruments work fine - plus you won't really mind in the heat of the action 😁
jm2c

I like this. For years I have been banging my head against a [virtual] wall trying to determine which is the most accurate to composition intent (even started my own topic, Intended Composition Device(s), in an attempt to get further input on how this affects games of my interest), getting to the point of acquiring one of the "GS Standard" SC-55s years back only to once again be questioning whether I really needed it for Duke Nukem 3D or Master of Orion. I notice the MT-32 list on VOGONS Wiki saying certain games like Dark Forces were allegedly composed on SC-88 because of polyphony issues on older Sound Canvases and I realize there is still a lot of information out there based on anecdotal assumption rather than scientific investigation.

And yet I still cannot manage to keep from bouncing between whether I would be better served with a unit that has capital-tone fallback or higher polyphony. The substance of my question has always been the same: how much of a difference does it actually make? I also have Gravis UltraSound and Yamaha XG support, and this topic has covered how some compositions (it is confirmed elsewhere Heretic is one) were done on Sound Blaster AWE, but nothing seems to suggest it would even be noticeable in gameplay like, say, the difference with a game composed for the original MT-32 being played on a succeeding Roland LA device (as an aside: I question some of the VOGONS Wiki listings suggesting games from EA and Dynamix, other than Heart of China which IS confirmed to be affected, exhibiting issues on CM-32L or LAPC-I due to being composed for the original MT-32).

The CM-32L is not accurate to the MT-32, the panning of the device is different (wrong), due to the different arithmetic used by the new chips. So broadly speaking a CM-32L should not ever be used as a MT-32, although it's hardly a life or death scenario. But it also basically is a third gen MT-32, which carries on the second gen improvements, which breaks strict compatibility with custom instruments programmed on a first-gen, which sometimes take advantage of firmware and sound bugs to create instruments that exploit these. This is also not a life or death thing, but still, a factual difference.

Capital Tone Fallback for SC-55's is something people talk about a lot and only occasionally useful, not often, but it depends what games you play. For example, it makes no difference in Sierra games for example, whereas a SC-55mkII is higher quality and ALWAYS higher quality sound wise, wih the extra 4 notes of polyphony (needed for later Sierra games, for example). For some games, especially ones that use different drumkits - and often do so with incorrect program changes - CTF is basically needed to compensate for these errors. 7th Guest is an example from memory.

I'm not aware of any game written specifically for the SC-55mkII. If a game has more instruments needed for the SC-55 then it was most likely composed on an SC-88.

Gabriel Knight 1, Quest for Glory 4 are two examples. However, they may have simply been not optimised for polyphony on a SC-55 mkI. Either way, a mkII gives better performance in all Sierra games.

Spikey wrote on 2026-08-09, 16:12:

Gabriel Knight 1, Quest for Glory 4 are two examples. However, they may have simply been not optimised for polyphony on a SC-55 mkI. Either way, a mkII gives better performance in all Sierra games.

I think that this anecdote may be hearsay. For example:

Cloudschatze wrote on 2020-10-29, 01:43:

It's worth mentioning that the demo version of Quest for Glory IV, which features MIDI music later used in the full game, contains a resource bundle that can reasonably be assumed to pre-date the release and/or general availability of the SC-55mkII.

Respectfully, I think both of you are living up to my assessment of coming to conclusions based on anecdotal assumptions instead of scientific investigation. Now to be fair this is MIDI we're talking about, thus instrumentation fidelity is always going to be a relative rather than absolute factor to some extent. This is why I phrased my question the way I did: how much of a difference does it make in (this game)? I threw Duke Nukem 3D and Master of Orion out there because those are a couple titles of my interest documented as being to one extent or another altered on devices that lack capital-tone fallback; I am attempting to determine what that extent is because I presently lack the means to test for it myself.

Things have come a long way since this thread started in 2013: Munt has been perfected enough to be basically indistinguishable from the real deal, and we now have Nuked SC-55 serving as an equivalent for the part of the Sound Canvas family that pertains to our era of gaming interest. I am probing to what extent anyone has tested for differences via these digital tools, as well as how accurate they compare to behavior of the real hardware, and have always tried to account for whatever device(s) a composition was developed on as a possibility of narrowing down whatever device(s) a composition was most ideally developed for the best composition fidelity. That how-much-of-a-difference factor can be, for some of us, THE difference in weighing the decision of what hardware (and/or software) to acquire for our own gaming purposes. I know this is a question that has been asked before in general form many times elsewhere, but there is much less information regarding to specific effects to which compositions on certain games.

Of course, this could be haggling over the most minor of details...but if it is I just want to be more certain how minor those details really are, and I doubt I'm the only one interested.

Kahenraz wrote on 2026-08-10, 01:34:

Spikey wrote on 2026-08-09, 16:12:

Gabriel Knight 1, Quest for Glory 4 are two examples. However, they may have simply been not optimised for polyphony on a SC-55 mkI. Either way, a mkII gives better performance in all Sierra games.

