The iPad has proven a tablet can be a powerful tool for music. It’s also been mostly alone. Android-powered tablets have suffered from lackluster audio performance. Compound that with low popularity in the marketplace and fragmented OS updates, and the platform has largely scared music developers away. Android devices also lack the richness of the iPad’s hardware accessory support, with multi-pin ports that lie dormant, giving accessory makers insufficient capabilities. Windows-powered tablets thus far show some promise, but absent high-quality multi-touch input or thinner form factors, they’ve also been a non-starter.

It’s unclear whether it’ll change the situation, but at least there’s a new horse in the race. Microsoft unveiled two tablets last night. One appears to be, in this initial preview, a real Windows 8 computer that happens to be a tablet. That’s good news, because it ought to mean that three things you expect from the desktop Windows experience are possible:

1. You can run your favorite applications. Desktop applications like SONAR, Renoise, Reason, Cubase, Ableton, FL Studio… you know, the list goes on. These applications rely on desktop Windows, not only because they were built to target that platform, but also in order to provide the functionality you need in a music app. That brings us to the next point:

2. You get reliable, low-latency audio. The OS tools that make this work involve a number of dimensions – things like the ability to prioritize audio threads and to avoid slow, clunky OS-wide mixers. But you care about the upshot. On desktop Windows, we know you can get the sort of audio performance that makes music apps usable.

3. You can use other hardware. MIDI keyboards, audio interfaces – okay, whether or not you can use all of them, you probably want to occasionally use some of them.

Some of you will want all three of these things, which means you want a tablet that runs Windows. (In fact, some people will want that instead of an iPad, because, while it offers points 2-3, it doesn’t give you the same software choices a Windows-, Mac-, or even Linux-powered desktop might.)

But items #2 and #3 are really essential. Without them, there’s no incentive for developers to make new applications. (What would that look like? Ask an Android user or developer.) So the question is, can you get anything like these three points on the non-“Pro” tablet, too?

Microsoft’s two tablets cover two very different scenarios. One runs Windows 8 – the real deal. That tablet, called the “Pro,” is thicker, heavier, and pricier – but it might also be a no-compromise, serious audio machine that justifies extra weight and cost for musicians on the go. Microsoft says the price will be comparable to an “ultrabook” PC; I’m guessing that’s closer to US$1000 than the iPad’s US$500 sweet spot. It’s even based on Intel: think ultrabook guts in a tablet body.

The other tablet is, well, more like an iPad – ARM processor, slimmer, lighter, and with a “tablet”-style cost, says Microsoft.

The slimmer Surface – and the development underpinnings for this new Metro and WinRT business – are really the wildcards here. Now, get ready, because everything about the naming scheme here is confusing. Let’s explain.

Whadaya Want, Windows or Windows?

It’s all in the name:

Surface, not Surface. This is “Surface,” just like Microsoft’s multi-touch table Surface, but … there’s no relation. Surface is now the name for Microsoft-produced tablets running Windows.

Windows RT, not WinRT. Windows RT is the name of the new version of Windows 8 designed to run on ARM, which has some restrictions in its hardware APIs and application model that differentiate it from Windows 8. Windows RT includes support for WinRT, the application development model for Metro that’s also on Windows 8. Windows RT – an operating system. WinRT – an application model that’s part of Windows RT, but also Windows 8, the, uh, Big Momma one. (Got that? Didn’t think so. CNET has a go at explaining this in more detail.)

Windows 8 applications, but not Windows 8 applications. Here’s where things get more critical. Microsoft touts “Windows 8” applications on these tablets, but they mean two different things. There are essentially mobile tablet applications, built for ARM, powered by WinRT. Then, there are the Windows 8 “Desktop” applications.

WASAPI, but maybe not WASAPI. It may sound like we’ve entered a sushi restaurant, but stick with me here: this is about whether audio developers can build viable apps for this platform. Various readers have pointed out to me that WinRT (and applications built in it for this Windows RT tablet) does support Windows’ WASAPI stack for sound. The problem is, it’s a specific mode of WASAPI – Exclusive mode – that some developers believe is necessary to deliver lower latencies you need for many music apps. Or, at the very least, WASAPI alone isn’t enough; the question is how audio signal is routed between the application and the user, and whether it has to compete with other threads or is routed through a mixer in such a way that adds latency overhead. Human translation: The ARM tablet edition of Windows 8 may well be crap for sound, possibly even more so than Android. This is a longer discussion, though, and the reason I keep qualifying these statements is that we’re looking for more detailed, more finalized information. But the alternatives are this: either these tablets add the high-performance features desktop Windows has, or Microsoft is doing exactly what Google and their partners did: they’re leaving creative music and high-quality sound to be the exclusive domain of Apple.

