iproducer #4 – Unlink and Conquer + Bonus from Gustavo Bravetti on Vimeo.

Exclusive to CDM (and Vimeo), Live guru Gustavo Bravetti offers a deep tutorial in sound creation in Ableton using “unlinked envelopes.” He assembles quasi-modular routings of effects into one another to create some unusual sounds. I think there are some inspiring techniques here, even beyond Ableton Live – I’d watch it even if I weren’t a Live user.

Gustavo writes:

“This workshop demo video is about using Live’s unique clip unlinked envelopes and Ableton’s effects to easily create elaborated sequenced sounds as well as a “how to” create a bass, a tuned bass drum, a clap and hi hats.”

So, what’s the beauty of the “unlinked envelope”? By separating the envelope length from the length of a sample, you can take a sample of one length and modulate it in time independently with an envelope of a different length. A short sample can warp in a longer block of time for more variety – a 1/16th note sample, for instance, transformed over a bar or two. Add the ability to route the envelope into effect modulation, and you can start to think extreme sonic freakiness. Here, that serves Gustavo’s own electronic palette, but it could suit a variety of musical aims.

You can also go the opposite direction: you could take a long sample and modulate it with a short envelope. For instance, you could select a 30-second oscillating texture, and apply a very short envelope, like a single sixteenth-note peak, in order to create an evolving high hat.

You can think of the unlinked envelope as a modulation source for a variety of sonic targets. It could be a short sample, a chain of effects, or a synth. Here, Gustavo sticks to built-in Live devices, but you could easily add your own plug-ins of choice – especially now that Live has added the ability to more easily choose which third-party plug-in parameters you wish to target.

Part of why I say this is about more than just Live is that you can begin to see how thinking in envelopes can make your musical content more dynamic, a technique you can apply to any environment that allows this sort of routing. It certainly illustrates the power of Live in a way that I think a lot of people miss.

To watch the full HD version, head to Gustavo’s Vimeo account.

I’m going to be organizing a compendium of our best Live tutorials soon; if you have external tutorials you’d like us to link up, let us know.

Previous Gustavo-fueled mayhem:
Ableton Live Follow Actions, Dummy Clips, Making Snares
DIY 808, IDM 101
Gustavo Bravetti Show Us How To Glitch out Ableton Live
Hexagonal Sequencer with vvvv, MIDI, Ableton, and Soon Wii, Camera Input
Live + FM8 = Drum Kit Love: Free FM8 Drum Kit Download

Out of the box, Akai’s APC40 has some lovely features for plug-and-play control of Ableton Live, with clip triggering, track control, device control, and dedicated buttons for command shortcuts. It also sends and receives standard MIDI messages for every last button and encoder. But what if you still want more? What if you need more controls to do multiple duties, or get bored with simple clip triggering and decide you want additional interaction? Enter the hackers. Already, using MIDI, clever APC40 users are squeezing more function out of this box. And while it isn’t solved yet, there are some clues to the infamous hardware handshake – a System Exclusive string exchanged between the APC and Live that locks certain Live software features to the APC and not to other hardware you might like to use.

Manual MIDI

Before we get too fancy, for power tricks, your first stop should be Akai’s own site:
Tips and Tricks June – APC40

Live allows you to manually override the APC’s dynamic control assignments using the standard MIDI Map. Let’s say you don’t use headphones for cueing. You can select the MIDI Map, pick a control to which you want the Cue Level encoder to be assigned, and you’ll manually assign just that control – the rest of the dynamic template remains in place. Akai has some tips for scrolling through scenes, selecting scenes with one of the two footswitch jacks on the back of the unit, scrubbing and nudging clips, fine-tuning tempo control, and more.

monome Emulation for APC40 and Korg padKONTROL

Our friend Michael Hatsis of trackteamaudio has been hard at work in Max/MSP patching an emulator for the creative patches for the open-source monome hardware. (Thanks on Twitter to ruaridhTVO, too.) By translating from the (and, cough, superior) OpenSoundControl messages the monome supports natively to MIDI, the emulator supports not only the APC but Korg’s padKONTROL, as well. This opens up the use of the APC for creative microsampling and other tasks.

Video demo at top (updated late Sunday night, so if you saw this over the weekend, here’s a tighter version).

