Shiny graphics, desktop widgets, a redesigned Start menu … most of what you’ve likely heard about Windows Vista means little to making music. So, we’ve grilled the senior technical expert and CTO at long-time Windows developer Cakewalk in a CDM exclusive on the guts of the new Windows and what it really means for audio production.

Cakewalk’s Noel Borthwick should know the answers. He’s not only worked on getting SONAR 6.2 to work on Vista; he’s also worked with Microsoft to make sure music customers get features they need. Rather than water this down, we’ll get all the way into the technical details, but here’s the brief summary of some of the most interesting benefits, as I see it:

  1. More stability and security should help keep misbehaved drivers and malware from taking down your system.
  2. Better, more “glitch-resistent” low-latency driver functionality will improve audio performance, particularly if using the new WaveRT driver model.

  3. Better performance from the built-in audio included on Vista-ready computers and other supported built-in hardware will make your computer’s sound hardware more useful for day-to-day work. Or, in other words, you won’t always have to lug around an extra interface to fill in for your laptop’s headphone jack. (Interestingly, I’ll bet this could extend to Apple’s Intel MacBooks and MacBook Pros, which use the common Realtek audio chipset — perfect if you’re dual-booting to run SONAR and FL Studio. We should know soon.)

Later this month or early next, I’ll be ready to throw my full arsenal of favorite Windows music tools at the new OS (SONAR, Ableton Live, Max/MSP, Audition Pro, and others). In the meantime, we can get to know the OS’ core better.

Noel Borthwick is Cakewalk’s CTO and Vista expert; he has played a key role in collaborating with Microsoft on the specification of the WaveRT architecture for professional audio applications.

The main performance benefits for DAW users can be attributed to the MMCSS and WaveRT implementation in Vista.

DAW users are aware that streaming low-latency audio comes at the price of CPU consumption. Also, there is a threshold beyond which lower latency can cause glitches in audio playback. To address this long-standing issue, Vista has new features designed to aid real-time streaming multimedia applications. The new Multimedia Class Scheduler (MMCSS) and WaveRT audio drivers offer applications a more efficient mechanism to stream low-latency audio with a lower CPU hit.

Graphics

Vista now takes advantage of the Graphics Processing Unit (GPU) on video cards for many graphics-oriented functions. This can translate into saved CPU cycles and smoother audio when you have many graphics-intensive plug-ins loaded. By offloading drawing functions to the GPU, you have more CPU resources left for your audio.

Using current hardware and drivers, users can expect equivalent performance to XP on Vista with the caveat that they will not be able to run Aero if their system is not robust enough. The Aero feature in Vista provides visual effects such as glass-like translucent window elements that you can see through. Vista scales to the hardware capabilities of the computer on which it is installed, so all computers that meet minimal hardware requirements will have Aero enabled. However, depending on your video hardware capabilities, Aero features can also use native CPU. If audio performance is compromised, the user can switch off Aero from the Control Panel.

Stability

Vista has numerous enhancements that will provide more system stability.

To start, the underlying audio subsystem has been re-architected. For example, driver components that were previously required to run in kernel mode now run in user mode. What this means is that a faulty driver cannot bring down the entire operating system in Vista. Additionally, Vista now enforces security to a much higher degree than XP or previous operating systems. The User Account Control (UAC) features protect the system from malicious software, such as spyware, etc. and programs that can change critical windows settings without the users knowledge. By running SONAR as a standard user without administrative privileges, DAW users can now keep their machines safer and less prone to failure.

Audio Guts, Explained

WaveRT driver: lower CPU consumption at lower latencies

WaveRT is the new Vista architecture for high-performance, real-time, low-latency audio drivers, designed for pro audio. The streaming model for such drivers is somewhat like [Steinberg's] ASIO and [Windows' previous driver model] DirectKS, but there are important differences. The WaveRT signal flow permits direct access to the internal audio hardware buffers and sample position counters, allowing a DAW application to stream audio to the hardware in the most efficient manner possible. Direct access to buffers and sample position means no costly user mode to kernel mode transitions on each audio pump cycle. With ASIO and Kernel Streaming (Direct KS WDM drivers) these transitions were unavoidable in Windows XP. This translates into lower CPU consumption while running at very low audio playback latencies.

Most off-the-shelf Vista-logo machines that will be available in 2007 will ship with onboard audio that is WaveRT-compliant. This promises low-latency audio playback from even consumer-grade laptops and desktops. When more pro audio vendors make WaveRT drivers available, DAW users will benefit from more efficient low-latency playback.

