They use a lot of the same physical parts from what I can tell, but the Wave ROM is entirely different, which means the sound is entirely different.
@Falcosoft Well it looks like the MT32 mode does not work at all as it just results in no sound at all. Oh well. There is a file Deadend.mid that I have that does not play correctly (random loud instruments are played around half way through the file) and it ID's as a GS compatible file.
I think the SC-88 fileTxplasma.mid also sounds weird as my synth is a strange combination of units in a sense. When I tried you instructions for setting it to SC-88 mode it sounded even more incorrect.
I have attached the files for you to take a look at.
Would it be worth it for me to attempt to dump the wave rom and system rom? I may not be able to dump the system rom as it is 16 bits wide and I am hoping to use my USB 2.0 8 channel 48Mhz LA to view the data when I run the built in rom tests. I don't think my programmer from the 80's can handle this chip layout and I don't want to risk un-soldering the chip so my plan was to clip the LA onto the data bus legs.
Also found the jumper for the CPU multiplier so I changed the FSB to 33Mhz and the multiplier to 3x as the system was not stable when running the video at 800 x 600 at 50Mhz FSB.
Poot36 wrote on 2025-02-17, 17:24:The playback issues I have had are missing instruments, too quiet instruments, corrupted and incorrect instruments. It seems to get worse the more MIDI files I play almost like it is not getting reset or something. I am using Windows Media player to play the files. I also have had the volume go to full even when I have turned it way down on the synth before playing a song (this happens without having played another song beforehand) I assume this may be normal as not all MIDI files do this.
I have randomly had this happen while using an ewave wavetable with multiple sound cards. Not sure what was causing it. What fixed it - using winamp for midi playback, going into midi playback options and changing "additional reset commands between tracks " from the default value of "none" to "gm".
Poot36 wrote on 2025-03-06, 02:38:@Falcosoft Well it looks like the MT32 mode does not work at all as it just results in no sound at all. Oh well. There is a fi […]
@Falcosoft Well it looks like the MT32 mode does not work at all as it just results in no sound at all. Oh well. There is a file Deadend.mid that I have that does not play correctly (random loud instruments are played around half way through the file) and it ID's as a GS compatible file.
I think the SC-88 fileTxplasma.mid also sounds weird as my synth is a strange combination of units in a sense. When I tried you instructions for setting it to SC-88 mode it sounded even more incorrect.
I have attached the files for you to take a look at.
Would it be worth it for me to attempt to dump the wave rom and system rom? I may not be able to dump the system rom as it is 16 bits wide and I am hoping to use my USB 2.0 8 channel 48Mhz LA to view the data when I run the built in rom tests. I don't think my programmer from the 80's can handle this chip layout and I don't want to risk un-soldering the chip so my plan was to clip the LA onto the data bus legs.
Also found the jumper for the CPU multiplier so I changed the FSB to 33Mhz and the multiplier to 3x as the system was not stable when running the video at 800 x 600 at 50Mhz FSB.
Hi,
Since I have no such hardware as yours I used Nuked SC-55 (which is the most accurate SC-55 emulator available) with mkII ROMs for testing.
Both of your attached problematic Midi files use 2 drum channels but this should not cause problems. Besides this DEADEND.MID is really a simple GS file that should be played without any problems on GS compatible devices.
TXPLASMA.MID is really written for SC-88+ so it uses SC-88 specific variations that are not available on SC-55. I used mkII ROMs for testing since mkII has no CTF feature so invalid variation tones results in silence on affacted channels. But with my Midi player's software CTF feature activated the Midi sounds pretty good on mkII.
I have made a test video so you can compare your results with mine.
In the 1st part of the video I tried to play TXPLASMA.MID without CTF so there were many missing instruments. If this sounds the same as your synth that means your synth is mkII-like and hardware CTF is missing.
In the 2nd part I tried to play TXPLASMA.MID with Midi player's software CTF activated and as can be heard it sounded pretty good. You should be able to get the same sound with software CTF even if your synth is mkII-like.
In the lat part I played DEADEND.MID without CTF since it does not matter (no such variation is used by the Midi that SC-55 is missing). It sounded good and I could not hear your mentioned 'random loud instruments'.
@Edit:
BTW, here is a video about how TXPLASMA.MID sounds as intended on SC-88+. This time I used Roland's SC-VA VST plugin that is not as accurate as Nuked SC55 but it can emulate even SC-8820/SC-88Pro/SC-88.
