m5215tx wrote:

If an MT-32 locks up due to the buffer overflow bug and it is in a chain connected to other MIDI sound modules would the MT-32 still be sending out a MIDI signal via the THRU MIDI port so that the other sound modules below it would still run?

Yes, MIDI in is routed to MIDI thru in hardware.

m5215tx wrote:

I use the Roland um-one mk2 USB cable to connect the sound modules to my system. Would it be allowable to have a second um-one cable connected to the same system just for the MT-32 so that it can't affect any other sound modules in a chain if it locks up?

Not because of lockup, no. But if you have multiple USB to MIDI adapters, then you can power up only the devices you need, so you don't have to power up MT-32 just because it needs to pass MIDI data through it.

Thanks for the information Jepael!

I am building a sizeable collection of sound modules so that info will be useful.

How do you like your UM-One MK2? Is it causing the buffer overflow for your MT-32 or is that just the inherent bug that's part of the hardware? I just picked up a Yamaha MU50 Tone Generator and plan on using the UM-One as the interface for my PC to the box.

The Roland um-one mk2 is a great accessory so I have been completely satisfied with it. From what I understand Yamaha also has a similar type of cable but I don't know if its as good as the Roland one. I am just sticking with what works for me.

The buffer overflow is a bug in the hardware of the MT-32. I have not personally run into the issue yet. I don't know if using the Roland um-one mk2 cable will mitigate the problem or not.

m5215tx wrote:

The buffer overflow is a bug in the hardware of the MT-32. I have not personally run into the issue yet. I don't know if using the Roland um-one mk2 cable will mitigate the problem or not.

First of all, it is not a bug in hardware but firmware and later firmware versions do not have this overflow issue that hangs the unit.

I don't have any statistics but the bug getting triggered with any modern USB MIDI adapter is possible due to how modern USB and operating systems work.

The original DOS game has more real time control how it sends MIDI data over MIDI interface, so it can keep the necessary delays between the necessary messages.

But even if you play these DOS games over emulators such as DosBox or ScummVM, they may not implement these delays between MIDI messages, and even if they do, most likely the OS/MIDI/USB drivers and the adapter itself will buffer these messages so the emulators have little control of actual delays between messages that come out from the USB MIDI adapter.

m5215tx wrote:

The Roland um-one mk2 is a great accessory so I have been completely satisfied with it. From what I understand Yamaha also has a similar type of cable but I don't know if its as good as the Roland one. I am just sticking with what works for me.

The buffer overflow is a bug in the hardware of the MT-32. I have not personally run into the issue yet. I don't know if using the Roland um-one mk2 cable will mitigate the problem or not.

Apparently Roland, EMU, and M-Audio USB adapters are not problematic. The Yamaha UX16 adapter might have issues, but it would be worth testing, I think. I've used the X3M from Serdaco and Mbox 2 from Avid without issues, as well.