Multimedia Class Scheduler (MMCSS): more glitch-resistant audio playback

The Multimedia Class Scheduler service allows multimedia applications to register their time-critical processing to run at an elevated thread priority, thus ensuring prioritized access to CPU resources for time-sensitive DSP processing and mixing tasks. You can think of this as anti-glitch insurance against other, non-audio-related processes running on your computer competing for CPU. Basically, with MMCSS you minimize interruptions or glitches in audio playback even when running other applications in the background.

CDM Q & A with Noel Borthwick

Peter: New validation requirements for drivers sound like a good thing as far as system reliability and security, but won’t this cause added compatibility issues? And I understand that the 64-bit version of Windows Vista (x64) won’t allow unsigned drivers at all?

Noel: You can get non-signed x64 drivers to work. You have to explicitly enable loading of unsigned drivers at boot up time of Vista x64. To allow unsigned drivers, you must press F8 at boot up time of Vista X64 and choose “Disable Driver Signature Enforcement” from the menu. As an example, we have tested the M-Audio FireWire Audiophile x64 drivers under Vista, although they are not signed. Note: Microsoft may remove this option in a future update of Vista.

Peter: What kinds of latency can we expect from Vista?

Noel: You will get the same millisecond latency in Vista as in XP with ASIO and WDM drivers. The latency is a function of the driver-supported buffer size, not the operating system. However, WaveRT drivers, when available, promise a more efficient low-latency experience under Vista. I.e., there will be lower CPU usage when running low latencies. WASAPI, as the name implies, is an API not a driver model. The new driver functionality that is relevant to Vista is WaveRT. See http://www.microsoft.com/whdc/device/audio/wavertport.mspx for more info on this.

Peter: I see that Vista includes new security measures for protecting buffer overruns from malware attacks. I’ll be honest, I have no idea what this would mean for apps and memory addressing.

Noel: Doesn’t mean much unless an application makes use of this. Most audio applications will not use this due to increased processing overhead.

Peter: We’ve gone through a lot of the technical details. The big question for many users is, how long should they wait before upgrading?

Noel: They should wait as long as they have working drivers for the components they need, which shouldn’t be very long, since most pro audio vendors are already releasing versions of their drivers that are signed for Vista. (Most of these are basically the same XP DirectKS or ASIO drivers that have been signed for Vista.) Many applications such as SONAR 6 are Vista-ready today.

Peter: I have some of my own opinions on the different Vista editions, but what version are you recommending?

A: That all depends on what features you want. A resource for sorting this out is Microsoft’s official Windows Vista Product Guide. Ed.: Noel notes that this is the best place to look partly because the non-Microsoft references are sometimes outdated; you’ll find some pretty exhaustive information on features. My personal feeling is that most CDM readers will want Vista Ultimate, because otherwise they’ll miss power features like Remote Desktop. We’ll talk more about this once I get a review copy of Vista’s final release. -PK

However, access to RAM is one of the most attractive aspects of native 64-bit computing so here is a breakdown of RAM supported in the different Vista offerings:

(64 bit OS)
Windows Vista Business: 128GB+
Windows Vista Home Premium: 16GB
Windows Vista Home Basic: 8GB

(32 bit OS)
Windows Vista Home Premium: 4GB
Windows Vista Home Basic: 4GB

Peter: Does MMCSS benefit all audio operations (e.g., if people are using ASIO drivers), or just WaveRT?

Noel: The MMCSS subsystem itself is independent of WaveRT. MMCSS is the Multimedia Class Scheduler service and affects the prioritization of application threads that are registered with MMCSS. It is up to the application to register its time critical processing threads with MMCSS. You will see MMCSS benefits using ASIO, WDM and WaveRT capable drivers in SONAR. The combination of MMCSS and WaveRT will provide the maximum benefit.

Peter: Offloading drawing functions to the GPU of course makes sense in theory, but have you seen noticeable real-world benefits?

Noel: Regarding graphics-intensive audio plugins, many plug-ins make use of transparency today. This is most commonly done using the “alpha blending” GDI operation which can be GPU-assisted. Additionally, the new Windows Display Driver Model (WDDM) in Vista makes heavy use of the GPU, so there are more graphics operations that use the GPU than before. Also, the Vista operating system itself is more graphics hungry than XP, so the GPU plays a strong role in offloading opeations that would otherwise use CPU cycles. The ability to offload graphics to the GPU can also open the door to applications and plug-in developers to create more graphically rich applications without taxing the native CPU.