Direct download:
http://www.warperparty.com/datter/Monomulator0.9.zip

Forum discussion:
http://forum.ableton.com/viewtopic.php?f=1&t=117307&start=0

And be sure to check out the Java- and Python-powered open-source library for the monome on which Michael’s work is based:
net.loadbang.shado

You’ll find plenty of documentation in Michael’s download, and the hope is that this is just the beginning — you Max patchers out there (and Pd, if we can port this) can keep hacking on it and try out some new ideas. One reason you might want to keep hacking on the padKONTROL is that you could find uses for velocity – unlike the monome and APC, Korg’s 4×4 drum pads are velocity sensitive.

APC40 Customization, Performance Tweaks

This is the best video I’ve seen yet with the APC40. The APC itself is strikingly limited for a MIDI device, without even basic abilities like preset switching or the ability to change default MIDI assignments. But because it’s connected to a computer, if you’ve got some MIDI programming skills and time on your hands, you don’t have to stop there. Stray411, the creator of the brilliant nativeKONTROL software for the padKONTROL, Korg nano series, and Akai MPD32 has turned his MIDI hacking superpowers to the APC.

First, he demos the manual remapping technique. But from 1:38 onward, he remaps and reroutes messages via Bome’s MIDI Translator, commercial Windows (and now Mac) software for more sophisticated mapping of MIDI messages. This allows him to create his own dynamic template for control that applies more functionality to the onboard hardware controls on the APC.

I’m not even sure you’ll want to do this – it can make for a more complex control scheme – but it’s impressive just seeing the ideas out there.

Note that this sort of thing should also be possible via any software that does MIDI input and output, including the free Pure Data (Pd) patching environment and Max for Live when it ships in the fall. (I’m not entirely sure how intercepting MIDI with Max for Live will work, though, especially with the hardware handshake to contend with… more on that in a moment.)

Korg fans (and Akai MPD owners), be sure to check out:
http://www.nativekontrol.com/
And see the nativeKONTROL videos:
nativeKONTROL YouTube Playlist

…and really, that deserves a separate post.

MIDI for Lights

Akai left out the MIDI Implementation that’s traditionally included with MIDI hardware (cough), but it does use standard MIDI messages both for outgoing control data (when you move an encoder or press a button) and incoming messages (like Live switching a light from off to amber to green).

I’m going to publish the Missing MIDI Implementation later this week here on CDM, but to get you started, Danny P on the Cycling ’74 forum has deciphered the toughest part – the messages that light up the clips:

Midi confusion with APC40

And even better, CerebralNektar (of the nativeKONTROL) project has already built a full-blown Max/MSP template for the clip grid:

OK, let’s hack the APC

The Hardware Handshake: First Clues

Ableton has worked with Akai to add a specialized MIDI implementation to Ableton Live, using a set of System Exclusive messages to prevent the hacker community from emulating certain APC features in other hardware. Specifically, this includes several abilities:

• Using bank buttons to trigger different sets of clips in a larger set, without running out of MIDI messages to do so
• Providing a red rectangle overlay to show which 8×5 (40 clip) array is selected in Live
• Sending MIDI messages for clip status back to the hardware (thus lighting up the lights)

Now, granted, as implemented this functionality may be of limited to use to hardware that isn’t the APC40 – particularly because it’s hard coded for an 8×5 grid of buttons, which is a non-standard size. But having talked even to some passionate fans of the APC, I know it’s bothering a lot of people. I think there are several reasons why.

In short, it’s the first time I know of that standard MIDI messages were used not for the purpose of interoperability, but to actually prevent you from using your own hardware. The APC itself won’t work properly with Live if this string is interrupted (and you’ll see complaints on the user forum in which people are having related problems). Also, while the functionality here is hard-coded to the 8×5 array on the APC, that raises another question – why not make a generic implementation for other hardware? Why not a rectangle that shows a 4 x 4 grid for hardware like the Akai MPD series, Native Instruments’ Maschine controller, and the popular Korg padKONTROL and M-Audio Trigger Finger?

In the meantime, cracking the handshake could be useful for owners of the monome or upcoming Ohm64, even with their 8×8 grid – you can use the last three rows for shortcuts.

Michael Hatsis writes (consistent with what I saw running MIDI through MIDI-OX):

from what I can see both the APC’s 2nd string and Live’s 3rd string have 24 bytes, both with bytes 8-23 different each time
- There’s your handshake…

I have set up two max patches that parse and output the SYSEX sent by both the APC and Live. the one called handshake only outputs the unique bytes for both the APC and Live to the Max window. There are more details inside.