Poot36 wrote on 2025-03-06, 02:38:@Falcosoft Well it looks like the MT32 mode does not work at all as it just results in no sound at all. Oh well. There is a fi […]
@Falcosoft Well it looks like the MT32 mode does not work at all as it just results in no sound at all. Oh well. There is a file Deadend.mid that I have that does not play correctly (random loud instruments are played around half way through the file) and it ID's as a GS compatible file.
I think the SC-88 fileTxplasma.mid also sounds weird as my synth is a strange combination of units in a sense. When I tried you instructions for setting it to SC-88 mode it sounded even more incorrect.
I have attached the files for you to take a look at.
Would it be worth it for me to attempt to dump the wave rom and system rom? I may not be able to dump the system rom as it is 16 bits wide and I am hoping to use my USB 2.0 8 channel 48Mhz LA to view the data when I run the built in rom tests. I don't think my programmer from the 80's can handle this chip layout and I don't want to risk un-soldering the chip so my plan was to clip the LA onto the data bus legs.
Also found the jumper for the CPU multiplier so I changed the FSB to 33Mhz and the multiplier to 3x as the system was not stable when running the video at 800 x 600 at 50Mhz FSB.
Obvious points:
1. The EM-10 does NOT have the MT-32 tones present (MT-32 tones are NOT part of the GS sound set). Therefore, trying to use those tones will result in silence or incorrect instruments.
2. The EM-10 is NOT SC-88 compatible, it's in fact sound wise equivalent to a SC-50 (SC-55mkII minus MT-32 tones). It is only GM and basic GS (level 1) compatible. It is not MT-32 compatible, or SC-88 or higher compatible.
Dumping the ROM could be fun, but as it's 1MB (2MB uncompressed) it's extremely likely to be the same as the SC-55 wave ROM, possibly slightly more compressed.
As per the demos here: https://synthmania.com/2004/12/26/roland-em-10/
You can hear it sounds like a SC-55.
@Falcosoft I have recorded playback of the Deadend.mid with both the Midi player from this forum and WMP. I think I have the settings for the Midi player set to default and for the WMP I used the GM GS reset MIDI before playing the track.
Judging from other users the Txplasma.mid will never play quite correctly on my synth so I have given up on that for now.
I have intentionally recorded the samples quiet at the start as it will go into clipping when the playback gets weird.
If you need the files as a different format I can re-export them as well.
I tried my best to have the recording times be identical as I really don't know how to use Audacity to make them identical.
There is a slight volume difference between the files as well most likely because I used the physical volume buttons on the synth to confirm that the reset MIDI file did not change the volume and I may have not quite been able to set it back to the volume from the other Midi player program.
Poot36 wrote on 2025-03-10, 00:35:@Falcosoft I have recorded playback of the Deadend.mid with both the Midi player from this forum and WMP. I think I have the s […]
@Falcosoft I have recorded playback of the Deadend.mid with both the Midi player from this forum and WMP. I think I have the settings for the Midi player set to default and for the WMP I used the GM GS reset MIDI before playing the track.
Judging from other users the Txplasma.mid will never play quite correctly on my synth so I have given up on that for now.
I have intentionally recorded the samples quiet at the start as it will go into clipping when the playback gets weird.
If you need the files as a different format I can re-export them as well.
I tried my best to have the recording times be identical as I really don't know how to use Audacity to make them identical.
There is a slight volume difference between the files as well most likely because I used the physical volume buttons on the synth to confirm that the reset MIDI file did not change the volume and I may have not quite been able to set it back to the volume from the other Midi player program.
Hi,
It seems that your synth misinterprets GS specific NRPN (1,8) Vibrato range messages on channel 4 (Program: Organ 3). To confirm or refute my theory try this modified Midi file that does not contain any such NRPN messages on channel 4.
@Edit:
I have just noticed that the same NRPNs are also used on channel 8 (Program: Organ 2). I have removed them, too. New modified Midi file is re-uploaded.
Well I tried it and it was just as bad. I have attached a recording of it.
I also found if I use the volume control in the Midi player program and I turn it down not much lower then I currently have it some of the sounds just disappear, they don't get quieter.
The strange thing is that my synth passes all onboard diagnostics for the ROM, Wave ROM and the RAM.
Poot36 wrote on 2025-03-10, 02:25:Well I tried it and it was just as bad. I have attached a recording of it.
I also found if I use the volume control in the Midi player program and I turn it down not much lower then I currently have it some of the sounds just disappear, they don't get quieter.