USB, but not USB. Then there are the USB ports. There are USB2 ports on the basic model, and USB3 ports on the Pro model. But will those support class-compliant MIDI, like the iPad does? The Pro model should, if it really is running the full version of Windows 8. The ARM model probably won’t, unless we get some new information. Other hardware could be supported if you can install drivers. Again, it seems likely the Pro model will allow you to do so, and the other tablet won’t.

Sound, but not good sound. Windows isn’t perfect – indeed, there’s no shortage of griping about any of the major OSes. But it does on some level solve some of the needs of musicians. The question is whether the “Windows” on these tablets will do the same. In a way, it’s even riskier than what Apple or Google has done. The iPad doesn’t run Mac OS; Android, too, runs something new. But Microsoft is calling this a “Windows” tablet. Musicians may find that that doesn’t mean what it’s meant in the past.

There is no reason a tablet can’t deliver low-latency, reliable audio for music applications. In fact, if people who knew something about audio ran the world, I think you’d see much more rigorous specifications for sound in general. Even telephony applications benefit from better performance – high latency and stuttering audio causes problems in voice calls. It’s something an average user is likely to notice in real-time applications, whether or not they’re a music pro.

We’re talking primarily about the ARM tablet here. If it delivers 100 ms latency in real-world applications, it’s a non-starter, and it’s time for those of us who are music lovers to start giving Microsoft and Google a piece of our mind. If it can do some of the things the iPad does – like give developers solid audio APIs and usable latencies, and, if we’re lucky, hardware support – then it’ll be worth talking about.

There’s more reason to be optimistic about the Pro tablet.

Surface Pro, Surface Amateur?

All of this can be summed up in three questions:

1. Has Microsoft got a viable alternative to the iPad in its slimmer Surface tablet? Can it hold a tune, or is it another tone-deaf entry that guarantees Apple will be the only company who makes stuff musicians want?

2. Does the Pro tablet deliver the full Windows experience?

3. Will WinRT development on either tablet let you make good music apps?

Signs look petty good for the Pro. And that makes yesterday’s announcement a potentially good thing. It wouldn’t be the first tablet to run a desktop OS, but looking at the design, it might be the first on you actually want to use.

But that could still be a mixed bag. With the iPad in the market – and plenty of good Windows, Mac (or, if you like, Linux) laptop choices out there, splurging on a tablet just to run Windows may not appeal to everyone. And even the Pro’s appeal is dependent on apps that are fun to use with multi-touch support, since most of them have been developed for a mouse and keyboard. (You could simply make desktop apps that use touch APIs, of course. See: Usine. But that would depend on the Pro tablet becoming a big sales hit, presumably.)

That means a whole lot still rides on finding answers to all three of these questions.

Even the absence of information is itself frustrating. The reality is this: Apple has a fantastic $500 tablet you can buy right now, and it’s got desktop-class audio support that makes responsive music apps fun to write, surprisingly-decent support for audio and MIDI hardware, and an impassioned audience that loves buying apps.

So, uh, yeah: that means the onus of proof is on Microsoft, not Apple. We’ll be watching.

Clarification: To be clear, high-latency performance on WinRT-based (Metro) apps would impact both tablets. It simply seems that only the Pro tablet lets you produce desktop apps. Yes, as commenters note, you’d want an app designed for the tablet, in order to justify using a tablet at all. But there’s no reason a developer has to use Metro and WinRT to make a touch-friendly app. Desktop Windows apps can support touch; it’s simply a matter of designing your user interface around that paradigm.

Oh, yeah, and one more question: Anyone else wondering what the bootloader is like on these two tablets? Linux, anyone? (It runs on ARM and Intel without issue. Cough.) So far, yes, it looks like at least the non-Pro tablet has a locked-down bootloader. Unknown: whether the Pro will, as well, and whether it will be possible to “jailbreak” the tablets to run alternative operating systems.

  • Korhan Erel

    At the moment, it seems like the only software that would shine on a Windows 8 tablet would be Audiomulch (, with its fantastic Metasurface feature, which allows for interpolation between an unlimited number of snapshots of the whole Audiomulch document (normally with a mouse/trackpad or a MIDI controller and in this case, simply by touch).  

    • Mulchy Mulch

      AudioMulch = the program that could be the greatest thing in the world but will never be more than what it is cos its creator doesnt/cant/wont see it or do anything about it.

      I love it so, but gave up on it long ago.

    • Grumpy

      An upgrade of Audiomulch is currently in beta test.

      Upgrades are done, not as frequently as most software but I see that as a good thing.

