Microsoft baking MIDI support into Windows

[ajay_m]

Encouraging.....

https://www.theregister.com/2025/02/06/ ... ervices_2/

[Luke W]

Indeed!

For those interested, it's covered here. The story also has a few links to some detailed breakdowns from the team at Scan and Microsoft's own developer blogs.

[shurikt]

I'm looking forward to that 75 year old Korg driver being retired...

Their 2024 NAMM show presentation is on Youtube here:

https://www.youtube.com/watch?v=-pe29zIVUCA

[The Elf]

I've heard something about this in relation to the ridiculous Korg requirement for their devices to be listed in the first ten MIDI device slots (what year is this?).

Apparently it will be important to keep the registry 'midi1' slot assigned to 'wdmaud2.drv' to make sure that all works as it should.

What could possibly go wrong?...

[The Elf]

Got around to watching the video of the Windows MIDI 2 presentation.

As soon as he reached the parts where things weren't working I began to see my future...

Also nice to see Microsoft at last recognising ASIO. It's only been... how long?

Call me a wrinkly old pessimist, but I predict a launch blessed with chaos and hasty software patches. I'd love to be proved wrong.

[Martin Walker]

Call me a wrinkly old pessimist, but I predict a launch blessed with chaos and hasty software patches. I'd love to be proved wrong.

Looking into my crystal ball, I suspect your prediction has a high chance of coming true

[OneWorld]

Call me a wrinkly old pessimist, but I predict a launch blessed with chaos and hasty software patches. I'd love to be proved wrong.

Looking into my crystal ball, I suspect your prediction has a high chance of coming true

Probably as much of a foregone conclusion as it is a prediction

If it is anything like the experience I am having with Skype, there is foreboding. Skuype used to be some simple and straightforward a demented parrot culd have used it. Now each time I sign in, it takes an age to to pull itself together (I use the browser version, which I always used) and when it is up and running, the Contacts List needs to re-populate itself whereas there are only two regular contacts I use, and on Chats, the list of not often used is shown but the one I use every day almost I have to dig out of Contacts, and the video chat function is worse than useless, whereas before, it came up in an instant and just worked, people could just call me and there they were. Not anymore.

Do I realkly want Microsoft meddling with my MIDI etc, I don't think so

[MOF]

Apparently it will be important to keep the registry 'midi1' slot assigned to 'wdmaud2.drv' to make sure that all works as it should.

What could possibly go wrong?... :lol:

That reminds me of one version of Cubase on the Atari that sent all controller data out on Midi channel 1. How it got missed by alpha testers, never mind beta testers, amazed me, that fault on channel 16 I could just about understand.

[N i g e l]

At least Microsoft are now doing somthing. Thats got to better than not doing anything and maintaining the current situation.

Is anyone here a beta tester ?

[James Perrett]

So what is actually new here apart from Midi 2 support? Windows has had Midi since 1992.

I know that certain German software writers couldn't get their head around how Windows worked so they created their own standards for audio but Windows Midi has always been there.

[The Elf]

A lot more to it than that. It's worth watching the video mentioned.

[James Perrett]

A lot more to it than that. It's worth watching the video mentioned.

I don't really have time for a waffly video (which most recording related ones seem to be). The V2 class compliant USB audio driver is nothing new - the only new progress is that MS seem to have come to a licensing agreement with the writer. It won't be up to RME standards if it is the Thesycon driver that has been around for years and is already used by a large number of audio interface manufacturers.

[The Elf]

Fair enough. As I said, there's a lot more to it than you seem to think.

[Aled Hughes]

A lot more to it than that. It's worth watching the video mentioned.

I don't really have time for a waffly video (which most recording related ones seem to be). The V2 class compliant USB audio driver is nothing new - the only new progress is that MS seem to have come to a licensing agreement with the writer. It won't be up to RME standards if it is the Thesycon driver that has been around for years and is already used by a large number of audio interface manufacturers.

I’m crossing fingers that it’ll be more stable at lower latencies than the ASIO driver that comes with the Soundcraft UI24R. That would be nice.

I don’t expect it to be much use with my RME interfaces though!