Have a download, folks – this gives you some of the MIDI to look at even if you don’t have an APC40:
http://warperparty.com/data/handshooken.zip

I couldn’t figure out what the algorithm was, but then, I’m not terribly good at that sort of thing. So we’ll be interested to see if anyone else can sort it out.

By the way, this is sent in the clear as MIDI messages. There’s no real reverse engineering here. It’d be like printing the secret password for your speakeasy on a billboard at a rush-hour bottleneck on the 101 highway. Nor is there any kind of theft involved. These are capabilities built into Ableton Live, which Ableton has effectively blocked from use with this System Exclusive communication.

In a matter of days since the hardware shipped, the APC40 user community has already done some incredible work. This to me makes a powerful argument for openness – and it says that the same community could do even more if hardware and software used more intelligent communication schemes like OpenSoundControl instead of being locked to the limitations of MIDI.

A Video to Close us Out

To close, here’s a reminder that part of why we expend this much energy on controllers is to make them personal instruments for ourselves. Here’s a YouTube demo that shows people can make the APC, well … shake.

If you think there aren’t still exciting things happening in synthesizer design in the age of software, you haven’t met people like Stephan Schmitt. Schmitt, founder of Native Instruments and the “mastermind” of Reaktor, could be seen as a mad sonic scientist behind NI. When I met him for dinner in Berlin in October, he had brought along a stack of signal flow diagrams and Reaktor screen grabs in plastic sheet protectors. I knew something brilliant was coming.

Native Instruments calls Schmitt’s latest creation “Spark,” but I like to think of it as the Schmitt Box – like a mysterious, powerful invention from a designer who loves to experiment. Stephan has been evolving the instrument in Reaktor through some 160 iterations. He uses foot pedals to modulate the sound live, and rails against the evils of dull, repeating LFOs. Spark ships as a Kore soundpack, so for US$59 you can fire it up right away and start playing, even without knowing how it works. Even better, though, is if you have a copy of Reaktor 5, because you can use the full-blown UI seen here to design your own sounds or even dig into the plumbing of the patch beneath. (It’s still worth looking at the Kore sound presets, because they’re consistent with Stephan’s approach of designing the sound for live performance.)

Reaktor Spark [info, download]

I think it’s telling that, while Stephan’s emphasis is on playability, he brought those signal diagrams. It’s tough sometimes to put the nuances of synths into words. NI’s own description, that Spark “combines powerful subtractive synthesis with a sophisticated array of internal feedback loops and various other special sound shaping features” doesn’t quite cover it.

So, I instead asked if we could use Stephan’s own words to describe the new instrument. The following is an excerpt from the guide he wrote for sound designers working on presets for Spark. (Scroll to the end for full diagrams of the signal routing inside, fellow geeks!)

If you’re new to this stuff, this will give you some insight into why Spark sounds the way it does. And if you’re a synth designer yourself, I think you’ll really appreciate Stephan’s personality and approach. And it encourages me that, even with a lot of repetition of basic elements (subtractive synthesis, a feedback loop), there are still many possibilities for personal, idiosyncratic instruments to explore.

Here are Stephan’s thoughts:

In the beginning, the design was inspired by "Weedwacker", a Reaktor instrument created in
the year 2000 by Siegmar Kreie. The main concept behind the Weedwacker was the feedback
of the filtered signal to the pulse-width modulation input of a pulse oscillator. The result was a
surprisingly-complex behavior of the simple one-oscillator/one-filter structure. The rich and
organic sound was appreciated by many users. Another influence was the Evolver from Dave
Smith, a hybrid concept with some interesting feedback paths.

The structure of Spark reflects my personal preferences in regards of synthesizers. Instead of creating a full-blown mega/multi-purpose synth, I try to keep it as small as possible and with a
special character.

A main goal is to allow dramatic real-time influence on the sound source. The synth itself offers
only a few simple automatic movements (2 envelopes and 1 monophonic LFO). Like a natural
instrument, it needs to be played expressively and therefore stays a challenge for the player
and the sequencer programmer.