Peter: Is UAC likely to then prompt users during any typical operations for music? I can imagine it would for music drivers, though would they simply approve a prompt to, say, install a new driver?

Noel: You will not be prompted for permission during normal operations of your DAW. With properly-designed UAC compliant applications, you typically need administrative privileges only to install new components such as plugins or drivers. Vista has really locked down the prohibited operations that an application can perform without administrator privileges. In “UAC Administrator Approval Mode”, the default mode in Vista even for administrators, if an application attempts to write to certain prohibited registry or folder locations, Vista will “virtualize” these operations writing to a different location instead. This can be very confusing to an end user since files are not written in the expected location. Additionally, virtualization only applies to the X86 flavor of Vista. Vista X64 versions will simply fail these operations and cause application malfunction. All these factors make it very important for applications to support UAC to guarantee a smooth Vista user experience.

CDM: Thanks, Noel! Stay tuned to CDM for detailed real-world information on Vista for audio and music you won’t find anywhere else. (I can guarantee that, because information I can’t find anywhere else I’ll get and put here — I need it, too!) I’ll be live at the launch event with Microsoft in New York (yeah, we’ll see whether Bill Gates is able to muster a reality distortion field or not, or maybe they’ll at least give me champagne), and then we’ll run Vista through its paces running a variety of music software from Cakewalk, Ableton, Image Line, and other developers.

  • http://geradorzero.com Fabio FZero

    Hm, and what about all the DRM watchdog features? I'll probably wait a *very* long time before even thinking of installing Vista. In fact, it's very likely that I'll buy a MacBook before that.

  • http://www.createdigitalmusic.com Peter Kirn

    Believe me, I'm looking at DRM. The deal is, there aren't any apparent ill effects for audio streams from audio applications in Vista. Unless you're trying to route audio from an HD-DVD in Windows Media Player into SONAR, it doesn't look like this will be much of an issue. At some point in Vista's development, there was talk of adding DRM features to pro audio apps — thankfully, that never happened.

    The only feature that really touches pro audio / music is driver validation — it's not DRM per se, but it is a component of Vista that impacts what we do. And I think we've gotten some thorough information from Cakewalk on this, including the revelation that you WILL be able to install unvalidated 64-bit drivers, which initially many had reported was not the case. This doesn't say anything definitive about future versions — but then, unless you can see into the future, that's never possible.

    We could follow up on this, but honestly, so much of that discussion has been theoretical — and often based on outdated information — that I think it'll be best to try the final Vista build and judge this on the OS itself.

    The bottom line on DRM in Vista is a lot like the bottom line with XP and OS X; despite some early (and legitimate) scares, if you really don't like DRM, run software that doesn't have it — like VLC, Songbird, and your music creation apps (well, not counting copy protection).

    The one thing I am concerned about is Vista's own validation requirements (for the OS, not media), but again, the best way to test that will be to test that OS. About that installation disc, Microsoft?

  • Anders Steele

    As you guessed, the Built-in drivers on the MacBooks do rock.

    Using a LightSnake (USB) with Garageband I can actually use the real-time monitoring with effects while playing my guitar – pretty much can use the MacBook (standard not even Pro) as a portable amp with effects.

    As of now, this just isn't possible with my Centrino 1.7 and the Window's OS built-in drivers. It will be great if this changes with Vista.

  • http://www.createdigitalmusic.com Peter Kirn

    Anders, do you mean it's not possible because of latency on Windows? I'm assuming Microsoft will update the class-compliant support for WaveRT; if you're using software (like Cakewalk's) that supports WaveRT and MMCSS, you should indeed see a major difference.

    Oh, and for what you're describing, Pro vs. standard MacBook should make no difference at all; the usual advice about fast hard drives and lots of RAM applies, but that's it. A Centrino 1.7 should be perfectly capable of low-latency performance, as well, with the same spec caveats.

  • http://blog.whats-your.name carmen

    more than enough new acronyms and creepy services and OS-level bandaids and certification processes to make anyone in the market for a new windows machine for music start to think twice

    havent even read that 'cost analysis of Vista' myself – Ubuntu, no weird acronyns…just a glance in the Swahili dictionaruy

  • http://blog.whats-your.name carmen

    also, i wouldnt hold your breath for working drivers – the reason i switched to linux, was beacuse EchoAudio failed to release win64 drivers, despite someone there personally replying to my email saying they were coming when win64 came out – and a year later, still vapor. the support still seems to be hit or miss depending on how new the product is. adding WAvRT and CCMS or whatever just makes more work for windows driver developers..

