VOGONS


88Emu is out (Roland SC-88/Pro/50 Emulator)

Topic actions

Reply 100 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
giulio wrote on 2026-11-04, 02:28:

For the SC-88/88pro, dumping the socketed eprom is enough

thanks, actually AFAIT, reading an old thread in this forum:

About Roland Virtual Sound Canvas 3

it seems SC-88 Wave ROMs were released even here and in that thread also the de-scrambling sequence was found. does, the 88Emu need scrambled or de-scrambled dumps?

also, in that same old thread here:

About Roland Virtual Sound Canvas 3

combining with the few documented information here:

https://github.com/dsp56300/gearmulator/blob/ … .md#rom-loading

it seems some (or the same?) Wave binaries are found in the "SCCore.dll". anyway, as a newbie, I still think a little better documentation on that part is needed. I have SC-88 and SC-88Pro - I guess I will dump all I can from their PCB and try to build myself some "map" what is what.

giulio wrote on 2026-11-04, 02:28:

For the SC-8820 and SC-8850 you will need to load a firmware upgrade that dumps the internal roms and the wave rom. It works only though on early units that don't use mask rom. For the late ones without the updater you'll probably need to do some microsoldering.

as far as I imagine it such approach uses some kind of Trojan-FW that makes the dump possible and I presume that Trojan-FW is not publicly available. so, I guess, that's out of questions. I have 2x (or 3x, don't remember exactly) 8820 units, need to check my storage room - at the moment cannot tell if anyone of them is using not masked-rom version of the mcu or not. I also have 1x SC-D70...

Reply 101 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie

ok, kind of got it running with SC-88, just to answer to some of my own questions that I was not able initially to deduct from what I read:

Wave binaries are fine scrambled, it seems Control ROM v1.01 is supported, but V1.02 is not recognized/supported.

also, the Control you can dump from the device you own with chip programmer, the Waves you can dump using the following tool:

https://github.com/shingo45endo/sc-mcu-wave-romdumper

so, it's not that hard when you know it, but I didn't know it.

Reply 102 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
giulio wrote on 2026-09-22, 02:47:

I attached this time the work we did on the CM-32P reverb chip, where I spent multiple weeks of my life tracing every single wire and annotating every single standard cell type manually to get their logic functions out.

totally off-topic, but please, could you look at the following new IC decap:

Roland MPU 401 Decap

and please tell us there your opinion as someone apparently very experienced - is it good enough for reversing? what tool you used for the tracing?

Reply 103 of 121, by Shiftyy

User metadata
Rank Newbie
Rank
Newbie

Would this be able to run on a RPi compute module 4 (CM4 Lite), making it possible to be used on a real hardware via the wavetable WP32 McCake, or is the CM4 Lite too slow?

Reply 104 of 121, by Karmeck

User metadata
Rank Member
Rank
Member
Shiftyy wrote on 2026-11-04, 04:27:

Would this be able to run on a RPi compute module 4 (CM4 Lite), making it possible to be used on a real hardware via the wavetable WP32 McCake, or is the CM4 Lite too slow?

(mostly) What you are looking for is here:

https://misterfpga.org/viewtopic.php?t=7858&start=30

Not tested by me.

Reply 105 of 121, by Karmeck

User metadata
Rank Member
Rank
Member
giulio wrote on 2026-11-04, 02:28:

For the SC-88/88pro, dumping the socketed eprom is enough
For the SC-8820 and SC-8850 you will need to load a firmware upgrade that dumps the internal roms and the wave rom. It works only though on early units that don't use mask rom. For the late ones without the updater you'll probably need to do some microsoldering.

Closest you can get is a post on the issues tracker for Nuked-SC55.

I "documented" my own dumping of the Roland SCC-1A.

The dumping of scc-1A / You can make a differens / Farwell.

You find a link to the relevant issue post there.

Reply 106 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
Karmeck wrote on 2026-11-04, 06:56:

I "documented" my own dumping of the Roland SCC-1A.

The dumping of scc-1A / You can make a differens / Farwell.

You find a link to the relevant issue post there.

thank you, interesting information! so, I own SC-88 and SC-88Pro and for them the unaffiliated project:

https://shingo45endo.github.io/sc-mcu-wave-romdumper/

is sufficient to backup your legally owned Roland hardware - it still requires to open the device, pull out the Control ROM chip and patch it using external chip programmer, but that is totally OK and I see nothing questionable about it. What I don't fully get through is why the Authors of 88Emu did not document that and create additional hurdles for newbies such way. especially when the documented:

https://shingo45endo.github.io/sc55mk2-ctf-patcher/
https://github.com/shingo45endo/sc55mk2-ctf-patcher

about 10 lines above the following line:

https://github.com/dsp56300/gearmulator/blob/ … d-midi-playback

in their documentation, but somehow documenting:

https://shingo45endo.github.io/sc-mcu-wave-romdumper/
https://github.com/shingo45endo/sc-mcu-wave-romdumper

is not OK to be put in the documentation. anyway, I guess it's their decision, but at least I don't see the motivation behind it.

