ASIO4ALL v2 Universal ASIO Driver For WDM Audio
- Instruction Manual Last reviewed: 01/20/08

Table of Contents

Introduction.....1 Getting Started..... 1 Setting up your audio software.....1 Basic Configuration.....2 Advanced Configuration.... 4 Common Usage Cases Optimizations....6 Using ReWuschel.... 7 Troubleshooting..... 8

Introduction

Welcome to ASIO4ALL v2! This manual enables you to make the most of your ASIO4ALL installation, especially as it comes to the advanced features newly introduced in this Version of ASIO4ALL. The ASIO4ALL v2 installer should have added two new icons to your desktop, one being a link to this documentation and another to launch the ASIO4ALL control panel in off-line mode.
Getting Started Setting up your audio software
In order to make use of ASIO4ALL, you need to configure your audio software accordingly. How you would do this depends on your particular application. Generally, you would always enter the audio configuration menu and select ASIO -> ASIO4ALL v2. There now should be a button to launch the ASIO control panel. How this button is labeled depends on your particular software. Once you press this button, the ASIO4ALL control panel should appear. Please consult the manual of your audio software for further assistance, if necessary! Sometimes, the audio software indeed would not provide an ASIO control panel option. This is the case with e.g. Winamp and Foobar 2000, for which ASIO output plugins exist. For configuring ASIO4ALL with these applications anyway, you can launch the off-line control panel from the desktop. Note that the off-line control panel launcher works like a mini-ASIO-host of its own and does not reflect the current device status of any instance of ASIO4ALL that may be open at the same time! Changes made in the off-line control panel will have an effect only after you restart the audio application. Settings made in the off-line control panel, however, will have no effect at all for applications that have been configured with the ASIO4ALL (on-line) control panel launched from inside the audio options dialog of the application. Once you made it into the ASIO4ALL control panel, you can now proceed with some basic configuration.

Basic Configuration

Illustration 1: Basic Control Panel View
1. Device List This is the list of WDM audio devices found in your system. Highlight the device that you want to make changes to. Note: All parameter changes always only apply to the currently highlighted device! If a device name appears in bold, this device is enabled. In the picture above, the CMI8738/C3DX PCI Audio Device would be enabled while all others are not. The current state of each device is shown by a small icon and can be either of these: Running - The device has been successfully started by the audio engine. Available - The device should be available for use in this session but has not been started yet. Unavailable - The device is in use by another process or otherwise unavailable to the current session. Beyond Logic - You get this if the device for some unknown reason refuses to start and/or displays erratic behavior of any kind. Sometimes, closing and re-opening the control panel may cure the situation, as may unplugging and re-inserting of USB devices. Sometimes, this can also mean the same as Unavailable, whenever the device does not report its current availability or the lack thereof back to ASIO4ALL. Note: If the device list is empty, this means that you do not have a single WDM audio device in your system. Please check with your audio hardware manufacturer for a WDM driver! 2. ASIO Buffer Size Use the slider to adjust the ASIO buffer size for the device currently highlighted. Smaller buffer size means lower latency. Once you hear crackles or audio becomes distorted, you need to increase the buffer size. ASIO buffer size directly relates to audio latency. Thus, you want to get a rather small value here.

3. Multi-Function Button Allows you to enable/disable/select the device list item currently highlighted. Use this to select the currently highlighted device for ASIO operation. In basic mode, you can only select one device at a time. If there is no valid action that can be performed with the current item, this button acts as an additional Exit- button. You can always close the control panel using the Close button in the title bar, which has the same effect. 4. Switch To Advanced Mode Switches the control panel into advanced mode, where you can fix things or completely mess them up at your disposal. Advanced mode is explained in the Advanced Configuration section of this document. 5. Load Default Settings Pressing this button will reset all configuration options to their initial defaults. Use when audio initially worked and you later got lost in the configuration process. 6. Status Bar The most important bit of the status bar is the overload indicator (right where the 6 mark is in the picture). If the overload indicator flashes, it means just that. You may either want to increase the ASIO buffer size, or, if it just flashes briefly when there is heavy system load (e.g. When loading a project), you may just want to ignore it and merely appreciate the coolness of this new feature.