  • http://www.createdigitalmusic.com Peter Kirn

    Well, this isn't an OS comparison. Both the Mac and Linux have a number of features and deeper integration of audio functionality that Vista, like its predecessors, lacks.

    But this is a step forward for Windows, and I wouldn't judge an OS by its number of acronyms. Protecting the kernel from misbehaved drivers and rebuilding the audio system for lower latency and better performance are good things. They might not switch you to Windows from Mac or Linux if you're happy on those OSes, but they'll make existing Windows users happier. And as someone running all three OSes (not necessarily recommended), I'm glad to see some improvement.

    The proof is in the shipping OS, which I should have any day now. But I do respect the insight of the actual developers working with the system, and there are few people with more experience than Noel. That's not an argument for any one OS; it's an argument for getting the technical details from the people who built the thing.

  • Adrian Anders

    What about tilt bits? From what I hear, some hardware interactions with the motherboard under Vista may cause a small(ish) crash when voltages reach a certain threashold. Could be bad news in a live performance situation where one needs to hardware-grade stability in one's DAW. Any sort of "crash" even a small one could cause a severe mishap in a musical performance (I know, because it's happened to me a couple of times).

    I'm going to hold back until I know what hardware brands and models cause these sorts of stability issues before building a Vista system.

    It's funny that all this trouble is stemming from some shit I don't even plan on buying into. Sorry MS, Toshiba, Sony, but fuck those next generation formats. Too much hassle for only a small bit of quality improvement. Hell, I don't even own a High-def television, why on earth would I want to rip a HD-DVD or blu-ray for anyway?

    ATA

  • http://www.thebravados.com Dano

    I haven't seen (or heard) anything in Vista that surpasses Apple's OSX CoreAudio integration.

    http://www.apple.com/macosx/features/coreaudio/

  • http://www.createdigitalmusic.com Peter Kirn

    Adrian, I'll have to investigate further on the copy protection mechanisms; there's nothing obvious that will cause problems at this point but it'll take more investigation on the subtler points and the range of possible configurations. As far as the question of whether there's incentive enough to upgrade from XP, I think these points demonstrate there is — it's just the other side of the equation we need to get to know better.

    Dano, like I said, this is not an OS comparison; I doubt very seriously that many people are looking at Vista as a reason to switch *to* Windows from Mac. There is no equivalent in Vista for device aggregation, the OS X Hardware Abstraction Layer and device sharing, inter-application MIDI, or shared MIDI configurations. Also, I'm still concerned about plug-and-play USB and FireWire functionality in Vista; there are some serious issues in XP and it's not clear yet whether they've been addressed. That's not to say there's not room for improvement in Core Audio — some of its functionality has caught on more than others, and I'd still like to see better inter-app audio, which is a whole other discussion. But, as I've said before and probably will say again, Vista narrows the gulf but doesn't bridge it.

    What Vista does promise to do is make life easier and more reliable for people running audio software on Windows. And I know some Mac users who want to dual-boot Intel Macs for the occasional project who want to know how Vista compares to XP, so this isn't even just a PC thing any more.

    A full-blown OS X / Vista comparison *might* be worthwhile at some point, but I think it's pretty clear who wins. Anyway, most people aren't trying to choose platforms on a day-to-day basis, and plenty of readers here (like myself) are dual-platform, so we'll focus more on how to get the most out of each.

  • Pingback: Create Digital Music » Vista Audio Improvement Details — For Consumers, At Least

  • http://jackit.sf.net/ Paul Davis

    (reposted from the wrong thread)

    Noel writes: "The WaveRT signal flow permits direct access to the

    internal audio hardware buffers and sample position counters, allowing

    a DAW application to stream audio to the hardware in the most

    efficient manner possible. Direct access to buffers and sample

    position means no costly user mode to kernel mode transitions on each

    audio pump cycle."

    This seems wrong given what I know about OS design. Every audio pump

    cycle is initiated by an interrupt, either from an audio interface, a

    timing device or network hardware. Therefore, every cycle necessarily

    involves a kernel-to-user space transition (preceded by the opposite

    in order to service the interrupt). The only thing the new

    architecture allows is avoiding user-to-kernel space data copying.

    This has been a feature of well engineered operating systems for at

    least 10 years, and its really quite amazing to me that Windows can

    only offer this now.

    I don't want to drag CDM deep into kernel coding, but I really have a

    strong reaction when technical jargon is used to sell technology in an

    apparently groundless way.