Reply 107 of 121, by mckuhei

User metadata
Rank Newbie
Rank
Newbie
costa wrote on 2026-11-04, 02:50:

as far as I imagine it such approach uses some kind of Trojan-FW that makes the dump possible and I presume that Trojan-FW is not publicly available. so, I guess, that's out of questions. I have 2x (or 3x, don't remember exactly) 8820 units, need to check my storage room - at the moment cannot tell if anyone of them is using not masked-rom version of the mcu or not. I also have 1x SC-D70...

I can provide modified firmware that allows for dumping the MCU firmware; however, I cannot release it publicly. If you need it, please send me a private message specifying which firmware you require, along with your contact information, so that I can send the modified file to you.
By the way, I've already dumped the MCU firmware for the SC-8820. I haven't started working on the firmware for the SC-D70 yet, because my SC-D70 is broken 🙁

Devices: Roland ED SC-D70, Roland SK-88 Pro, Yamaha MU2000, Edirol UA-101

Reply 108 of 121, by Karmeck

User metadata
Rank Member
Rank
Member
mckuhei wrote on 2026-11-04, 13:36:
costa wrote on 2026-11-04, 02:50:

as far as I imagine it such approach uses some kind of Trojan-FW that makes the dump possible and I presume that Trojan-FW is not publicly available. so, I guess, that's out of questions. I have 2x (or 3x, don't remember exactly) 8820 units, need to check my storage room - at the moment cannot tell if anyone of them is using not masked-rom version of the mcu or not. I also have 1x SC-D70...

I can provide modified firmware that allows for dumping the MCU firmware; however, I cannot release it publicly. If you need it, please send me a private message specifying which firmware you require, along with your contact information, so that I can send the modified file to you.
By the way, I've already dumped the MCU firmware for the SC-8820. I haven't started working on the firmware for the SC-D70 yet, because my SC-D70 is broken 🙁

The dumping firmware is public on github, found in the issue post I link to. Posted by the grate nukeykt them self.

Reply 109 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
mckuhei wrote on 2026-11-04, 13:36:

I can provide modified firmware that allows for dumping the MCU firmware; however, I cannot release it publicly. If you need it, please send me a private message specifying which firmware you require, along with your contact information, so that I can send the modified file to you. By the way, I've already dumped the MCU firmware for the SC-8820.

I probably will contact you privately, but see my comment about SC-D70, because the same applies for my SC-8820 units as well. so, for the time being it's more important information to me that the MCU rom of SC-8820 can be dumped and it was dumped. I remember in past looking in the Service Manual of SC-8820 that the Wave Flash chip was kind of weird to dump - they used TSOP variant with pins mirrored compared to usual TSOP Flash chip - but maybe I recall wrongly....

mckuhei wrote on 2026-11-04, 13:36:

I haven't started working on the firmware for the SC-D70 yet, because my SC-D70 is broken 🙁

I have SC-D70, but it's somewhere buried in my storage room - I haven't seen it for years. however, if it helps the project, I can start digging to get it out - it was working last time I tried it. so, let me know - it will be a lot of effort to find it and it will certainly take some time, but maybe it's faster solution than fixing a broken unit. I also have collection of chip programmer - if anything is needed to be read that way in hardware and I can solder/desolder.

Karmeck wrote on 2026-11-04, 13:43:

The dumping firmware is public on github, found in the issue post I link to. Posted by the grate nukeykt them self.

you mean for SC-8820 or for SC-55, because what you're saying I can find there for SC-55 only?

Reply 110 of 121, by Shiftyy

User metadata
Rank Newbie
Rank
Newbie
Karmeck wrote on 2026-11-04, 06:52:
(mostly) What you are looking for is here: […]
Show full quote
Shiftyy wrote on 2026-11-04, 04:27:

Would this be able to run on a RPi compute module 4 (CM4 Lite), making it possible to be used on a real hardware via the wavetable WP32 McCake, or is the CM4 Lite too slow?

(mostly) What you are looking for is here:

https://misterfpga.org/viewtopic.php?t=7858&start=30

Not tested by me.

Damn, an optimized Nuked-SC55 AND Munt/Nuked-MT32 for the CM4? That's exactly what I was looking for haha. Thanks, will try it out soon.

Reply 111 of 121, by giulio

User metadata
Rank Newbie
Rank
Newbie

@costa

> but V1.02 is not recognized/supported

If you can provide the dump or a hash I can add support

Reply 112 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
giulio wrote on 2026-11-04, 15:11:

@costa

> but V1.02 is not recognized/supported

If you can provide the dump or a hash I can add support

thanks! MD5: cd72a33ca10295740bc924fdef52ab00

BTW, it seems there is public dump as well, when do search for "Roland SC-88 Control ROM v1.0.2" in Google.

