Microsoft are bringing native ASIO support, audio-optimized USB drivers to Windows 11 in late 2025

[Starship Krupa]

From known Cakewalker Pete Brown of Microsoft:

https://devblogs.microsoft.com/windows-music-dev/making-music-on-windows/

I was most interested to see that the initial testing will be on the arm64 platform. Probably because they knew it would need a longer lead time, but there seems to be a push there to get DAW's working well in Windows 11 on arm64.

I count at least 2 interface manufacturers and 3 DAW's that are releasing native arm64 builds.

Most interesting. I had only thought of arm64 processors as something aimed at mobile device use, for lower heat and power consumption. Looks like it's going to be making inroads at least for laptop producers.

Yoiks, I don't know of any plug-in manufacturers who supply native arm64 versions of their DLL's. Here's hoping for their sake that it's just a recompile.

[Amberwolf]

From the very very little coding I have done (mostly arduino, esp32, rpi recently), *nothing* is "just" a recompile for a different platform and/or cpu/mpu.     

With luck, that's not true of situations like this, but....

[Joseph Gresham]

That’s awesome! native ASIO support on Windows 11 will make a big difference for music production. Can’t wait to see how this improves DAWs and plugins!

[Jim Roseberry]

FWIW, I don't get the excitement.

 

• Windows 11 x64 is already a great DAW platform.
• Laptop users can currently run ASIO drivers.

 

The only "advantage" of native ASIO is that the motherboard's onboard audio interface can be used via ASIO.

Most DAW users are running a dedicated audio interface... for a multitude of reasons (higher quality A/D D/A, lower noise-floor, balanced I/O, etc).

Some current audio interfaces can run sub 1ms total round-trip latency.  Native ASIO isn't going to improve upon this.

 

Apple has had "Core Audio" for years (akin to Native ASIO).

Core Audio doesn't outperform current ASIO drivers on a PC.

 

[pwallie]

i guess they've bought asio4all haha

[TheMaartian]

3 hours ago, Jim Roseberry said:

FWIW, I don't get the excitement.

 

• Windows 11 x64 is already a great DAW platform.
• Laptop users can currently run ASIO drivers.

 

The only "advantage" of native ASIO is that the motherboard's onboard audio interface can be used via ASIO.

Most DAW users are running a dedicated audio interface... for a multitude of reasons (higher quality A/D D/A, lower noise-floor, balanced I/O, etc).

Some current audio interfaces can run sub 1ms total round-trip latency.  Native ASIO isn't going to improve upon this.

 

Apple has had "Core Audio" for years (akin to Native ASIO).

Core Audio doesn't outperform current ASIO drivers on a PC.

 

I don't for one microsecond believe that Microsoft will, much less can, write an ASIO driver that will perform better than the multi-client driver provided with my RME Fireface UCX II. There could be some useful developments like device aggregation that RME could take advantage of. I'm more interested in the improved support for MIDI and MIDI 2.0.

Edited Thursday at 04:26 PM by TheMaartian

[Starship Krupa]

2 hours ago, Jim Roseberry said:

The only "advantage" of native ASIO is that the motherboard's onboard audio interface can be used via ASIO.

Did you read the article? Pete lays out very clearly why they are doing it.

The two biggies are first, that as we all know well if we've been trying to help people on the forum, not everyone is as savvy and squared away as we are about downloading ASIO drivers.

Take the example of someone completely new to computer recording. People have been working for many years with USB devices that just work when they plug them into their computer. So what is going to alert the brand new user that they should download and install this other custom driver? Paper manual? Good luck getting anyone to read one these days.

And anyone who's inclined to say "well, that's the price of admission, we all have to pay our dues," well, nope. What computer peripheral these days, other than an audio interface, requires jumping through all of those hoops to get working optimally?

ASIO can't perform as well as Core Audio if ASIO never gets installed. What was it that Byron Dickens used to say around here whenever someone was having trouble getting their audio to work? Something to the effect that 99% of those issues would be solved if the person would just use an interface with a good ASIO driver. I didn't and don't entirely agree with him on that, but he's not the only one who believes it.

Second, and this is also huge, Microsoft want to make system audio compatible with ASIO, as in, system sounds will still get through, and according to the article, this isn't practical if they're at the mercy of the interface manufacturers' driver. My system is a multi-use computer. I watch movies on it (and often wish to take dialog samples from them), listen to music recreationally, etc.

I'm picky enough about audio that I want to use my Focusrite interface to play this stuff back.

I tolerate a certain amount of the interface getting confused about sample rates when I start up the DAW while Apple TV is running, because I know the rain dance it takes to get it to settle down, and I know that Windows is at the mercy of Focusrite's driver.

