Multipacket SysEx is not received correctly on macOS

[James-Morrell]

I'm trying to use DryWetMidi on MacOS, to receive SysEx messages from a MIDI USB device.

The sysEx size is 251 Bytes. The Windows build works fine, but with the MacOS build only the first two bytes are received in the SysExEvent.

(MacOS Version 12.61 Monterey)

Let me know if you need any more information. Thanks, James

[melanchall]

Hi James,

Yes, I definitely need more information:

1. Can you please check this sysex message receiving by any other MIDI software? Is it OK?
2. Does the problem occur only with this specific 251-byte sysex message or with any sysex one?
3. Can you please provide bytes of the message so I can try to reproduce the error within my unit tests?

Also please be patient, I'm not a macOS user, so you can be asked a lot of questions by me to clarify what's going on :)

Thanks, Max

[James-Morrell]

Hi Max,

Thanks for the quick response, here are my answers:

1. It works fine in other midi software including SysEx Librarian & MidiView
2. I'll check this next, will post a response asap.
3. Here are the bytes in Hex format: 00 21 5F 00 00 01 03 01 01 06 00 00 02 00 00 4C 00 00 41 00 00 52 00 00 49 00 00 20 00 00 42 00 00 20 00 00 53 00 00 4F 00 00 4C 00 00 4F 00 00 20 00 00 20 00 00 20 00 00 20 00 00 20 00 00 02 00 00 3D 04 00 00 00 00 74 03 00 64 00 00 74 03 00 00 00 00 74 03 00 64 00 00 01 00 00 1C 00 00 00 00 00 00 00 00 7F 7F 03 05 00 00 64 00 00 7F 7F 03 05 00 00 64 00 00 7F 7F 03 05 00 00 64 00 00 00 00 00 3C 00 00 00 00 00 00 00 00 64 00 00 64 00 00 64 00 00 64 00 00 00 00 00 00 00 00 2C 02 00 14 00 00 28 00 00 3C 00 00 64 00 00 64 00 00 64 00 00 00 00 00 6C 7F 03 6C 7F 03 04 00 00 0A 00 00 00 00 00 7F 7F 03 01 00 00 48 01 00 00 00 00 6C 7F 03 7F 7F 03 04 00 00 7F 7F 03 00 00 00 7F 7F 03 01 00 00 48 01 00 00 00 00 30 7F 03 14 00 00 48 01 00 00 00 00 F7

Feel free to ask me anymore questions I will try to help get this issue solved. James

[melanchall]

Regarding bytes: are you sure thare is no F0 byte at the start of the message?

[James-Morrell]

Sorry, there is a F0 at the start of the message. Just that SysExEvent.Data's print has truncated it.

I've just tried this with another device and send a variety of SysEx messages, and it seems to do the same. The simplest one which displays the issues is: F0 7F 60 40 F7

Thanks James

[James-Morrell]

Just as an update, the sysEx messages are fine with 2 bytes or less.

I've just tried another API using RtMidi on mac os, and that works: RtMidi.Core

Maybe that can be some help for reference?

[melanchall]

Thanks! I'll look into it tomorrow.

[melanchall]

Well, I've created unit tests with the provided bytes arrays. Running them on virtual machine with Monterey 12.6.1 ends with success, sysex events are received correctly. So we need to do some tests on your machine.

I've prepared the archive: TestKit.zip. Unzip it please somewhere on your computer.

---

First of all, run the LoopbackDevice program from the folder in a separate Terminal window:

./LoopbackDevice "MIDI A" "MIDI B" "MIDI C"

This commad will create three virtual loopback devices. Devices are present in the system as long as the program running, so please don't close the Terminal.

---

Now let's open the SendDataViaDwm.sln solution in IDE (VS for Mac or Rider) and run it. We expect to see the message that sysex F0 7F 60 40 F7 is received. Please write me if bytes are correct in the message (F0 will be omitted).

---

Then let's open the SendDataViaBytes.sln solution in IDE (VS for Mac or Rider) and run it. We expect to see the message that sysex F0 7F 60 40 F7 is received. Please write me if bytes are correct in the message (F0 will be omitted).