I deliberately do not make use of the following techniques:

• samples
• complex waveforms or wave-tables
• complex shaping curves
• noise or random (except a randomization for pitch at note-on)
• multiple and extensively routable LFOs (the LFO is intended to replace human control or an automation curve)
• multi-breakpoint envelopes or step-sequencers
• multi-oscillator structures (like an FM matrix or additive osc bank)
• pitch envelopes

The idea behind this is that the signals are generated and modified by a small number of very basic mathematical functions. Applied in a certain structure, they can create complex signals that might have their very own nature, behavior, and sound character. That is what fascinates me in synthesizers.

The two oscillators deliver two very basic waveforms which are contrasting and complementary: Pulse and Sine. The Pulse has a bright sound and a wide spectrum of overtones which can be filtered. The Sine has a very soft sound and a narrow spectrum that can be widened up by FM, amplitude modulation or wave-shaping. By using feedback loops which contain linear filters and non-linear functions, complex and quasi-chaotic waveforms become possible. This is why the sound spectrum includes more organic and aggressive timbres than a classical analogue synth.

In some settings, the structure behaves similar to a physical modeling synth. The Oscillator section becomes the exciter, while the feedback loop around the filter, shaper, and delay behaves like a resonator.

My favorite range of sounds is percussive – similar to pianos, mallets and plucked string instruments. Since I was always fascinated by electric guitars, you can get a lot of distorted and feedback sounds out of it.

By its sound and playability Spark might be suitable for:

• Rock and Blues
• Jazz, World Music
• Noise, Free, Experimental
• Industrial, Dark Wave, Gothic, Psychedelic
• IDM, Elektro and EBM
• Sound Tracks
• Sound-Design (where the output is sampled)

In my music projects, the instrument has replaced the DX7-II (+ external Effects). Spark gives me similar aggressive, non-harmonic spectra, the fast percussive response, and the very wide dynamic and spectral range controlled by velocity. But it also delivers dramatic and fat filter sounds that are hard to get from the old FM synths (not to mention the crazy feedback stuff).

The pedals allow much deeper sound modulation and the effect chain adds a lot of processing. Spark’s character is not always impressive, fat, brilliant and shiny. It can often sound cheap, ugly, and mean. I made no big efforts to minimize aliasing or to optimize the filter behaving in order to achieve the typically-favored analog "sound quality".

Sometimes the behavior is unpredictable and hard to tame. The feedback structure can cause surprising self oscillations and levels, where the limiter has to jump in. But that’s part of the
concept. It can feel a bit like beast in a cage…

How to play it

The velocity sensitivity of the two envelopes plays a great role in the real-time variation of the
sound. It can be used to create a very wide dynamic range of loudness and timbre.

The full expression potential becomes available by moving the three Macro Controllers. The set
of controls is designed to be a great environment for improvising musicians.

The Macro Controllers would be typically assigned to a Volume pedal, an Expression pedal and
the Mod Wheel. But they can also be easily controlled by sequencer automation curves.
Spark needs to be played in an expressive way. In the sequencer environment it takes the role
of the wave generator, that needs intense work with the "movements".

Recently I have added an LFO as an internal source of periodic movements. It can replace
automation or pedal movements (maybe only temporarily) and helps you to get the hands-free
for sound modifications.

Full Block Diagram

(click through for larger version)

Feedback Diagram

(click through for larger version)

slicedbread, on behalf of The Covert Operators, has released a free Windows utility that generates “endless harmonic oscillators” for Ableton Live’s Simpler and Sampler instruments. (Since this was a released, a Mac build has been made available, as well; see link below.) Even if you don’t intend to use the utility directly, pay attention – The Covert Ops already have a sample pack up full of oscillators, and you can bet the presence of this utility means more will come. (Even Robert Henke was impressed on the forums.)

Live 6 introduced the file format for “Ableton Meta Sounds.” Bjorn Vayner is currently breaking down how the format works, but the short upshot is that you can make oscillator sources that won’t alias for sound design in Simpler and Sampler. The AMS File Utility does more, too – export tunings (even microtuned stuff), and make oscillator variations. It’s sampling for people who like synthesis. In fact, not only is it fun to make additive synthesis-style oscillators dragging individual harmonics, but it’s a total breeze to change the offset and make equal-tempered stuff, negative scales, and other tunings.