Advanced Configuration

Illustration 2: Advanced Control Panel View
Once the control panel has been switched to advanced mode, things begin to look a little more complicated. You will have noticed that the items in the device list are now expandable. By expanding the device list, you can now fully explore the WDM audio architecture of your system. The device list contains Devices, Device Interfaces and so called Pins (this is Microsoft(r) lingo). The figure above illustrates how they interrelate in the WDM KS hierarchy. Using the Action-button, you can now selectively enable/disable each particular item in the device list. This way, you can also create multi-device-setups. Multi-device-setups require that all the devices involved are running from the same clock source. You can achieve this by daisy-chaining devices via S/PDIF etc. Fortunately, most USB devices will automatically syncronize themselves for as long as the host controllers they are connected to have a common clock source, which is trivially true for the USB host controllers embedded in the south bridge on any mainboard. Note: If devices are not accurately synced, their audio streams are likely to drift apart over time! On the right side of the panel are the advanced controls (1.5). 1. Latency Compensation Since ASIO4ALL does not have sufficient knowledge of the underlying hardware/driver architecture, it can only guess the actual latencies involved. With these sliders you can compensate for the latencies unknown to ASIO4ALL such that recordings in your sequencer Software are properly aligned with the rest. Note: In multi-device-setups the largest respective value of all devices will be used. Therefore, if different devices have different inherent latencies, audio placement will not be accurate for some devices!

2. Hardware Buffer on/off Enables the hardware buffer for the highlighted device. This only works for so called WavePCI miniports, as other types of WDM drivers do not usually allow direct access to the hardware buffer. Note: For WaveRT drivers (Vista), this box is always checked! Adjustment for best hardware buffer performance involves the ASIO Buffer Size slider and the Buffer Offset slider (see below). Hardware buffering works best for rather small ASIO buffer sizes. Try something between 128 and 256 samples as a starter! The biggest advantage of using the hardware buffer is that this method uses a lot less CPU. In addition, it may be possible to decrease latencies even further. In multi-device-setups, it is possible to mix Hardware-buffered devices with devices that are not. This, however, is not particularly recommended! If hardware buffering is not supported by a particular audio device, there will be an additional latency of a couple hundred milliseconds, which is clearly audible. 3. Kernel Buffers/Buffer Offset If hardware buffering is disabled, this control lets you add up to two more buffers to be queued for audio output. Each additional buffer increases the output latency of the device by the time it takes to play one buffer. Therefore, the initial setting of 2 should only be changed on less powerful machines, where reasonably small ASIO buffer sizes cannot be achieved with the default setting. If hardware buffering is enabled, this control determines the amount of clearance (in ms) between where ASIO4ALL will insert data into/read data from the hardware buffer, and the position where ASIO4ALL currently thinks the hardware read/write position is. Sound complicated already? You haven't even seen the code that calculates this. As a general rule: Higher settings increase latencies and stability, lower settings have the adverse effect. You should, however, be able to achieve a setting that is very close to zero (4ms would still be considered very close to zero, while 10ms, the default, would indicate that there is room for improvement.) With Envy24-based PCI-sound cards, there may be an option in your sound card control panel that reads DMA Buffer Transfer Latency (Seen with Terratec products) or similar. You should set this to the lowest possible value, e.g. 1ms for best results. 4. Always Resample 44.1<->48 kHz ASIO4ALL can do real time resampling of 44.1 kHz audio to/from 48 kHz. Resampling will automatically take place whenever ASIO4ALL is opened for 44.1 kHz and the WDM driver does not support this sample rate. There may, however, be instances in which case an AC97 will support 44.1 kHz by resampling internally. More often than not, however, AC97 resampling quality is extremely poor and/or prone to stability issues. To work around this, you can enable this option. With at least one incarnation of the SoundMax WDM driver (smwdm.sys), this option absolutely must be enabled in order to make it work at 44.1 kHz at all. 5. Force WDM driver to 16 Bit This option only has an effect if the supported bit depth of the WDM driver is larger than 16, but less than 24. Some AC97 devices report e.g. 20 Bits resolution but cannot actually be opened for more than 16 Bits resolution. Should this be the case on your system, this option provides a workaround. Originally, this was introduced as a workaround for an issue with the SigmaTel AC97 WDM driver.