    If I am mistaken, perhaps Noel could explain just how the user/kernel

    mode transitions can be avoiding with WaveRT drivers?

  • http://www.cakewalk.com Noel Borthwick

    Paul, the Create Digital Music article was not intended for a developer audience and intentionally did not go into technical detail about the waveRT model. I'm surprised that you concluded that it was groundless, and the article was certainly not intended to be a comparison with any OS other than the Windows platform itself. Although I have worked on Linux(as a developer for WINE the open source Windows Emulator), I don't have any direct experience with the audio driver architecture on Linux, so I'm not qualified to make comparisons with that platform.

    Have you read Microsoft's WaveRT whitepaper, yet? It has a fairly detailed explanation of the WaveRT architecture(although it misses some detail on the pull model).

    Here are some links with some more detailed tech reference info on WaveRT:

    A Wave Port Driver for Real-Time Audio Streaming:
    http://www.microsoft.com/whdc/device/audio/wavert

    Microsoft's WaveRT Patent:
    http://www.freshpatents.com/Low-latency-real-time
    0406ptan20060074637.php?type=description

    If you look through the links above, I am sure most people will conclude that WaveRT has some truly innovative performance enhancements for Windows. We have seen noticeble benefits in our tests with this in SONAR.

    To cover your questions I quote an excellent explanation from Ken Cooper, one of Microsoft's Audio Driver guru's..

    "In WaveRT we map the physical memory buffer used for audio DMA directly into the user mode virtual address space of the client. Additionally, if supported by the hardware, we map the DMA position register into the user mode virtual address space of the client. This allows the client to read the DMA position and read/write the DMA buffer directly from user mode without the need for IRPs or kernel/user transitions.

    Back in the 'old' days of Win98, et cetera, we had the WaveCyclic model for streaming audio. In WaveCyclic, the client would read the audio position by submitting an IRP and then submit audio data in another IRP.

    The data IRP would be queued in the driver. Meanwhile, the audio DMA hardware would be programmed to generate a periodic hardware interrupt.

    The ISR (interrupt service routine) would schedule a DPC (deferred procedure call). When the DPC ran it again read the DMA position and then COPIED data from the queued IRPs into the audio cyclic buffer.

    This resulted in IRP-based data transfers and an extra data copy occurring in the kernel between the IRP buffers and the audio DMA buffer.

    When we were rolling out the WDM audio driver model back then we also developed a WavePci model, which initially was not highly used because it was more complex and many of the DMA controllers of the day didn't support scatter-gather DMA very well or at all. In the WavePci model we eliminated the extra data copy in the kernel by mapping the buffers in the IRPs and submitting these mappings to the audio driver to be programmed directly into the audio DMA controller. While this eliminated the extra data copy it turned out that the management of the buffer mappings was somewhat complex.

    The WaveRT model not only eliminates the data copy and mapping management from the streaming dataflow, but it also eliminates the IRPs themselves and associated user-kernel transitions by mapping both the audio buffer and the position register into the user mode virtual address space of the client.

    Of course, something still needs to drive the streaming. With WaveRT we support both 'push' and 'pull/event driven' models. In the 'push' model a user mode thread runs off of a timer or other thread scheduling mechanism to periodically process the audio data to or from the audio DMA buffer mapped in the user mode virtual address space. Note that 'push' mode is done without stream interrupts from the audio hardware. In the 'pull/event driven'

    model we enable stream interrupts to fire at the end of the cyclic buffer, and optionally at the mid-point of the buffer. The ISP schedules a DPC and when the DPC runs it signals any registered notification events (user mode events are mapped to kernel events and fired from the kernel). Any thread that is waiting on a notification event is then woken up to process audio data, once again doing this processing in it's user mode virtual memory space."

  • http://www.cakewalk.com Noel Borthwick

    In response to the post about X64 driver unavailability, I believe there will be a bigger surge of X64 drivers in 2007. There are a few factors that will contribute to this:

    - The Windows Vista SKU's ship with an X64 version on the disk, unlike XP where you pretty much had to buy an X64 machine preloaded with XP X64, or you had to have an MSDN subscription to get the OS.

    - There are MANY off the shelf X64 capable machines now available from all the top vendors. Also X64 capable laptops are now freely available, and many vendors offer a free Vista update which will allow installing the X64 version of Vista. Vendors will have to ensure working X64 drivers for all these systems.

    - Windows XP X64 was really a beta OS IMHO. There were many non functional or unsupported features, incomplete codecs, etc. Most users who chose XP X64 were typically bleeding edge early adopters. Windows Vista X64 on the other hand is far more robust and complete in that regard.

