Sound is an area of the PC that has been largely overlooked in early systems. Aside from a simple, oscillator-driven speaker, the early PCs were mute. Driven largely by the demand for better PC games, designers developed stand-alone sound boards that could read sound data recorded in separate files, then reconstruct those files into basic sound, music, and speech. Since the beginning of the decade, those early sound boards have blossomed into an array of powerful, high-fidelity sound products capable of duplicating voice, orchestral soundtracks, and real-life sounds with uncanny realism (Fig. 47-1). Not only have sound products helped the game industry to mature, but they have been instrumental in the development of multimedia technology (the integration of sound and picture), as well as Internet web phones and other communication tools. This chapter is intended to explain the essential ideas and operations of a contemporary sound board, and show you how to isolate a defective sound board when problems arise.

Before you attempt to troubleshoot a problem with a sound board, you should have an understanding of how the board works, and what it must accomplish. This type of background helps you when recommending a sound board to a customer, or choosing a compatible card as a replacement. If you already have a strong background in digital sound concepts and software, feel free to skip directly to the troubleshooting portion of this chapter.

All sound starts as pressure variations traveling through the air. Sound can come from almost anywhere - a barking dog, a laughing child, a fire engine’s siren, a person speaking - you get the idea. The process of recording sound to a hard drive requires sound to be carried through several manipulations (as shown in Fig. 47-2). First, sound must be translated from pressure variations in the air to analog electrical signals. This is accomplished by a microphone. These analog signals are amplified by the sound card, then digitized (converted to a series of representative digital words each taken at a fixed time interval). The resulting stream of data is processed and organized through the use of software, which places the data (as well as any overhead or housekeeping data) into a standard file format. The file is saved to the drive of choice - typically a hard drive.

Simply speaking, the playback process is virtually the reverse of recording (Fig. 47-3). A software application opens a sound file on the hard drive, then passes the digital data back to the sound card. Data is translated back into equivalent analog levels - ideally, the reconstructed shape of the analog signal closely mimics the original digitized signal. The analog signal is amplified, then passed to a speaker. If the sound was recorded in stereo, the data is divided into two channels which are separately converted back to analog signals, amplified, and sent to their corresponding speakers. Speakers convert the analog signal back into traveling pressure waves that you can hear.

To appreciate the intricacies of a sound card’s operation, you must understand the concept of digitization - otherwise known as sampling. In principle, sampling is a very straightforward concept; an analog signal is measured periodically, and its voltage at each point in time is converted to a digital number. The device which performs this conversion is known as an analog-to-digital converter (ADC). It sounds simple enough in principle, but there are some important wrinkles.

The problem with sampling is that a digitizer circuit has to capture enough points of an analog waveform to reproduce it faithfully. The example in Fig. 47-4 illustrates the importance of sampling rate. Waveforms A and B represent the same original signal. Waveform A is sampled at a relatively slow rate - only a few samples are taken. The problem comes when the signal is reconstructed with a digital-to-analog converter (DAC). As you see, there are not enough sample points to reconstruct the original signal. As a result, some of the information in the original signal is lost. This is a form of distortion known as aliasing. Waveform B is the same signal, but it is sampled at a much higher rate. When that data is reconstructed, the resulting signal is a much more faithful reproduction of the original.

As a rule, a signal should be sampled at least twice as fast as the highest frequency contained in the signal - this is known as Nyquist's Sampling Theorem. The lowest standard sampling rate used with today's sound boards is 11kHz - this allows fair reproduction of normal speech and vocalization (up to about 5.5kHz). However, most low-end sound boards can digitize signals up to 22kHz. Unfortunately, the human range of hearing is about 22kHz. To capture sounds reasonably well throughout the entire range of hearing, you would need a sampling rate of 44kHz - this is often known as CD-quality sampling since it is the same rate used to record audio on CDs. The disadvantage to high sampling rates is disk space (and sound file size). Each sample is a piece of data, so the more samples taken each second, the larger and faster a file grows.