Common Usage Cases Optimizations
Playing Software Synths Live In this scenario, you do not need audio inputs. Therefore, you best disable them all, which normally will provide you with a better stability at very small ASIO buffer sizes, or allow smaller buffer sizes in the first place. Further, you should also disable all audio outputs you do not really need. To disable channels, use the advanced control panel, expand the items in the WDM device list and disable everything you do not want to use in this setup!
Computer As Effects Processor Obviously, in this scenario you do need inputs. But, as always, you should disable all channels you do not want to use. Disable 44.1KHz resampling if it is not really necessary!
General Purpose Sequencer Setup Normally it matters most that you do not get any dropouts even when the CPU load goes through the roof. Thus, it is recommended that you relax the latency requirements a little and work with an ASIO buffer size that feels comfortable with all your favorite VST plugins active. This especially applies when you are recording audio, in which case dropouts are a little worse than just moderately annoying. If your sequencer provides latency compensation, you probably want to check that recorded audio is aligned properly, and, if not, make the necessary adjustments in the Latency Compensation section in the advanced settings dialog.
Latency Does Not Matter A Lot In certain configurations, ASIO4ALL allows for bit transparent audio where the Windows driver stack does not. Hence, audiophiles prefer ASIO output over DirectSound or MME, which most likely does mangle audio data. In these scenarios, latency is of little concern and audio input is not asked for. Naturally, you would make sure that all inputs are disabled, set the ASIO buffer size to the maximum and be happy!

Using ReWuschel

If you selected the ReWuschel install option, the audio inputs ASIO4ALL provides to your host application are now also accessible as ReWire inputs. This makes sense for applications that only allow for ASIO output, but support ReWire. One such application is Reason. Now, in order to get real-time audio input in Reason, you need to make sure that ASIO4ALL v2 is the current ASIO driver. Then: Create->ReBirth Input Machine - that's it! The ASIO4ALL inputs are now mapped to the outputs of the ReBirth Input Machine. Note that, once you have ReWuschel installed, you cannot use ReBirth in Reason. In order to re-enable Reason. ReBirth input for Reason you need to re-install ASIO4ALL with the ReWuschel install option left unchecked!

Troubleshooting

Since ASIO4ALL presents itself to the audio software as a single ASIO driver, but due to its various configuration options, can act like a chameleon, there are numerous things that can go wrong without ASIO4ALL being at fault. Most notably, if you change the device setup in the ASIO4ALL control panel, the number of available input and output channels is likely to change, as well as the names of the channels that are seen by the host application. Therefore, it is always advisable to restart your audio host application after any change in the audio device setup whenever you find that the particular application does not appear to be able to handle these kinds of changes on the fly. More potential problems and possible solutions:
ASIO4ALL v2 not visible in host audio configuration menu There are two possible reasons for this: Either your audio application does not support ASIO or you installed ASIO4ALL v2 as an underprivileged user. In the latter case, please log on as Administrator and install ASIO4ALL v2 again. Once successfully installed, ASIO4ALL v2 should not require Administrator privileges anymore in order to run.
Audio device flagged as Unavailable or Beyond Logic even though it is not in use elsewhere You want to make sure the MS GS Software Wavetable Synth or anything by a similar name is not enabled anywhere in you MIDI setup. The ASIO4ALL Web Site has further information on that. If any such Software Wavetable Synth (sometimes in disguise as.DLS Synth.) can be ruled out as the cause, try to restart the audio host application. Sometimes, when switching from another driver to ASIO4ALL v2, the previous driver will not release the audio device in time. If the device is an USB/PCMCIA/FireWire device, close the ASIO4ALL control panel, unplug the device, plug it in again and re-open the ASIO4ALL control panel.
Cannot play sound from another application when ASIO4ALL is active This by design. As close to the hardware as possible means that all the software mixing provided by Windows will be bypassed. Without any software mixing - and the associated latency and bit mangling you are stuck with the hardware mixing capabilities of your audio device. Most of the time, there simply are no hardware mixing capabilities at all.
(Vista) The meters are moving, everything looks o.k. - But there is no sound Drivers for High Definition Audio devices have a habit of listing SPDIF outputs before the analog ones. As a consequence, the first two or more ASIO channels will be routed to SPDIF, if you use the default configuration. You can change this by either disabling the SPDIF output in the ASIO4ALL control panel or correctly assigning ASIO channels in your audio application.