I think that this anecdote may be hearsay. For example:

Cloudschatze wrote on 2020-10-29, 01:43:

It's worth mentioning that the demo version of Quest for Glory IV, which features MIDI music later used in the full game, contains a resource bundle that can reasonably be assumed to pre-date the release and/or general availability of the SC-55mkII.

What I said is correct, it's either that the game was composed for it, OR, that it performs better with it due to using excess polyphony with a mkI. Either way, a mkII is preferable.

BaronSFel001 wrote on 2026-08-10, 13:50:

Respectfully, I think both of you are living up to my assessment of coming to conclusions based on anecdotal assumptions instead of scientific investigation. Now to be fair this is MIDI we're talking about, thus instrumentation fidelity is always going to be a relative rather than absolute factor to some extent. This is why I phrased my question the way I did: how much of a difference does it make in (this game)? I threw Duke Nukem 3D and Master of Orion out there because those are a couple titles of my interest documented as being to one extent or another altered on devices that lack capital-tone fallback; I am attempting to determine what that extent is because I presently lack the means to test for it myself.

Things have come a long way since this thread started in 2013: Munt has been perfected enough to be basically indistinguishable from the real deal, and we now have Nuked SC-55 serving as an equivalent for the part of the Sound Canvas family that pertains to our era of gaming interest. I am probing to what extent anyone has tested for differences via these digital tools, as well as how accurate they compare to behavior of the real hardware, and have always tried to account for whatever device(s) a composition was developed on as a possibility of narrowing down whatever device(s) a composition was most ideally developed for the best composition fidelity. That how-much-of-a-difference factor can be, for some of us, THE difference in weighing the decision of what hardware (and/or software) to acquire for our own gaming purposes. I know this is a question that has been asked before in general form many times elsewhere, but there is much less information regarding to specific effects to which compositions on certain games.

Of course, this could be haggling over the most minor of details...but if it is I just want to be more certain how minor those details really are, and I doubt I'm the only one interested.

I'm talking possibilities, there isn't a way to time travel to 1993 and be a fly on the wall. All one can do is make logical assumptions based on what facts we can examine.

That's why I gave specific game examples.

CTF is relevant where game devs make errors, essentially. The most common being, they code the drum part as a non-Standard part but put the wrong number down. If I'm remembering correctly - on a mkII, this yield the Standard drums always. On a mkI, this yields the nearest neighbour, which is good because usually the error is the drum patch being off by 1.

Composition fidelity is an interesting one, because MUNT can yield identical, better or worse results than the original hardware, and a key weakness of that implementation is that it gives the user too many options which is likely to result in them making an error in setting up, or a hodgepodge of options that aren't designed to synergistically be used together. However, it CAN be made to sound identical to the hardware.
NUKED SC-55 however does not emulate the output/DAC stage of the SC-55, which leads to different, potentially better or worse audio than actual hardware, depending on your POV.

I have made a comprehensive guide for Sierra games and which devices should be used to play them before. You could consider me an authority on the subject. For non-Sierra games, I am much less well-versed.

Spikey wrote on 2026-08-13, 23:45:

CTF is relevant where game devs make errors, essentially. The most common being, they code the drum part as a non-Standard part but put the wrong number down. If I'm remembering correctly - on a mkII, this yield the Standard drums always. On a mkI, this yields the nearest neighbour, which is good because usually the error is the drum patch being off by 1.

Not quite with an mkII. If the drum set is invalid, the unit will display "NO DRUM SET" and use the last used drum set. It will not revert to patch 001 Standard. So if the last patch used was 049 Orchestra, it will keep it there. Of course, if you just reset your synthesizer, it will be on Standard anyway and will keep it there as well. I was able to confirm this behavior with a MIDI that specifies patch 050 on the drum set, which doesn't exist on the mkII.

Agent of the BSoD wrote on Yesterday, 17:21:

Spikey wrote on 2026-08-13, 23:45:

CTF is relevant where game devs make errors, essentially. The most common being, they code the drum part as a non-Standard part but put the wrong number down. If I'm remembering correctly - on a mkII, this yield the Standard drums always. On a mkI, this yields the nearest neighbour, which is good because usually the error is the drum patch being off by 1.

Not quite with an mkII. If the drum set is invalid, the unit will display "NO DRUM SET" and use the last used drum set. It will not revert to patch 001 Standard. So if the last patch used was 049 Orchestra, it will keep it there. Of course, if you just reset your synthesizer, it will be on Standard anyway and will keep it there as well. I was able to confirm this behavior with a MIDI that specifies patch 050 on the drum set, which doesn't exist on the mkII.

Well, close enough and of course one can easily change it to Standard. But thanks for clarifying the actual behaviour, that's good to know.