Description on the forums:

AMS File Utility for Microtonal/Traditional Tunings

And from the very awesome Covert Operators site, some of the behind-the-scenes action, plus the Mac build (updated with additional links!):

Meta Files: Uncovering the .ams format, Part 1

Meta Files: Uncovering the .ams format, Part 2

Meta Files: Uncovering the .ams format, Part 3

Mac OS X Meta Application

Thanks for reminding us of this, Tony. I’m a bit behind on all this, but better late than never. Since I am lagging, has anyone made some AMS packs since this came out in September?

Earlier this fall, we got an inside look at the various Vista optimizations in Cakewalk’s SONAR 8. It was all fairly technical stuff, but odds are you may have been puzzled by a mention of support for custom “MMCSS task profiles.” Don’t worry – the correct response here is, indeed, “say WHA?” These refer to specialized task profiles, or configurations by which Vista determines how to prioritize access to the CPU. It’s not strictly necessary for end users to ever have to touch these. You can theoretically squeeze some additional performance and reliability out of your system by using them with SONAR 8, but that would require knowing how to do so.

Well, you’re in luck. I convinced Noel Borthwick, CTO at Cakewalk, to share his personal MMCSS profile with CDM. This will only work in SONAR 8 for the moment, as far as I know, until another host implements the same feature. Now, don’t get me wrong — it’s not necessary to use this tweak with SONAR 8 and Vista; I’ve been testing SONAR without it with no issue. This is just “because you can.” Here’s how to use it:

From the manual:

The Multimedia Class Scheduler Service (MMCSS) provides Windows multimedia programs, including SONAR, with prioritized access to CPU resources for time-sensitive processing.

By default, SONAR uses the MMCSS task profile named Pro Audio. If desired, you can instruct SONAR to use a custom MMCSS task profile.

To use a custom MMCSS task profile in SONAR

1. Determine the name of the MMCSS task profile that you want to use. MMCSS task profiles are listed in the Windows registry under HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Multimedia\SystemProfile\Tasks. Ensure that the registry entry exists for the profile you want to use.

2. If you have added any new profiles, reboot your computer to allow the MMCSS service to see the new profiles.

3. On SONAR’s Options menu, select Audio and then click the Advanced tab.

4. Under Configuration Settings, click Edit Config File. The AUD.ini file opens in the default Windows text editor.

5. In AUD.ini, locate the entry called MMCSSTaskKey and set it to the the desired profile name. For example:

MMCSSTaskKey=Pro Audio

6. Save AUD.ini and close the Windows text editor.

7. Click Reload Config Settings to reload the current audio configuration settings from AUD.ini.

SONAR will now use the new task profile for MMCSS.

And here’s the actual file:

Cakewalk Audio.regbak

As usual, use at your own risk. It’s unwise to make Registry modifications without first backing up your Registry, just in case something goes wrong.

Side note: if you other technical types at other software developers of all the attention we’re giving Noel at Cakewalk, it’s because he plies me with large amounts of alcohol when I’m in Boston. Kidding. Seriously, it’s because Noel has put a lot of time into sharing this stuff. We’re up for bringing any discussion to CDM, so get in touch. If I owned a car and didn’t live in Manhattan, I’d want to know how to work on the engine. I think it’s just as important to know what’s going on under the hood of software. Talk to your PR reps if you have to.

Previously on CDM:

Optimizing for Vista: Inside the Mechanics of SONAR 8 with Cakewalk Engineering

SONAR 8 Preview: Instrument Tracks, Beatscape Instrument, Transient Shaper, Enhanced Performance, Other Goodies

On a purely technical level, how does Windows Vista work with a leading pro audio application? A lot of that depends on just how much the developer does to tune their software, and to hook into some of the new audio technologies in the OS.

I wanted to know more about how Cakewalk’s SONAR 8 DAW was specifically optimized for Windows Vista. SONAR 7 was already out in front with support for Vista’s audio plumbing, but SONAR 8 promises quite a bit more. With the help of CTO Noel Borthwick, we were able to go through internal engineering documents and communication and get the full story. (Noel, who has a resume as a Linux programmer as well as leading the top Windows audio developer’s technical efforts, has been very frank in the past about Vista, walking us through changes before launch and real world experiences one year later.)

As Noel puts it, this is “more minute detail than most sane people would care to read,” but that’s just the kind of detail we like. If you just want to know whether your software will work, this is overkill. But for those of you who, like us, are interested in why some things work, this is a good place to start. I also sincerely hope other developers on Windows will start to invest more time in some of these details.