Changes made in the control panel do not propagate between different audio applications.Neither do they propagate between different users! This is not a bug, it's a feature! ASIO4ALL v2 stores settings per host application/per user! This makes it possible to have several instances of ASIO4ALL run at the same time for as long as they do not try to use the same piece of audio hardware exclusively. This further allows having ASIO4ALL run in educational/computer lab type environments without user666 being able to f**** things up for any user on the same machine, with the exception of user666 him/her/itself.
The latencies displayed in e.g. Cubase SX 3 do not match the values that would result from the ASIO buffer size. ASIO4ALL supports the latency compensation features of ASIO hosts that perform latency compensation. This support is still a bit under development and will be improved as time passes. The values reported here are not just the latencies ASIO4LL adds to the audio stream, but rather the represent the whole shebang as far as driver/OS/hardware inherent latencies. If the guess was correct, that is. Earlier Versions of ASIO4ALL did not make this attempt at guessing, so you may obtain a smaller latency display with v1.x and other WDM->ASIO wrappers. The true (i.e. Perceived) latencies are at least as good as with earlier versions and on top of that, they do not change anymore when CPU utilization goes up!
Copyright 2004-2008, Michael Tippach. All trademarks are the property of their respective owners and used for product identification purposes only. This document contains statements that may/may not be true. Particular combinations of colors or black and white contrast may/may not cause epilepsy, nausea, or the urge to do weird things to domestic animals. Either way, I shall not be held responsible for any of that.

Titelsida Development of a realtime signal framework (Utveckling av ramverk fr signaler i realtid) Mattias Claesson
Abstract The department of Telecommunications and Signal Processing are currently conducting research projects where a realtime framework has the potential to much shorten the time from the design of the algorithm is finished until it can be proven to work in real world. In addition to research projects the framework will be used when developing courses and improving already existing courses. It should be possible to use the framework both for laborations and lectures.
Copyright ASIO is a trademark of Steinberg Soft- und Hardware GmbH ActiveX, BackOffice, Developer Studio, Direct3D, DirectDraw, DirectInput, DirectMusic, DirectPlay, DirectShow, DirectSound, DirectX, DirectAnimation, JScript, Microsoft, Microsoft Press, MSDN, MS-DOS, MSN, Natural, OpenType, Visual Basic, Visual C++, Visual Studio, Win32, Win32s, Win64, Windows, Windows NT, and XENIX are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries.