[James-Morrell]

Hi Max,

Just ran the test's on the Mac device I have, The "SendDataViaDwm.sln" worked correctly. But the "SendDataViaBytes.sln" did not. I received no message with the "SendDataViaBytes.sln" and only the Press any key to exit... displayed.

[melanchall]

Thanks James!

Well, I'll try to provide new version of the SendDataViaBytes to try to understand why it's not working. For now, can you please confirm that you see wrong data with SendDataViaDwm if you send sysex via your device the problem occurred with?

[James-Morrell]

Just checked SendDataViaDwm with two different MIDI USB devices, I sent the SysEx from my windows machine. With both devices it only shows the first two bytes 7F 60. On the loopback it is fine.

[melanchall]

Pretty strange... Hm, it looks like we need to write some code in C in XCode. I'll try to provide you the code tomorrow.

[melanchall]

Hi James,

I've prepared a new solution: ViewPacketsData.zip. Replace name of the device with the desired one from where you receive broken data:

InputDevice.GetByName("MIDI A");

Just follow the instructions from the utility and provide me an output text please. I made the solution without any tests so please write me about any errors if any.

[James-Morrell]

Hi Max,

I've just run the ViewPacketsData, here's the output from the terminal:

Packet 0 / 1 data:
    F0 0 21
Packet 0 / 1 data:
    5F 0 0
Packet 0 / 1 data:
    1 3 1 1 6 0
Packet 0 / 1 data:
    0 0 0
Packet 0 / 1 data:
    0 4C 0 0 41 0 0 52 0 0 49 0 0 20 0 0 42 0 0 20 0 0 38 0 0 54 0 0 48 0 0 20 0 0 4E 0 0 4F 0 0 54 0 0 45 0 0 20 0
Packet 0 / 1 data:
    0 1 0 0 F 2 0 0 0 0 74 3 0 64 0 0 74 3 0 0 0 0 74 3 0 64 0 0 1 0
Packet 0 / 1 data:
    0 18 0 0 1E 0 0 1C 0 0 7F 7F 3 28 0
Packet 0 / 1 data:
    0 64 0 0 7F 7F
Packet 0 / 1 data:
    3 5 0
Packet 0 / 1 data:
    0 64 0
Packet 0 / 1 data:
    0 7F 7F
Packet 0 / 1 data:
    3 5 0
Packet 0 / 1 data:
    0 64 0
Packet 0 / 1 data:
    0 0 0
Packet 0 / 1 data:
    0 2C 0 0 1 0 0 0 0 0 64 0 0 64 0 0 64 0 0 64 0 0 0 0 0 0 0 0 2C 2 0 14 0 0 28 0 0 3C 0 0 64 0 0 64 0 0 64 0
Packet 0 / 1 data:
    0 0 0 0 6C 7F 3 6C 7F 3 4 0 0 A 0 0 0 0 0 7F 7F 3 1 0 0 48 1 0 0 0 0 6C 7F 3 7F 7F 3 4 0 0 7F 7F 3 0 0 0 7F 7F
Packet 0 / 1 data:
    3 1 0 0 48 1 0 0 0 0 30 7F 3 14 0
Packet 0 / 1 data:
    0 48 1
Packet 0 / 1 data:
    0 0 0
Packet 0 / 1 data:
    0 F7

[melanchall]

Hi James,

Is that output on sending one single event or multiple ones? If multiple, can you please just send a single one (for example, F0 7F 60 40 F7)?

[melanchall]

Can you please run a new version of the solution: ViewPacketsData.zip? I have an assumption what's going on.

[melanchall]

@James-Morrell Any news?