  • Fabio FZero

    Ableton would be pretty cool with a touch interface – specially if they add multitouch support.

    • Stefan Rijkse

      i’m using a touch screen on my pc with ableton, it works nicely but i do think that with the smaller screen (i have 24 inch monitor now) it might not be ideal working with ableton using touch, even if you can zoom in. 

  • Stefan Rijkse

    you’re going to need gui’s that are specifically designed for tablets with all these windows apps. it’s nice if you can run all sorts of software that we’ve all been using for years on the windows desktop. but almost all of them are not meant for a small touch screen and won’t be practical at all. defeating the purpose.

    of course you can use traditional input devices with the tablet, mouse keyboard trackpad… but then what’s the advantage over a laptop. we allready have that.

    the succes of the ipad is largely because of the great new software that’s being developed for it. apple providing the proper infrastructure for this, as said in your article.

    i think the windows tablet can only rival the ipad as music making tool, if they have the same content of TABLET applications.  

    • Peter Kirn

      I agree. However, while I have some concerns about Metro and WinRT, I should add – there are APIs for supporting touch in conventional, desktop Windows. So it seems possible you could simply make a desktop app. It’s going to be a niche application, for sure, unless sales of the (non-RT, Intel-based) tablets take off, but I could see someone doing it.

    • Philippe Pascal

      This is my main concern.

      Mouse/keyboard apps is THE main reason why previous tablet Windows PC never succeeded.They exists since Windows 7, witch is the first multitouch Windows.
      Using desktop apps on them is like using a cursor ten times bigger 😀

      If at least Microsoft could create some ergonomic tricks to make them better on tablet…
      Like autozoom to see the options bigger when you move your finger on them,gestures linked to keyboard shortcuts…simple things like this i never seen on them.

      For now, Usine still my only hope as multitouch app on Windows 😀

  • kent williams

    This is all about the fundamental divide between what serious and casual users need. My wife cares about storing and editing the digital photographs she takes, e-mail and facebook. She’s using a 5 year old Dell laptop with an 80 gig disk, and I’m pretty sure she hasn’t cracked 25 gig of disk use.  She’d be perfectly happy with an iPad or a Tablet PC — though she wouldn’t be crazy about learning a whole new user interface.

    I make music with my PC, and I’m not looking forward to the dominant consumer OS and hardware choices being less hospitable to the software and hardware I’ve invested in.  I read with horror that the NI audio interface I bought for my laptop would become useless if I upgraded to a newer USB3-only laptop.

    And at my day job, where we do scientific image processing, I shudder to think what the consumer OS people are going to do to screw up our environment.  We already have to deal with Apple and Windows wasting our time with their own bespoke gratuitously different development environments.  We’ve wasted man-months alone on the various platforms not being able to agree what the end of a line of text is!

    What I’m left with is the sinking feeling that anyone doing serious work is always going to be considered last when it comes to commodity computing.  There’s been a revolution in applied computing that came about because commodity-grade computers became powerful and inexpensive.  Many-CPU Linux clusters are revolutionizing scientific computing.  It’s sad that Apple and Microsoft continue to work at cross purposes with real computer users.

    If they’d just ask the engineers, they’d consider a different approach. A robust, standards-driven OS is the perfect foundation BOTH for casual computer usage and serious computing, and that’s true BOTH on desktop/laptop/server machines and portable touch screen devices.

    • Peter Kirn

      Well, this is an easy explanation for sound, though, and perhaps a separate article.

      Low latency concerns, audio prioritization, matters only to one “niche” demographic:

      People with ears.

      If you don’t have ears, or, alternatively, if you exclusively use computing devices on mute and never with any kind of interactive sound or telephony, then you’re fine.

      No, the problem isn’t the divide between advanced users and everybody else. The problem appears to be a divide between the people who know something about audio engineering requirements in making the *system* for those users, and those who don’t know squat.

      I’ve noticed one particularly circular argument. You’ll regularly see references to telephony users noticing round-trip latencies of greater than 250 ms. Therefore, there are specs that call for 150 ms latency one-way. 

      But … wait a second here. What’s the source for those minimum round-trip latencies that are supposed to be perceptible? You’ve got it – existing specs. It’s like the old fallacy of people imagining that the human eye is only capable of seeing movement of less than 1/30 of a second … when in fact they’re referring to the 30 fps film rate, and human perception is capable of far greater accuracy.

      These discussions get even more circular. Back to the telephony discussion, we’re talking a potential *breakdown of communication* between two people once latencies hit the quarter second mark. But a lot of hardware then uses that as acceptable latency for the *device* – before you begin adding transmission latency. Worse, you’ll sometimes even see *one-way* latencies of 100-250 ms, when the mention above is dealing with round-trip.