Contents

Chapter 1 Introduction... 1 Chapter 2 Physical Interface... 2 2.1 Overview.... 2 2.2 Sound card.... 2 2.3 Microphone.... 3 2.4 Microphone Amplifier... 3 2.5 Equipment list.... 3 Chapter 3 Windows audio architecture.. 4 3.1 Windows Kernel Streaming Architecture... 4 3.2 WDM Audio Drivers... 5 3.2.1 SysAudio and KMixer.. 6 3.3 Audio APIs... 7 3.3.1 DirectSound.... 7 3.3.2 ASIO... 7 3.3.3 Direct Kernel Streaming Technique.. 8 3.4 Universal Audio Architecture... 8 3.5 Multiple Channel Audio Data and WAVE Files... 8 Chapter 4 The Application... 9 Chapter 5 Evaluating the framework... 10 Chapter 6 Summary and Conclusions.. 11 Chapter 7 User Manual... 12

Chapter 1 Introduction

This thesis will result in a programming interface for realtime signalprocessing. The framework user will write an implementation of an algorithm in the industry standard MATLAB programming language. Instead of executing the MATLAB code directly, the MATLAB compiler is used to automatically transform the algorithm to the C/C++ programming language. The C/C++ code of the algorithm is then merged with code developed in this thesis, code that input and output the realtime data to and from the soundcard. Together all code is compiled to a single executable using the Visual C++ 6.0 compiler. The framework development in this thesis will mainly consist of: Interface description MATLAB Implementation of one realtime input/output interface The idea is that the interface description towards MATLAB will be general enough so that it is possible to develop (and change to) another input/output interface instead of the one provided in this thesis. It should be possible to switch to a new interface without changing the MATLAB code in an implementation of an algorithm. Chapter 2 Physical Interface The initial chapter present the physical interface in this thesis. A brief system overview is presented followed by detailed descriptions of every component. Chapter 3 Windows audio architecture Description of the most important parts of the architecture for sound support in a Windows operating system now and in the future. Chapter 4 The Application framework. Development of a application using the
Chapter 5 Evaluating the framework properties of the developed system. Chapter 6 - Summary and Conclusions results of this thesis. Chapter 7 - User Manual developer.
Presentation of methods to evaluate the Summarizes and concludes the work and
Contains required information for the application
Chapter 2 Physical Interface

2.1 Overview

The physical interface used in this thesis consists of a multichannel sound card connected to a microphone amplifier. The microphone amplifier is connected to good quality microphones. As the selected equipment makes it possible to input eight and output eight mono analog signals, it provides a good base for most applications. A general overview of the complete system is presented in Figure 1.

Sound card

Microphone Amplifier Microphone

Amplifier

Speaker

Figure 1 System overview

2.2 Sound card
The inputs and outputs of the sound card can individually be set to balanced or unbalanced mode using the +4/-10 switches on the back. Both modes use the same size connectors , but the number of used sections of the TRS connector (see Figure 2) to transmit the three components of a balanced signal (T =plus, R = minus, S = ground) differs as opposed to the two sections on a unbalanced style connector. A balanced signal originating from a microphone uses often a XLR connector (see Figure 3).

Sleeve

Figure 2 TRS connector

Figure 3 XLR connector

The balanced mode (+4 dBu) is used for professional equipment and the unbalanced mode (10dBV) for consumer equipment. These levels are the nominal signal levels. A 0 valued +4 dBu signal has a voltage level of 1.23 volts and a 0 valued 10 dBV signal has a voltage level of 0.316 volts. For optimum performance it is recommended to use balanced cables. The analog inputs and outputs provided are capable of 24-bit data width and sampling rate from 8kHz to 96kHz.

2.3 Microphone

The used microphone is of condenser type and will require external power to function properly. The microphone has a XLR connector.

2.4 Microphone Amplifier

The chosen microphone amplifier can provide phantom power to the microphone thru the standard input XLR connector. In general most professional microphone amplifiers does provide the option to power the microphone by means of phantom power. All channels have independent gain and clip indicator. The outputs have balanced TRS type connectors.