[James Perrett]

I've just watched most of the video and I'd have to say that it is less waffly than I had feared. However, the demo was painful. My lad could have put an intuitive graphical control panel together in a couple of days which would have saved all that typing and would have given a much better overview of what was happening in the system.

[N i g e l]

The crux of the matter is that apple does it better, windows needs to catch up
[alledgedly, from what I read, I have no apple experience]

Windows has had Midi since 1992.

Installation of the afforementioned Korg driver [ ~2020] required me to strip out my usb devices, install the korg so it was in the top 10 USB devices, then reinstall all my USB gear. Thats not good.

[forumuser828745]

"...so not everything will work out of the box. Microsoft wants users to report issues and backward compatibility problems with existing hardware and software."

You are the beta tester

The comments are worth a read on that El Reg article.

[The Elf]

Installation of the afforementioned Korg driver [ ~2020] required me to strip out my usb devices, install the korg so it was in the top 10 USB devices, then reinstall all my USB gear. Thats not good.

Same here, but I found a point where I couldn't achieve this. Every time I opened up a first ten slot, other drivers would leap in immediately. I was made aware of a utility to do the job:
https://github.com/gtsiber/MidiDriverOrderForKorg

Use at your own risk. It's worked well for me. If you *do* use it, then take note of what I said about device position midi1.

[N i g e l]

Thanks, I had not seen that before.

[James Perrett]

I’m crossing fingers that it’ll be more stable at lower latencies than the ASIO driver that comes with the Soundcraft UI24R. That would be nice.

I don’t expect it to be much use with my RME interfaces though!

As far as I can work out, they've probably licensed the same driver from Thesycon that your Soundcraft already uses. Many devices that are class compliant on the Mac or Linux use the same Thesycon driver in Windows. The only difference in the future will be that the driver comes with Windows rather than the user having to download it from the AI manufacturer's website.

In fact, as I understand it, the driver has been included in Windows for a while now but it will currently only install itself if you are using an interface on its whitelist.

Of course the Thesycon driver isn't up to RME's standard although it will work with the RME interfaces that have a class compliant mode.

[Aled Hughes]

I’m crossing fingers that it’ll be more stable at lower latencies than the ASIO driver that comes with the Soundcraft UI24R. That would be nice.

I don’t expect it to be much use with my RME interfaces though!

As far as I can work out, they've probably licensed the same driver from Thesycon that your Soundcraft already uses. Many devices that are class compliant on the Mac or Linux use the same Thesycon driver in Windows. The only difference in the future will be that the driver comes with Windows rather than the user having to download it from the AI manufacturer's website.

In fact, as I understand it, the driver has been included in Windows for a while now but it will currently only install itself if you are using an interface on its whitelist.

Of course the Thesycon driver isn't up to RME's standard although it will work with the RME interfaces that have a class compliant mode.

Thanks James, I feared that would be the case! Oh well. It works well enough with the Soundcraft, and I’ve never had trouble with it - it’s just that I noticed it wasn’t happy at low latencies when I was experimenting with using VSTs on live inputs on the desk. No big deal in this case!

[Pete Kaine]

As far as I can work out, they've probably licensed the same driver from Thesycon that your Soundcraft already uses.

The Windows USB 2.0 ASIO driver? Developed by Yamaha for Microsoft.

[James Perrett]

The Windows USB 2.0 ASIO driver? Developed by Yamaha for Microsoft.

Thanks for the correction Pete - I had read on the Thesycon website that they had created a generic Windows Audio driver for class compliant devices which MS included in newer versions of Windows so assumed that they had done the same for ASIO (as I would guess that they already have the code to do that).

[SteveFlesh]

I've been half following this for a little while now (not watched the video yet though)
Not sure about the ASIO side of things, but the midi driver appears to be developed by AMEnote and funded by AMEI.
Link below to a blog entry from Oct 2024 which also mentions other partners, (altough his seems to be the same one in the Register pirece)

https://devblogs.microsoft.com/windows- ... 24-update/

The Windows Midi Service project has been on github for a while and available to developer, so no 'beta testers' as such, but as the article states, it's in the Canary build of Windows 11 now if you're brave enough...

[BJG145]

Does ASIO have any connection with MIDI? It never occurred to me they were related?