If SONAR 8’s music features are what interest you, see our first look:

SONAR 8 Preview: Instrument Tracks, Beatscape Instrument, Transient Shaper, Enhanced Performance, Other Goodies

I’ll repeat my current recommendation on Windows. If you’ve got a system running XP and you’re happy, I’d stick with it. But I see no reason to avoid SP1, unless you have a specifically incompatible setup, now that SP1 and driver updates have brought some maturity to the platform. I find the new OS more usable and, with the right software and drivers, even more stable. (This is not what I would have said before SP1, especially early after launch.) In fact, this walk-through with Noel demonstrates why, in the case of a SONAR system, you might be happier under Vista, on a point by point basis. I’ll avoid the “versus Mac” or “versus Linux” arguments, as those tend to get more complex and abstract, but there’s some good stuff here for those of you with Windows experience.

And yes, if any developer would like to do a similar tour on the Mac or Linux, I’m all ears. But just in case you think I’m crazy when I say there is some really robust stuff for audio in Vista, now you’ll know just what I mean.

CDM Executive Summary

Okay, here’s the quick run-down.

SONAR 8 involves improved performance optimizations, even relative to previous versions of SONAR. As far as I know, this is unparalleled progress on the Windows platform; few, if any developers are doing this kind of work. That’s not to say you won’t get good performance out of other Windows apps under Vista – certainly, I spend a lot of time in hosts like Ableton Live, Kore, and FL Studio on Vista — just that SONAR is leading here. (Developers out there doing similar work, we’d love to hear from you; developers not doing this kind of work, I hope this can piqué your interest and/or point you in the right direction, and certainly I don’t think anyone on Windows is trying to protect “trade secrets” – the better this stuff works, the better for everyone.)

The upshot of all of this should mean:

• Reduced audio glitches and performance bottlenecks, thanks to improvements to the audio engine itself, as well as the way the software makes use of the kernel (which impacts drivers also using the kernel)
• Smoother peformance in the UI as you zoom, draw, and the like
• More efficient performance, through optimization at the driver level and by better supporting Vista’s new audio API and threading mechanisms
• Better support for both WDM (Windows Driver Model) and ASIO drivers, and better performance at lower latency (that’s always the measure – you can always reduce latency, the question is whether you can maintain reliable audio performance when you do so)
• x64 is now, effectively, ready for prime time latency-wise

Now, here’s Noel to say that in more technically precise terms.

Performance Optimizations

Although every version of SONAR we shipped in the past had some degree of optimization work, SONAR 8 is the first version of SONAR to which we applied the same engineering process to performance optimizations as we do with other more user visible features.

i.e. we established goals, built a specification for the optimizations, split up the work into milestones and tracked the progress of these tasks just as we do for other features. To make testing more deterministic, we devised various internal profiling tools in order to track and measure changes in performance across a variety of hardware platforms on XP as well as Vista.

Systems tested included brand new cutting edge platforms from Intel and AMD as well as earlier generation machines.

We split up this work into the following classes of performance enhancements for SONAR 8:

1. CPU and kernel level optimizations – use less of your CPU to do the same amount of work
2. User Interface optimizations – faster drawing, scrolling, zooming
3. Driver level optimizations – more efficient access to drivers, minimizing driver state transitions
4. Vista OS specific optimizations – Better use of MMCSS thread priorities, support for custom MMCSS task profiles, new WASAPI support
5. Audio engine optimizations – optimize "hotspots" in our bussing, streaming and mixing code

As a result of all these changes, SONAR 8 has the following benefits:

• greatly minimized kernel usage. This helps provide more "kernel bandwidth" to drivers who need it the most. More kernel bandwidth translates into less potential for audio glitches.
• Lower CPU usage – translates to better performance at low latency
• More efficient use of audio drivers – esp with ASIO drivers
• Better performance on Windows Vista esp X64 [64-bit Windows]. Many of the complaints of Vista performance as compared to XP have been solved with SONAR 8. X64 low latency performance should now be on par with X86.
• Faster application launch
• Less flicker in GUI. Track view splitters no longer flicker when resizing.
• More responsive zoom and scroll with large projects. Zooming with wave files now uses 1/2 the RAM with 24-bit or less stereo or mono files used.
• Better meter performance.
• Improved thread scheduling by insuring threads are properly distributed on processors.