    Many hardware companies will be far more motivated to write X64 drivers for Vista because of this.

  • Bill Ruys

    Having read the WaveRT white paper myself, I can only add that there are some truly innovative technologies coming. One particularly exciting feature of WaveRT (requiring hardware support), is the new "clock register". According to MS, this feature will allow an Audio application to sync disparate cards without the need for a physical clock link.

    All the predictions of doom are just a tad early. I think there is a lot of reason to look forward to a bright future of solid audio performance on Vista.

    Again, reading through the white papers, it does seem to me that MS has finally taken pro audio seriously.

  • http://www.thefractal.org Ryan

    i am pasting this from an earlier post about vista audio to show how incredible my audio performance has been even without waveRT .. probably due to a better kernel and the media thread priority service.

    "well.

    I have been using vista for months and its been like a dream for me.

    I tried osx on a macbook and ironically had alot of technical problems, hardware problems (metal chassis is retarded and continously shocks you.. nice one apple!) and found the software selection (even of PPC apps) to be dissmal at best.

    I went back to pc and decided to jump into vista back before RC1. The best part is that the performance and stability are unmatched. It has been my best computing experience thus far. IT JUST WORKS.

    All my software works. If a poorly written program crashes, it doesnt effect the overall system and quickly responds to a close command. I have had no blue screens and the drivers arent even released yet.

    The audio performance is top notch. I am beating 2.0+ghz core2duo macbook pros with my 1.66 core1duo toshiba tablet. I get 18-19% in the ableton live 5 performance test in live 6.01.

    I can run 64 samples for 2ms latency like never before. I can have 100 tabs scrolling through firefox at 2ms and there are no pops or clicks. Many other apps can be running in the background. I can truly do it all at once without having to shut things down to make audio perform better. Its really remarkable how much better the audio performance is. My echo pcmcia drivers are even written for vista yet and they are flawless and powerful like never before.

    The overall smoothness is incredible. no more jerking mouse when the computer is busy. its always smooth. performance is incredible. i have a 4gb sd card in my built in reader that i use with “readyboost”.. this is killer… it stores an index of the files on your main drive which makes the drive perform much much more quickly… instead of scanning the index and then finding a file it is instantly loaded from the sd flash memory and the hardrive is reading the file is less than half the time.

    really speeds things up when you are maxing out your computer."

  • http://jackit.sf.net/ Paul Davis

    Noel,

    I am utterly flabbergasted and more than a little bit dismayed. The WaveRT design (including the patent) are absolutely NOT innovative. The design you outlined and the details of the patent echo something that the ALSA driver architecture on Linux has done for at least 6 years, and large parts of it were done by the older OSS driver architecture on Linux for the 5 years before it.

    Mapping the hardware buffer into userspace ? check. Mapping control registers into userspace? check. Driver never touches the data? check.

    Now of course, one proviso is that doing this requires hardware designs that make it possible. In the late 1990's, there were not many devices that could be fully supported in this way, but by the early 2000's most audio interfaces have a design that is amenable to at least the buffer mapping and quite a lot support mapping the position and/or clock register. As is generally the case, ALSA is much more flexible in this regard than MS driver architectures, since it continues to support older hardware where direct mapping isn't possible while also supporting newer hardware that does.

    WaveRT maybe a step up from the naive models that preceded it, but it is really deeply disgraceful that MS was granted a patent on this design, and it should be challenged at the nearest opportunity. I have no issues with Windows users getting improved audio performance (we all know they need it!), but passing off these changes as some kind of technological innovation let alone deserving of a patent is not the same thing at all.

  • http://jackit.sf.net/ Paul Davis

    just so that i don't look so ridiculous, i know that for every linux audio user there are umpteen gazillion windows audio users. i know that linux is a marginal enterprise for audio (for now). it just really sticks in my throat when MS manage to get away with selling stuff as "Shiny! New!" that has been in use in other systems (yes, not particularly widely used systems) and people who should not be required to know any better buy the sales pitch.

  • http://www.createdigitalmusic.com Peter Kirn

    Only on CDM can you read a discussion between the creator of JACK and Ardour and the CTO of Cakewalk, so please, carry on.

    I'm not going to touch the patent question, because I know nothing about patent law. The patent does appear to be very specific, however; it's hard to imagine it being applies to anything beyond Windows. Someone with more legal experience perhaps can comment.