2.5 Equipment list

Description Sound card Microphone Microphone Amplifier Type number M-Audio Delta 1010 AKG C417III PP SM Pro Audio PR8 Comment Analog mono 8 in/8 out Condenser, XLR connector 8 channel, phantom power
Chapter 3 Windows audio architecture
All recent Windows operating systems use the same basic foundation as their audio architecture. The generic device driver model in these operating systems is called WDM (Windows Driver Model) and is used for all kinds of device drivers like video, sound and storage. WDM is based on the layered device-driver model in Windows NT, and it provides additional support for Plug and Play. All streaming devices share the same functionality provided by the Windows Kernel Streaming Architecture (KS) and is based on WDM. The personal computer is more and more used for applications that require streaming of data. These applications differ from those used for traditional business tasks in that they deliver large amounts of data to hardware, and that data is generally timesensitive. Before WDM and KS implementation of streaming data was hindered by an inconsistent driver model, an architecture that wasnt modular or extensible, inefficient use of system resources, and delivery delays.

3.1 Windows Kernel Streaming Architecture
Kernel Streaming is kernel-mode processing of streamed data. KS does enable efficient real-time streaming for streaming devices. The Microsoft kernel streaming model is used in Microsoft Windows 2000 and later operating systems, and in Windows Me. KS drivers can be both filter and functional drivers. KS allows audio stream drivers to communicate directly with audio adapter hardware. The WDM Kernel Streaming architecture divides the processing of data into blocks of functionality. These blocks are called filters, and serve as the atomic components in the system. The properties of a filter is partly described by a number of connection points called pin factories. These pin factories can normally be used to create instances of a pin. Each pin instance can consume, produce, or both consume and produce a data stream, such as digital audio. Tasks can be performed by connecting these filters together into a filter graph. In the Windows Driver Development Kit (DDK) there is a program called KS Studio. The program can be used to create filters and connect them into graphs. For a simple example of such a graph see Figure 4.
Figure 4 Kernel Streaming graph
See Reference [1] for a more detailed description of Kernel Streaming.

3.2 WDM Audio Drivers

Microsoft Windows 98, Windows Millennium Edition (Windows Me), and Windows 2000 and later include enhancements to simplify developing and managing WDM streaming audio drivers, which are based on the WDM kernel streaming architecture. The KS filter is a convenient abstraction for representing an audio device. Input pins on an audio filter consume audio streams and route them to the audio hardware. Output pins on the filter produce audio streams from data produced by the hardware. Avoid confusing the term KS filter with the term filter driver, which is another WDM concept. To make audio hardware drivers easier to write, the WDM audio driver model isolates the hardware interface issues from the filter-implementation issues. The model organize the driver code into components that address these issues separately and by precisely defining the interfaces between these components. See Figure 5 for a overview of the audio achitecture in Windows XP.
Figure 5 Simplified view of WDM audio driver architecture in Windows XP [Redraw, add ASIO, add Direct KS]

SysAudio and KMixer

KMixer allows all streams or some subset of the active streams in the system to be played at the same time. KMixer makes intelligent tradeoffs between audio quality and CPU usage. It also perform synchronization and (sample-rate conversion) SRC if needed. To perform SRC KMixer uses a high-quality, high-performance, multi-tap, Finite Impulse Response (FIR)-based SRC algorithm.
The SysAudio and KMixer components enable multiple audio applications to play sound simultaneously. The system audio device (SysAudio.SYS) constructs a filter graph to perform the format conversion, sample-rate conversion (SRC), and mixing needed to combine digital streams. The kernel mixer (KMixer.SYS) performs the actual format conversion, SRC, and mixing operations on the streams. For an introduction of WDM Audio drivers see Reference [2]. However the Windows DDK is the document for the driver developer or advanced application developer. The DDK documentation is available online at Reference [3] and it contain all the detailed information about WDM, Kernel Streaming and WDM Audio drivers.

3.3 Audio APIs

The audio application programming interfaces (APIs) are built on top of the KS architecture. These APIs include Microsoft DirectSound, WinMM (for Win32 compatibility), and MMSystem (for Win16 compatibility). A more specialized API for streaming audio is the Audio Streaming Input Output (ASIO) developed by Steinberg. A technique called Direct Kernel Streaming, where a user mode application connect directly to a kernel mode KS filter (the sound card device driver).