Benchmark: Overall benefits of SONAR 8 as compared to SONAR 7

CDM: Will we see these improvements in Project5? What about your instruments, like Rapture and Dimension, running as apps (instead of inside another host)?

Noel: We don’t have any firm plans for P5 at this time. Our “minihosts” for instruments don’t have these enhancements. It’s less relevant since these are far simpler with just a single instrument.

Windows Vista – New WASAPI driver mode

WASAPI (Windows Audio Session API) is the new multimedia API to talk to audio devices in Windows Vista. It represents the first real general purpose audio API from Microsoft in over 15 years, since MME from Windows 3.1.

The primary advantages of WASAPI are:

• Supports a wide variety of audio devices (any device that has a WDM driver should work in WASAPI mode if Windows can use it in exclusive/event mode)
• Provides low latency access to the device in WASAPI exclusive mode. (unlike older Microsoft API’s like MME or DirectSound)
• Model that is the closest to the low level WDM kernel streaming approach, while yet providing a higher level of abstraction for the device.
• Primary audio API for next generation of Windows.

Windows Vista: WaveRT

In WDM mode WaveRT is automatically used internally when SONAR detects a WaveRT capable WDM driver. A [WaveRT] suffix is added to the driver name in this case.

In SONAR 6.2 we first introduced support for Vista-specific WaveRT. Direct WaveRT mode is available in SONAR when you are using WDM as a driver mode and a WaveRT capable device is detected. In Vista SP1, Microsoft made some fundamental changes to WaveRT API’s that return the device sample position,
returning a cyclic position rather than a monotonically increasing position. In SONAR 8 WaveRT support has been updated and enhanced so that it works properly with Vista SP1.

CDM: You talk about using a WaveRT driver. Do you have to do anything to switch to WaveRT?

Noel: You don’t switch to WaveRT. In WDM [Windows Driver Model] mode, there is no choice – if a driver exposes itself as WaveRT, that’s the only mode of communicating with it. The same applies in WASAPI except the difference is that WASAPI itself communicates with the driver in WaveRT mode internally.

CDM: I have to admit, I’m used to working with ASIO. How do I know if a device has a WDM driver?

Noel: Any audio interface that works in Windows itself (i.e. if it shows up as a device in control panel -> sounds and audio devices) HAS to have a WDM driver. Otherwise Windows will not recognise it.

I don’t know of any interface that provides ONLY an ASIO driver. Some provide minimal WDM drivers that aren’t tested very well.

Windows Vista: MMCSS task profile support

The Multimedia Class Scheduler Service (MMCSS) provides Windows multimedia programs, including SONAR, with prioritized access to CPU resources for time-sensitive processing.

By default, SONAR uses the MMCSS task profile named Pro Audio. If desired, you can now instruct SONAR to use a custom user created MMCSS task profile.

CDM: Under what circumstances would you switch task profiles? What profile other than Pro Audio would you want to use?

Noel: You can tweak the MMCSS thread settings via a custom profile. I have a profile I tweaked that works marginally better than Pro Audio. It’s pretty technical stuff – You will have to read this link to get the details.

http://msdn.microsoft.com/en-us/library/ms684247(VS.85).aspx

Ed.: That’s a very technical explanation of a very technical topic; for more resources –

Mark Russinovich tours the enhanced Vista kernel

MMCSS, the Cliff Notes-style version on Wikipedia

Mark Russinovich on the interaction of MMCSS and networking

Interesting thread on the Cakewalk forum

More Information

You’ll want to see, previously…

Overview of Vista

Adieu, XP; How Vista SP1 is Doing, and Why This OS Generation Has Been So Tough

Vista for Audio, 1 Year Later: Talking OS Plumbing with Cakewalk’s CTO

Vista for Music + Pro Audio: Exclusive Under the Hood with Cakewalk’s CTO

Windows Sound Glitches Explained, Plus Glitches and the Fight-or-Flight Response

Compatibility

Cakewalk Vista Musicians’ Resource Page, Lots of Vista Drivers

Digidesign Talks Latest Windows, Mac Releases, Compatibility, Drivers

M-Audio Responds: Update on Driver Situation, New Drivers

Helpful utilities

12 Free and Cheap Must-Have Music Utilities for Windows

10 Free Non-Musical Windows Software Every Musician Should Use