If they do their own ASIO driver, they can give that driver the ability to communicate to the OS and tell it not to switch the interface's sample rate, to prioritize traffic from certain programs, etc. As it is now, people who use ASIO do it under the assumption that ASIO will take control of all audio, and anything other than one program at a time using the driver is far from guaranteed.

I've seen people struggle plenty with driver issues, and "YouTube doesn't work when I have the DAW running." I also know that for every person who takes the trouble to ask about it on an internet forum, there are probably at least 5 who gave up on it. And that's not good for pro audio on Windows.

Also, unlike Apple's Core Audio, ASIO as it is now can't aggregate devices, and that's a disadvantage. He mentioned that they're going to see if they can get device aggregation to work, too.

As for laptop users being able to use ASIO drivers, well, they can as long as the CPU in their laptop is an amd64 CPU. Correct me if I got the wrong impression, but I gathered from the article that it's currently uncommon for interface manufacturers to supply an arm64 ASIO driver.

BTW, is nobody intrigued by the other part of the news, Windows' USB getting optimized for audio? I think that is GREAT news. The reason I still use Firewire is that I've never taken to the idea of using an asynchronous bus that was designed to connect printers, keyboards, and mice. I know that it's time tested and zillions of happy users rely on their USB 2.0 audio interfaces (I even have a small one for portable use), but I don't think Jim's main interface is USB either. Quantum Thunderbolt, if memory serves?

14 hours ago, Amberwolf said:

*nothing* is "just" a recompile for a different platform and/or cpu/mpu.

A recompile followed by a LOT of testing.

[pwallie]

15 minutes ago, Starship Krupa said:

A recompile followed by a LOT of testing.

and fixing/changing, usually

[Starship Krupa]

22 minutes ago, TheMaartian said:

I don't for one microsecond believe that Microsoft will, much less can, write an ASIO driver that will perform better than the multi-client driver provided with my RME Fireface UCX II

Well, given that nobody else can either, that's not a stretch.

It used to be that I never wanted to be stuck with any important driver that was supplied by Microsoft, but they have gotten WAY better during the Windows 10 era. Working more closely with hardware manufacturers to incorporate their drivers in Windows updates, etc.

It wouldn't surprise me if we get a situation similar to what we now have with video drivers. The Microsoft ones will work reasonably well (and reliably), but to squeeze the last bits of performance out of a device, we'll still be downloading them from the manufacturers. nVidia (and AMD, I think) offer me two different drivers for my video card, I can install the "game ready" driver if I want the most performance for gaming. Perhaps we'll have "studio optimized" audio drivers.

Then the forum people will still be able to have fun telling people that they should download the manufacturer's ASIO driver rather than just using the generic Microsoft one.😄

And hey, don't knock finally having a reliable driver for those Realtek CODECS. There are times and places where it's not desirable or practical to lug an external audio interface around just to mix or even perform.

I watched a video of Four Tet, an artist whose work I admire, do a show in New York, and he plugged his battered Dell notebook into the mixer via its headphone jack. I would not do that, but some would and do.

Edited Thursday at 04:43 PM by Starship Krupa

[pwallie]

2 minutes ago, Starship Krupa said:

And hey, don't knock finally having a reliable driver for those Realtek CODECS. There are times and places where it's not desirable or practical to lug an external audio interface around just to mix or even perform.

exactly, plug n play asio with the onboard realtek (without any nasty evil asio4all)

[Starship Krupa]

Just now, pwallie said:

exactly, plug n play asio with the onboard realtek (without any nasty evil asio4all)

As you know well, Pwal o'mine, ASIO4ALL will infect your mind, curve your spine, and keep the country from winning the war (as George Carlin used to put it about the Seven Words You Can Never Say on Television).

[Jim Roseberry]

1 hour ago, Starship Krupa said:

Did you read the article? Pete lays out very clearly why they are doing it.

The two biggies are first, that as we all know well if we've been trying to help people on the forum, not everyone is as savvy and squared away as we are about downloading ASIO drivers.

Take the example of someone completely new to computer recording. People have been working for many years with USB devices that just work when they plug them into their computer. So what is going to alert the brand new user that they should download and install this other custom driver? Paper manual? Good luck getting anyone to read one these days.

And anyone who's inclined to say "well, that's the price of admission, we all have to pay our dues," well, nope. What computer peripheral these days, other than an audio interface, requires jumping through all of those hoops to get working optimally?