      I mean, are we really talking about what any users want, or are we talking about tech makers lowering the bar as far as they possibly can?
      I’m not nearly as concerned about the multiple standards here as I am about the lack of quality. Developers aren’t upset spending those human-hours on Apple’s development environment precisely because it more often results in software actually working, versus spending those hours banging your head against a brick wall.

    • papernoise

      I totally agree with you on this. I also have the feeling that big software houses care less and less for real pro users. At least in certain fields that have become more open to the casual, semi-professional user. While I like the idea of making software more accessible for people who are not top-notch professionals this has also several negative effects.
      One negative effect is that software gets dumbed down. At the studio where I live we had to switch back to Avid from Final Cut because the latest version was just not suitable for our standards. While some people (I’m thinking about Editors working on their laptops at home, or people how shoot video on a semi-pro level) Final Cut X was a great upgrade since the user interface is clearly better but for us it just would not work.Another example is how Photoshop gets filled up with functions clearly aimed at the semi-pros, while the software is still filled with horrible bugs and things that make the user experience totally terrible.

      The same way, since OSs are not making money with specialized pro users, they get filled with useless eye candy that usually don’t improve performance, and functions that make these systems more intuitive and easier to use but lock you up in a golden cage. The whole iPad trend is clearly heading us in this direction.

    • Peter Kirn

      Well, I think this is perhaps an oversimplification.

      On the iPad, to me the issue is that they copy some of the superficial elements but not the engineering work on things like audio. It’s a bit like the iMac fiasco, when everyone copied the translucent plastic. They missed the stuff beyond the surface.

      On Final Cut, Photoshop… who knows. Adobe is introducing some pretty hefty features for pros, and often in narrow niches. It’s a mature product where they’ll find any reason to get people to upgrade.

      Apple had the opposite problem with FCP X. They shipped a product that I think wasn’t aiming at casual users so much as it was shipped before it was ready. When you have a look at the multi-camera support and audio sync stuff they added in the most recent FCP X update, it’s clearly of interest to pro users. So, it’s the reverse of Adobe CS: because they rebuilt it from scratch when they could no longer rely on legacy QT API code, they wound up shipping something that wasn’t entirely mature.

      In other words, I wouldn’t read *too* much into these trends. This is about engineering problems, not marketing or different classes of users.

      And when they get it wrong, as with sound, it can quickly become wrong for everyone. That ought to be the lesson of the iPad – advanced functionality as far as things like battery life engineering and high-quality visual and sonic APIs winds up appealing to everybody.

    • bam!

      I agree, and I think that some of the frustrating limitations of some of the new Windows 8 stuff are really in the same “not entirely mature because rebuilt (sort of) from scratch” category as well.


    i’ll spend that $1000 if it runs iOS, OSX, win8, and Linux and allows me to connect my Rane+Serato  box.

    • Cocovyt

       The Pro will probably do 3 out of the 4.

  • papernoise

    Usine would be probably cool on the pro version! But I still have my doubts… theoretically if this tablet become a hit, software houses would start to build touch-friendly versions of their DAWs and that would be nice. The problem is, I don’t see how Microsoft can succeed with such a shaky marketing strategy. They are not even able to name their products properly…

    On the other hand, they could fail, but still launch a trend, and others might be able to succeed where they don’t So let’s just sit back and wait, and see what the market comes up with :)

  • Greg Miernicki

    Wait a minute… so this has an ARM processor… not x86? It also runs Windows RT? Doesn’t this mean that currently, there are about 0 apps for this devices? You can’t run all your old Windows applications that were compiled for x86, like Reason, Cubase, etc….

    • Peter Kirn

      There are two tablets. One has ARM, runs Windows RT; one (“Pro”) has Intel, runs “Windows 8” — the full version.

      There are zero music apps for the Windows RT device at the moment, yes, more or less. Whether that changes is dependent on the other factors I mention in regards to the non-Pro tablet.

  • Philippe Pascal

    @Korhan Erel
    Audiomulch, yes for sure.

    But you forget THE GUI killer DAW on Windows :
    Sensomusic Usine

    You can create your own interface from scratch, like Lemur.
    You can automate anything, even interface look itself.
    And more than Lemur, because it is a full DAW will VST support, Rewire, OSC, custom scripts and more 😉
    I t can as used as VST inside another host or standalone, or OSC interface (Ableton anyone ?)