DirectSound

The DirectSound API is the audio component of the DirectX 9.0 SDK. DirectSound provides low-latency mixing, hardware acceleration, and direct access to the sound device. DirectSound use the kernel mode components KMixer.SYS and SysAudio.SYS, thus the sound input and output of DirectSound is subject to the time delay introduced by these components. On the other hand sound captured or played thru the DirectSound API take advantage of the SRC and mixing capabilities of the above mentioned kernel components. DirectSound allows application programs to use the hardware in the most efficient way possible while presenting a device-independent interface that insulates applications from the specific details of that hardware. DirectSound applications work well with the simplest audio hardware, but they also take advantage of the special features of cards and drivers that have been enhanced for use with DirectSound. See Reference [4] for the complete DirectX 9.0 Software Development Kit (SDK).
The ASIO API provide multiple channel input and output streaming. The application (called host application) load the API as Component Object Model (COM) object. The sound card manufacturer develop the actual ASIO API driver (the COM object mentioned above). The ASIO API driver usally interact with the sound card hardware thru the WDM driver of the sound card. Both the host application and the ASIO API driver part execute in user mode. The ASIO API between the host
application and the ASIO API driver is based on a rather standard dubble buffer architecture. For the description of ASIO see Reference [5] from Steinberg. As a side note, several implementations of mappings between the ASIO API and the standard WDM driver has been developed. The one in Reference [6] was the first but developement seem to have stopped. For the latest and also the one that work best of the two see Reference [7].
Direct Kernel Streaming Technique
TBD See Reference [8] for an application example and its documentation. Please note that the example application requires either the Windows Platform SDK or the Windows DDK to be able to compile.

3.4 Universal Audio Architecture
http://www.microsoft.com/whdc/hwdev/tech/audio/UAA.mspx The Universal Audio Architecture (UAA) describes a class driver architecture for PC audio solutions supported in the next version of the Microsoft Windows operating system, code-named Longhorn. This paper provides an overview of UAA for system manufacturers and manufacturers of audio devices. A Wave Port Driver for Real-Time Audio Streaming http://www.microsoft.com/whdc/hwdev/tech/audio/WaveRTport.mspx This paper provides information about the WaveRT port driver for the Microsoft Windows family of operating systems. It provides guidelines for audio hardware vendors to develop WaveRT miniport drivers for their audio devices.
3.5 Multiple Channel Audio Data and WAVE Files
http://www.microsoft.com/whdc/hwdev/tech/audio/multichaud.mspx This article describes the standard for storing and transporting multiple channel audio data using the WAVE file format. The reader should have a basic understanding of multimedia file formats, and especially of audio file formats. This article describes the method used to author multi-channel audio streams that require well-defined channel/speaker locations. Any of these formats could be used to indicate the number of bits of precision in a high-resolution stream.
Chapter 4 The Application
Here will there be a description of a simple proof of concept application.
Chapter 5 Evaluating the framework
Here there will methods to evaluate the framework be presented.
Chapter 6 Summary and Conclusions
Slutredovisningen ska innehlla en diskussion av huruvida den valda metoden lst problemet som angavs i inledningen. Under arbetets gng har du skert ftt flera egna ider om frbttringar. Det r viktigt att du tar fram sdana synpunkter.

Chapter 7 User Manual

Om ditt examensarbete, t ex en apparat eller ett datorprogram, ska anvndas av andra behvs i allmnhet ngon form av anvndarhandledning. Med hjlp av denna ska man kunna anvnda din slutprodukt utan att behva lsa igenom hela rapporten. Tnkt och snt:

ASIO API BufferSwitch()

Engine (Convert to double)
MATLAB [1,2,3,.]=Data(1,2,3,.)

Log data to files

Bibliography
[1] Microsoft Corporation April 2, 2003
http://www.microsoft.com/whdc/hwdev/tech/stream/csa1.mspx
Windows Kernel Streaming Architecture
[2] Microsoft Corporation February 28, 2003
http://www.microsoft.com/whdc/hwdev/tech/audio/wdmaudio.mspx