ASIO can't perform as well as Core Audio if ASIO never gets installed. What was it that Byron Dickens used to say around here whenever someone was having trouble getting their audio to work? Something to the effect that 99% of those issues would be solved if the person would just use an interface with a good ASIO driver. I didn't and don't entirely agree with him on that, but he's not the only one who believes it.

Second, and this is also huge, Microsoft want to make system audio compatible with ASIO, as in, system sounds will still get through, and according to the article, this isn't practical if they're at the mercy of the interface manufacturers' driver. My system is a multi-use computer. I watch movies on it (and often wish to take dialog samples from them), listen to music recreationally, etc.

I'm picky enough about audio that I want to use my Focusrite interface to play this stuff back.

I tolerate a certain amount of the interface getting confused about sample rates when I start up the DAW while Apple TV is running, because I know the rain dance it takes to get it to settle down, and I know that Windows is at the mercy of Focusrite's driver.

If they do their own ASIO driver, they can give that driver the ability to communicate to the OS and tell it not to switch the interface's sample rate, to prioritize traffic from certain programs, etc. As it is now, people who use ASIO do it under the assumption that ASIO will take control of all audio, and anything other than one program at a time using the driver is far from guaranteed.

I've seen people struggle plenty with driver issues, and "YouTube doesn't work when I have the DAW running." I also know that for every person who takes the trouble to ask about it on an internet forum, there are probably at least 5 who gave up on it. And that's not good for pro audio on Windows.

Also, unlike Apple's Core Audio, ASIO as it is now can't aggregate devices, and that's a disadvantage. He mentioned that they're going to see if they can get device aggregation to work, too.

As for laptop users being able to use ASIO drivers, well, they can as long as the CPU in their laptop is an amd64 CPU. Correct me if I got the wrong impression, but I gathered from the article that it's currently uncommon for interface manufacturers to supply an arm64 ASIO driver.

BTW, is nobody intrigued by the other part of the news, Windows' USB getting optimized for audio? I think that is GREAT news. The reason I still use Firewire is that I've never taken to the idea of using an asynchronous bus that was designed to connect printers, keyboards, and mice. I know that it's time tested and zillions of happy users rely on their USB 2.0 audio interfaces (I even have a small one for portable use), but I don't think Jim's main interface is USB either. Quantum Thunderbolt, if memory serves?

A recompile followed by a LOT of testing.

Read the article... and it basically stated exactly what I thought it would.   

• Class-compliant ASIO driver
• Arm CPU support

I'm coming from the perspective of looking for ultimate performance.

 

I'm using an Antelope Orion Studio Synergy Core (main DAW) and a Presonus Quantum (secondary machine)... which are both Thunderbolt audio interfaces.

I chose those specifically because they're pushing the limits of ultra low-latency performance (sub 1ms RTL).

The Orion Studio Synergy Core has great fidelity, onboard DSP, flexible routing, etc.  IMO, All the best features rolled into a single audio interface.

 

I don't think native ASIO is a bad thing.  I'm certainly not against it.

For me, it isn't bringing anything new/improved to the table.  Same with ARM CPUs...

 

When it comes to CPUs, I want the fastest "workstation" CPU I can get.

Don't care if it's Intel, AMD, etc...  (whatever offers highest performance at a given point in time).

Right now, I'm running AMD's 9950x.

Certainly Laptops have their purpose... and I see why some folks are interested in ARM CPU development (lower TDP, lower power demand, etc).

Power-management and performance-management typically don't equate to top-performance.  Thus, my lack of enthusiasm. 

Edited Thursday at 05:32 PM by Jim Roseberry

[Starship Krupa]

4 hours ago, Jim Roseberry said:

I'm coming from the perspective of looking for ultimate performance.

Sure. Makes sense, and I agree with you. Regarding native ASIO and arm4 support, there is nothing to see there for bleeding edge performance.

I'm intrigued by the USB thing, though. I was really surprised and disappointed) when PreSonus stopped making Thunderbolt interfaces.

Mostly, I'm happy to see Microsoft putting effort into making Windows work better for audio production. I think this is the first big news I've heard since WASAPI, and WASAPI was some time ago.

[Amberwolf]

17 hours ago, Joseph Gresham said:

native ASIO support on Windows

I can see it now--they buy the ASIO4ALL project and build it in as a layer...like they did with certain things many years ago when they bought Bars&Pipes from Blue Ribbon Software (or maybe they bought the whole company?) so they could have something in the DirectX devkit (for what reason, I don't know) just as BRS was getting the Windows version of B&P ready (it had been Amiga-only before that).   

 

(this was the reason I ended up getting into Cakewalk, because it meant that B&P would never be available for Windows and I had to move to it from Amiga....)