Reply 113 of 121, by Karmeck

User metadata
Rank Member
Rank
Member
costa wrote on 2026-11-04, 13:53:
Karmeck wrote on 2026-11-04, 13:43:

The dumping firmware is public on github, found in the issue post I link to. Posted by the grate nukeykt them self.

you mean for SC-8820 or for SC-55, because what you're saying I can find there for SC-55 only?

So what I did was move the mcu (read: paid someone) from the SCC-1A into a Roland SC-55, mostly because I did not know how to power the sound card but also, how to use the control rom to dump try mcu if there are no slot for it.

Now, like I said in that thread, I know nothing really. But I guess the firmware to dump the sc-55 would also work for the sc-8820.

Reply 114 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
Karmeck wrote on 2026-11-04, 15:52:

But I guess the firmware to dump the sc-55 would also work for the sc-8820.

no, that cannot be the case, because SC-55 MCU is H8/532 and SC-8820 is using AFAICT SH7016 - they are totally different.

Reply 115 of 121, by Ishisashi

User metadata
Rank Newbie
Rank
Newbie
costa wrote on 2026-11-04, 00:28:
yes wrote on 2026-10-29, 15:10:

Now we have SCB-55/RLP-3237 internal ROM published and it looks like it's a BAD_DUMP speaking MAME terms - because its last 920 bytes (7C68-8000) resemble a few instruments definitions from a SC-55 series program (!) ROM.
Once the extreme definitions are incomplete and no other SC-55 series internal ROM contains its program ROM data, I believe that's an editing error probably overlapping R15199827 and R15209486 ROM images.

So please recheck and redump if it's really an error.

I disagree it's BAD_DUMP, because those work in Nuked-SC55, i.e. they are good no matter how they look. however, the overlap you're describing, while I can see it, I see it in almost all variants of Roland SC-55 devices internal ROM dumps - they cannot all be corrupted (presuming done by different people using different hardware) and at the same time working in the Nuked-SC55. so, that overlap maybe is something Roland did in their factory...

Corrupting a small portion of the ROM doesn’t necessarily render the entire device broken. For example, the SC-55mkII internal ROM could be used for testing—modifying its tail will not cause it to fail.

Reply 116 of 121, by costa

User metadata
Rank Newbie
Rank
Newbie
Ishisashi wrote on 2026-11-04, 18:11:

Corrupting a small portion of the ROM doesn’t necessarily render the entire device broken. For example, the SC-55mkII internal ROM could be used for testing—modifying its tail will not cause it to fail.

in theory I agree, but in this particular case the observed overlap that is supposed to be a corruption - I can observe it even after re-dumping those same alleged corrupted dumps using entirely software approach:

https://shingo45endo.github.io/sc-mcu-wave-romdumper/

where in my opinion any corruption is out of question and after all the original dumps were made with hardware approach like what can be seen on the picture here:

https://x.com/nukeykt/status/1752965047179981212
(taking that link from a previous post in this thread : Re: 88Emu is out (Roland SC-88/Pro/50 Emulator) because I did not know such in-depth details)

so, this discussion in my opinion is pointless and a dead-end, because that supposed "corruption" persists - no matter how the dump is made and by whom is made and what device is used. however, a corruption is a random thing - if it is really a corruption then it will change, but it doesn't change, which means those are the correct dumps, i.e. that alleged corruption is just an overlap made by Roland and how those ICs were actually programmed/produced in their factory. At least I see no other explanation that makes sense based on the facts.

Reply 117 of 121, by stgiga

User metadata
Rank Newbie
Rank
Newbie

Does anyone have a VA-76, VA-7, or VA-5 to dump (VA-3 has the same patches but is 8820 based rather than 8850 based) because all of those as well as the VA-3 have an 8850 map but with quite a lot of extra patches in the Bank Select MSB 40+ range, up into the 60+ range? It would be nice to hear the true final SC patches

I'm Nonbinary and neurodivergent.
I use they/them. It can also be written as ᵺ㏟.

Reply 118 of 121, by stgiga

User metadata
Rank Newbie
Rank
Newbie

Also I have a real SC-D70 if you want my help

I'm Nonbinary and neurodivergent.
I use they/them. It can also be written as ᵺ㏟.

Reply 119 of 121, by zaphod77

User metadata
Rank Member
Rank
Member

To elaborate on these arranger keyboards.

The G800, G1000, and va-76 all have a complete sc55 map, with all the sc55 rom patches. The G800 also has a sc88 map, with a large number of added patches. It is strictly better than a sc88 in that regard.
The G1000 adds an sc88Pro map, as well as a cut down version of the effects. it keeps the entire roms of the previous one in backwards compatibility maps
The VA76 adds a sc8850 map, as well as extra instruments to it. it does drop a few from the g800 map though. It only has 32 parts, though, and only 16 can play back external midis.

All of them provide CM64 support in all their maps, and appear to contain the actual samples for previous versions.

From a playing midis perspective, the g1000 and the va-76 could use a dump to gain access to unique instruments.