    Patents aside, I do think developers automatically earn the right to claim "innovation" when they add a new feature to software. Or, at least, they always do. My favorite moment: Digidesign put out a press release talking about the addition of groove quantize to Pro Tools a couple of years ago, a little late to the party. (That's not to single them out; everyone does it.)

    That said, while this wasn't intended to be a comparison, I think it is worth comparing operating systems on these details. Why?

    1. It helps de-mystify why performance is variable across platforms.

    2. It helps make people's OS choice an educated one. (And if we leave aside complaints, I could easily make a "pro" argument for any of these OSes.)

    3. All three operating systems have room for improvement.

    4. We live in a multi-platform world. Our server logs show a lot of people read the CDMs from Linux. (And even Solaris — hello, work machines, I'm guessing.) Dual Mac/Windows setups are extremely common, and I've heard from a surprising number of readers who run all three. We're disproportionately platform-balanced here, I think — but maybe that's an indication of the future as far as music, especially when you can buy a $600 Mac mini that will run all three OSes in parallel or multi-boot.

    I hope that someday we see Cakewalk software on Linux; I understand that the time may not be right business-wise, but the future may be a different story.

    In the meantime, I'd love to do a round table with a Core Audio, Vista, and ALSA developer. Consider the gauntlet thrown.

  • http://jackit.sf.net/ Paul Davis

    The patent claims do not mention Windows. It mentions a "computing device". I can see nothing except the terms "adapter driver" and "Wave RT port driver" that is not generic across all platforms. "adapter driver" could be argued to be generic as well, and "Wave RT" is the subject of the patent.

  • http://www.createdigitalmusic.com Peter Kirn

    I agree; it's troubling. And you're right, the patent covers adapter driver and port driver. I just don't know enough about patent law or the technologies involved to tell whether this would apply to the specific combination of techniques here or could be applied to something like ALSA. Companies are being so aggressive with their patent portfolios — and often, as here, with stuff that is technically involved — that I would imagine there's a lot of this theoretical overlap, and patents granted when they shouldn't.

    I would at least separate the patenting, marketing, and technical issues here.

  • http://www.cakewalk.com Noel Borthwick

    Talking about Linux, I haven't really kept up with this OS, although I spent a year writing some internals for WINE. Most of the clipboard code in Wine is my code :-)

    When I started at Cakewalk, I even proposed that we take a look at porting our app over to Linux using WineLib. Everyone looked at me really strange, so that idea flew like a rock at the time!

    Linux is cool but it will be some time before companies like us can justify the development cost porting to that platform.

  • Pingback: Create Digital Music » Vista Launch is Here; Developers, Help Us Get Ready

  • Pingback: Create Digital Music » Vista Launch Day: What’s Compatible for Musicians?

  • http://volikituyahoo.com Neal Down

    sorry, trouble with mu e-mail.

    Most of the industry hardware driver developers are now considering Linux?! because

    of this amazing VISTA features scare them. Shame. Unfortunately, we could all end up in Ubuntu.

    Then i can not use my Aero desktop making musick with my Wista.

  • Steve Schow

    Very interesting article. No mention of midi related improvements in Vista. I'm sad to hear that. Inter-application Midi and more performant midi event scheduling has needed improvement for a long time.

  • Pingback: Create Digital Motion » Vista Journal: Vista for Visualists - Not Ready for Primetime Yet

  • Lobiando

    Paul Davis wrote: "The WaveRT design … and the details of the patent echo something that the ALSA driver architecture on Linux has done for at least 6 years, and large parts of it were done by the older OSS driver architecture on Linux for the 5 years before it."

    Paul, if the WaveRT design "echos" something, I have to assume that it is "SIGNIFICANTLY different" and NOT "substantially the same" as the thing being echoed – which words, although you did not use, were certainly strongly implied.

    Paul Davis also wrote "just so that i don’t look so ridiculous, i know that for every linux audio user there are umpteen gazillion windows audio users. i know that linux is a marginal enterprise for audio (for now). it just really sticks in my throat when MS manage to get away with selling stuff as “Shiny! New!” that has been in use in other systems"

    Well, considering that the WaveRT design is merely an ECHO of something in Linux THAT MIGHT OR MIGHT NOT BE IMPLEMENTED AS WELL AS, OR BETTER THAN, THE WaveRT design, I see NO reason and you have give NO reason for thinking that is really not "shiny new". In fact, by saying that WaveRT is a mere "echo" of something, we have good grounds for thinking that yes, maybe it IS "shiny new".