[James-Morrell]

Hi Max,

Sorry I've been away for a couple of days, will look at it some point today :)

James

[melanchall]

Hi James,

Here another one new version: ViewPacketsData.zip :)

[James-Morrell]

Hi @melanchall

Here's the output for the latest ViewPacketsData.zip

Waiting for packets information...
Just send problem sysex and wait some time, then just press any key to see the log
 Message received at [2022-12-16T14:47:01.2838960Z]!
Packet 0 / 1 data:
    F0 00 21
Message received at [2022-12-16T14:47:01.3909050Z]!
Packet 0 / 1 data:
    5F 00 00
Message received at [2022-12-16T14:47:01.3910180Z]!
Packet 0 / 1 data:
    01 03 01 01 06 00 00 00 00 00 4C 00 00 41 00 00 52 00
Message received at [2022-12-16T14:47:01.3911710Z]!
Packet 0 / 1 data:
    00 49 00 00 20 00
Message received at [2022-12-16T14:47:01.3912260Z]!
Packet 0 / 1 data:
    00 42 00 00 20 00 00 38 00 00 54 00 00 48 00 00 20 00 00 4E 00 00 4F 00 00 54 00 00 45 00 00 20 00 00 01 00 00 0F 02 00 00 00 00 74 03 00 64 00
Message received at [2022-12-16T14:47:01.3912940Z]!
Packet 0 / 1 data:
    00 74 03 00 00 00 00 74 03 00 64 00 00 01 00 00 18 00 00 1E 00 00 1C 00 00 7F 7F 03 28 00 00 64 00 00 7F 7F 03 05 00 00 64 00 00 7F 7F
Message received at [2022-12-16T14:47:01.4502970Z]!
Packet 0 / 1 data:
    03 05 00
Message received at [2022-12-16T14:47:01.4503970Z]!
Packet 0 / 1 data:
    00 64 00
Message received at [2022-12-16T14:47:01.4505440Z]!
Packet 0 / 1 data:
    00 00 00
Message received at [2022-12-16T14:47:01.4506370Z]!
Packet 0 / 1 data:
    00 2C 00 00 01 00 00 00 00 00 64 00 00 64 00 00 64 00 00 64 00 00 00 00 00 00 00 00 2C 02 00 14 00 00 28 00 00 3C 00 00 64 00 00 64 00 00 64 00
Message received at [2022-12-16T14:47:01.4507270Z]!
Packet 0 / 1 data:
    00 00 00 00 6C 7F 03 6C 7F 03 04 00 00 0A 00 00 00 00 00 7F 7F 03 01 00 00 48 01 00 00 00 00 6C 7F 03 7F 7F 03 04 00 00 7F 7F 03 00 00 00 7F 7F
Message received at [2022-12-16T14:47:01.4507960Z]!
Packet 0 / 1 data:
    03 01 00 00 48 01 00 00 00 00 30 7F 03 14 00
Message received at [2022-12-16T14:47:01.4558240Z]!
Packet 0 / 1 data:
    00 48 01
Message received at [2022-12-16T14:47:01.4559230Z]!
Packet 0 / 1 data:
    00 00 00
Message received at [2022-12-16T14:47:01.4560460Z]!
Packet 0 / 1 data:
    00 F7

Press any key to exit...

Do you want the other output from the previous version, it is the same but without the time stamps.

James

[melanchall]

Hi James,

Thank you a lot! Now I know for sure what's going on. Your message was split into multiple separate messages. DryWetMIDI right now assumes that single packet == single complete message. Which is obviously wrong according to your tests.

It's definitely the bug in my library. It should be pretty easy to fix, but I think I can handle it in next year only (I'm going to vacation right now :)).

I appreciate your assistance with diagnosing the problem. Thanks!

[James-Morrell]

Hi Max, Glad I could help, let me know if you need anymore tests this end!

Have a good holiday! James

[melanchall]

Thanks James! If I'll be able to fix the bug before next year, I definitely let you know.

[melanchall]

Hi James,

Looks like the bug has been fixed now according to my unit tests. I've published new prerelease version of the library - 6.1.4-prerelease2. @James-Morrell Can you please install it into your project and let me know whether the problem is gone or not?

Thanks, Max

[James-Morrell]

Hi Max,

Sorry for the late response, been away for the holidays!

Just tested 6.1.4-prerelease2 and it's all working!

Kind regards, James

[melanchall]

Hi James,

Great! Thanks for the issue and your assistance! You've made the library better.

Thanks, Max

[melanchall]

:rocket: 6.1.4 version is released now!

Prerelease NuGet packages will be unlisted soon, so please update the package references to the new version.

Thanks for using the library!