    For now, this is the only audio application on Windows already supporting multitouch, gesture and such 😉

    Look at the vids on this page :

    • PaulDavisTheFirst

      you might want to note what usine’s author says about its relationship with “a full DAW”.  its a very cool program. it does a vast array of cool stuff. but even the author says “It’s not an in depth multi-track recording and wave editing software. It
      doesn’t replace software like Steinberg™ CUBASE® or Magix™ SAMPLITUDE®,

  • erja

    A keyboard with accelerometer could make for better soft-synth control from the computer keyboard.  But I really wonder how usable this will be on a lap as opposed to a table top (not the most common use case for a tablet IMO).

    For audio, I’m skeptical that it’s constraints can make it better than a Macbook Air or even an iPad.  For a desktop-class music app, I don’t think I want anything slower than my MacBook Air’s i7 (even though I use the power-sipping Reason) and I suspect that it would demand enough power to be tethered to a power cord (in which case, an ultra book makes more sense IMO).  And they haven’t even started to play catch-up with iOS when it comes to tablet optimized music creation.  I’m really curious to see what iOS 6’s announced “multi-route audio” and “inter-app audio” mean for us – they sure sound like something that could open up real audio processing and soft-synth options (possibly before Windows 8 even arrives).  If MS’s history is any indication (leaving basic pro audio functionality to outside developers) I’m not counting on the ARM Surface being too capable in the music department.

  • Robert Halvarsson

    It’s interesting.

  • Robert Halvarsson

    It’s interesting, but still to little info to really get into the nitty gritty (understandable since it’s some time before they plan it to hit the market).

  • Noel Borthwick

    One clarification. The ~100 msec latency limitation applies to WinRT aka Metro. 
    i.e even the Pro Surface tablet will be subject to this latency when running Metro apps. IOW the limitation is not the ARM version of Windows, its Metro due to its forced WASAPI shared mode audio stack.
    Desktop apps on the tablet should not have this latency as far as I can see (hopefully). 

    • Peter Kirn

      Agreed, yes – sorry, I try to get that across; the appeal is desktop apps running on the Pro Surface. And despite the comment thread here, I see no reason why software couldn’t use desktop APIs to make a touch app; that’s an interface design question. The APIs are all there. Usine is a great example.

      Of course, what remains variable even then is what the audio hardware will be like and whether, to put it bluntly, the hardware and drivers will be any good.

      And whether the Pro tablet alone would hold enough developer appeal that anyone would bother writing touch-friendly desktop apps – other than Usine. 

    • Dustinw

      And whether the Pro tablet alone would hold enough developer appeal that anyone would bother writing touch-friendly desktop apps – other than Usine. ” 

      — but the thing is, MS isn’t trying to own the Windows Tablet market … instead they are trying to kick start the Windows8 eco-system, they won’t under-price Acer, Asus and whoever else makes Windows8 tablets. Microsoft doesn’t care if the Surface sells, it only cares about Windows8 unit sales.It might take a few years, but Windows8 will have touch baked-in and those all-in ones that Dell sells and the various third party-tablets and touch screens, and new convertible laptops will be come increasingly popular … and large chunk of them will ship with the Intel version of Windows 

  • Nay Seven

    Yes, Usine will be a good candidate for this tablet, now some infos are missing like tactile technology used and number of point ..if it ‘s at least 4, can be interesting .And I’ve yet tried Usine on Win 8 preview without problems , now we work on the version 6 and this one will be simplified and more “object” wait and see .

  • filiphnizdo

    I’ve been using the Samsung Slate 7 running Windows 8 as my main PC since February (mostly got it for music use).  I’ve mostly just used Bidule and Renoise so far so not many touch features but I’m really looking forward to trying Usine out properly (I’ve only tried the demo so far).  Some better Metro and touch specific apps would be brilliant.

  • vpsaxman

    Not being a developer, there is something I don’t understand: there were a few Win 8 beta versions released to the public, right? Couldn’t someone get a definite answer as to whether or not it’s possible to create low latency audio apps in the metro environment?

    • Peter Kirn

      Yes, and right now that definite answer is no. But I’m trying to get a more complete version of that answer.

    • vpsaxman

      Thanks Peter! Keep us posted. Again, I’m not a dev so I’m very limited to how I can help but this is definitely a critical subject for a lot of us.

  • West

    “UPDATE: I have received confirmation from Microsoft that WASAPI exclusive mode is indeed supported even for Metro apps (via C++ only). This is great news since it means that Metro apps can in theory have low latency support just like desktop apps. If this is indeed as good as desktop mode, audio latency in Metro should be better than Android and potentially on par with IOS. Of course until we do some real world tests on Win8 tablets we won’t know for sure so stay tuned for future tests. See sidebar below for the full Microsoft quote on Exclusive mode. “