  • Pingback: Article Feed » Vista for Music + Pro Audio: Exclusive Under the Hood with …

  • http://www.ralphbuckley.com ralph buckley

    i know this isn't a forum…but i'm completely frustrated so anyways. I have a new duo-core laptop with vista basic. My recording software is ableton live 5…I have downloaded the asio4all drivers to cure the latency problems. I'm told that this should work, in fact i can see the levels in ableton but I can't hear them, although when i switch back to the MME/Dircet X the sound is there again but the latency in recording is back or course!

    I'm told to stick with the Wave Rt drivers and i'm not even sure that i have them…

    so…is this whole episode not working because of the Vista Basic OS…if I upgrade to vista home edition will this work?

    i'm not a computer tech obviously…just wanna get back to recording…

    help

  • http://www.blessomania.com Blessing Okechukwu O

    Guys. I need help!!!!

    I have an M-audio FW410 and thankfully have downloaded the vista driver. The FW410 now seems to run well with Ableton Software that was bundled but does not produce any sound when used with Cakewalk's Sonar 6.0.

    I prefer working with Sonar as I'm not too familiar with Ableton. I would appreciate it if you could pls email suggestions to me baokere@hotmail.com

  • http://N/A Curtis Smith

    I bought a Cakewalk Music Creator 3 and i cant get it to work on my window vista, what can i do?

  • Greg

    I have Guitar Tracks Pro and it will not work on Vista. I have worked many hours trying and even asked for help from a Cakewalk tech who told me I was out of luck.

  • Jeff

    Wonderful – I have a M Audio Delta 1010LT sound card that I can't use because there are no Vista drivers available for it! My Guitar Tracks Pro 2 also will not run on Vista! Downgrading the Vista Home Premium Edition to XP is expensive and very difficult – so I sit in front of my brand new $800 Gateway computer dead in the water! I bought the new system to facilitate the creation of music – now I am just plain f—-d! Thank you Bill Gates, M Audio, Cakewalk, and the rest of you assh—s!

  • Pingback: Create Digital Music » Vista for Audio, 1 Year Later: Talking OS Plumbing with Cakewalk’s CTO

  • chirag ramani

    my male could not load my massege

  • Pingback: … nachbelichtet » Tipps zu Problemen mit Audiointerfaces und Latenzen

  • http://www.thugplay.com Cheru

    Keep the site up. It's ya boy Cheru out of Thug Town holla at me on Thug Play

  • Pingback: Create Digital Music » Optimizing for Vista: Inside the Mechanics of SONAR 8 with Cakewalk Engineering

  • Steve Sleep

    Damn, this is depressing, I need a new pc, like 6 months ago, But Vista is all there seems to be. I'm not going to go for mac. Sorry, too expensive, for trite "Indusrty Standard" sound. So What If MAC is totally great for recording what The RIAA Lawyers and Middle school brats want to hear at the mall,I don't wish to create a "Competative" or "Marketable" Musical partial abortion. That's why I've chose windows. I still use GASP! ANALOG TAPE AND USE A RAZORBLADE TO SPLICE MY REEL TO REEL TAPES FOR SOME PARTS, I didn't become a Musician to ROCK The MTV ZOMBIES, I wanted to create experimental music, and I do. and I sell, and I use Windows XP for my mastering/post/editing..HMMM Curious I've never had to worry about sounding banal and irritatingly plastic as the " INDUSTRY STANDARD" recording studios Produce. They can make all the disposable stuff they wan't. I just wan't to make interesting music If the Beatles could record Sgt. Peppers on analog 4 track, and Zappa record Absolutly Free on 12 Track analog, Why do I need Mac, I Wan't My WinXP! or was it MTV? I Just want may hardware and software to work, so I can work. Microsoft….Do Something, not for The Payola, help out the Indies, and you might just take a larger chunk of the market share away from MAC, by pricing and features that i'm sure if you can make sparkly shit like vista you can whip up a renamed XP without all the other crap and market expressly to Musicians of the Indie, and lower end, who by the way sell more physical CDs than the BIG LABLES DO, and there are a hell of a lot more of us than the Hannah Montana's of the world. MAC can Keep it's Big Boy market share, but MS might actually do something that garners the respect of the much larger Indie scene and move more units in the long run. But right now….I'm at at dilemma, Vista, too much hassle, Mac Too POP plastic, Linux? I have no clue and don't care.

    c'mon….work with us here…fuck the techno improvment theory bullshit, and ask some poor,home studiio musicians how they would like a system to function, then do it. don't get all technoid about it, just make something work….leave the sparklies for the Short bus and soccer mom market…