The strange thing is that my synth passes all onboard diagnostics for the ROM, Wave ROM and the RAM.
Hmm, interesting.
Then let's try to find what channel/instrument causes the issue. Here is a version wih channel 8 (Organ 2) removed.
And here is the one with channel 4 (Organ 3) removed.
@Edit:
The only other controller that I could find and can be relevant (i.e. it is modified at the moment the problematic sound starts) is the Modulation Wheel (CC#1). Here is a version the removes modwheel 120 events from both channel 4 and channel 8.
Well it looks like removing channel 4 fixed it. When I played the channel 8 one it still went strange. I also noticed with the channel 8 one the instant it went strange the BPM dipped and then came back to normal speed and near the end of the song stayed lower the the original BPM. That may explain why it sounded strange. Is it possible that this song is to fast for my system to play correctly?
Poot36 wrote on 2025-03-10, 03:09:Well it looks like removing channel 4 fixed it. When I played the channel 8 one it still went strange. I also noticed with the channel 8 one the instant it went strange the BPM dipped and then came back to normal speed and near the end of the song stayed lower the the original BPM. That may explain why it sounded strange. Is it possible that this song is to fast for my system to play correctly?
No, there are tempo changes even right at the beginning (at 16-17 sec.) and while it's true that there is one tempo change near at the problematic part it is such a tempo change that causes less than 10% difference in speed/BPM so it cannot cause such problem we can hear.
Ok, so it is limited to channel 4 having some sort of strange issue then?
Poot36 wrote on 2025-03-10, 03:31:Ok, so it is limited to channel 4 having some sort of strange issue then?
Yep, but it is hard to say what the exact problem is.
If my lastly attached DEADEND_NoModWheel.mid results in the same problematic sound then I have no other idea.
Sorry I did not see that you edited your previous post. I just tried the NoModWheel and it played just fine. I am going to assume that my synth does not correctly support or handle that command?
Poot36 wrote on 2025-03-11, 00:48:Sorry I did not see that you edited your previous post. I just tried the NoModWheel and it played just fine. I am going to assume that my synth does not correctly support or handle that command?
Hi,
If the NoModWheel version plays fine then I have a plausible explanation:
Your synth does not support properly one or more of the SysEx messages that modify the modulation wheel LFO behavior on channel 4. The used SysEx messages also can be seen on the screenshot that I posted previously related to the Tempo Change events.
Your synth somehow misinterprets these SysEx messages but the problem only manifests itself when the modulation wheel (CC#1) is set to a higher value (120) on channel 4.
So without changing the default modulation wheel parameters your synth can interpret modulation wheel messages without problems but if some GS specific SysEx messages modify some LFO parameters then problems can occur.
You can try to fix the problematic SysEx messages with the help of my Midi Player:
Load your original DEADEND.mid file with Midi Player
1. Open the Event Viewer/Debugger.
2. At the very beginning of the event list you can see some SysEx messages.
3. Remove/delete the channel 4 mod wheel LFO related ones:
F0411042124024055542F7
F0411042124024063066F7
4. Save the file and re-load it with Midi Player.
PS.
If somehow deleting the above mentioned SysEx messages does not fix the problem you can try to delete other SysEx messages until you find the right one(s).
Well modifying the file by removing the SysEx messages worked. I do find it kind of funny that the file is made by Roland and it won't play correctly on their hardware.
Poot36 wrote on 2025-03-11, 02:29:Well modifying the file by removing the SysEx messages worked. I do find it kind of funny that the file is made by Roland and it won't play correctly on their hardware.
The problem is not the Midi file but the synth. The Midi file works with the original SC-55 and even with SC-88/88 Pro etc. without problems.
This seems to be a specific hardware implementation related bug.
Interesting, I guess there is no way to fix it?
Poot36 wrote on 2025-03-11, 02:57:Interesting, I guess there is no way to fix it?
I do not know, maybe Spikey can tell.
Poot36 wrote on 2025-03-11, 02:57:Interesting, I guess there is no way to fix it?
It depends what you're hoping for by way of a "fix". The EM-10 should work fine for GM gaming. Playing MIDI files will depend on whether the file is GM/basic GS or not, and then the EM-10 would appear to have a somewhat limited SysEx compatibility. I would say that example with the organ and complex SysEx shouldn't be a super-common experience, although obviously it depends on what files you want to play!
I'm not familiar enough with the EM-10, but Falco would appear to have identified the specific issue in that MIDI's case. If you run into others, feel free to provide a recording and the original MIDI and we'll take a look, I guess!