Not only does the number of samples effect sound quality, but also the precision (or number of bits) of each sample. Suppose that each sample is converted to a 4-bit number. That means each point can be represented by a number from 0 to 15 - not much precision there. If 8 bits are used for each sample, 256 discrete levels can be supported. But the most popular configuration is 16-bit conversion which allows a sample to be represented by one of 65536 levels. At that level of resolution, samples will form a very close replica of the original signal. Many of today’s sound boards are 16-bit.

Although the majority of a sound card is geared toward handling the recording and playback of sound files, the musical instrument digital interface (MIDI) port has become an inexpensive and popular addition to many sound card designs. MIDI is a standard protocol that is defined by hardware, software, and electrical interconnections. At the core of a MIDI interface is a synthesizer IC. Unlike a sound file which basically contains the digital equivalent of an analog waveform, a MIDI file is a set of instructions for playing musical notes. Each note is sent to the synthesizer, along with duration, pitch, and timing specifications. The synthesizer can be made to replicate a variety of musical instruments such as a piano, guitar, harmonica, flute - you name it. The high-end sound boards are capable of synthesizing a small orchestra. Since most synthesizers can process several channels simultaneously, the MIDI standard supports playing a number of "instruments" (or voices) at the same time. Thus, very high quality music can be produced with MIDI on a PC. The two most common synthesizer types are FM and Wavetable.

Figure 47-5 illustrates the kinds of things MIDI is capable of. Pre-recorded MIDI files can be read from a storage device like a hard drive file, or from CD-ROM (many games include an orchestral-quality MIDI soundtrack on the CD). The MIDI data is passed through to the sound board’s synthesizer which reproduces the sound, and out to the amplified speakers. If you plan on composing music yourself, you can interface a MIDI instrument to the sound board’s MIDI port. Using MIDI sequencer software, the notes played on the instrument will be heard through the speaker, as well as recorded to the MIDI file on the hard drive. Note that you do not need a MIDI instrument to playback a MIDI file, but you need an instrument and sequencer software to create a MIDI file. Also, since MIDI is not sound (but rather sound "blueprints") the same MIDI composition entered on a keyboard can be played back as a harp, or a guitar, or a flute.

Now that you are aware of the major functions a sound board must perform, you can see those functions in the context of a complete board. Figure 47-6 shows a simplified block diagram of a sound board. It is important to note that your own particular sound board may differ somewhat, but all contemporary boards should contain these subsections.

The core element of a sound board is the digital signal processor (DSP). A DSP is a variation of a microprocessor which is specially designed to manipulate large volumes of digital data. Like all processor components, the DSP requires memory. A ROM contains all of the instructions needed to operate the DSP and direct the board’s major operations. A small quality of RAM serves two purposes; it provides a "scratch pad" area for the DSP’s calculations, and serves as a buffer for data traveling to or from the PC bus.

Signals entering the sound board are passed through an amplifier stage and provided to an A/D converter. When recording takes place, the DSP runs the A/D converter and accepts the resulting conversions for processing and storage. Signals delivered by a microphone are typically quite faint, so they are amplified significantly. Signals delivered to the "line" input are often much stronger (such as the output from a CD player or stereo preamp), so it receives less amplification.

For signals leaving the sound board, the first (and often most important) stop is the mixer. It is the mixer which combines CD-audio, DSP sound output, and synthesizer output into a single analog channel. Since most sound boards now operate in a stereo mode, there will usually be two mixer channels and amplifier stages. The audio amplifier stage(s) boost the analogs signal for delivery to stereo speakers. If the sound will be driving a stereo system, a "line" output provides a separate output. Amplifier output can be adjusted by a single master volume control located on the rear of the board.

Finally, a MIDI controller is provided to accommodate the interface of a MIDI instrument to the sound board. In many cases, the interface can be jumpered to switch the controller to serve as a joystick port. That way, the sound board can support a single joystick if MIDI instrument will not be used. MIDI information processed by the DSP will be output to the on-board synthesizer.

An important aspect of sound boards is their audio benchmarks. Unlike logic and processing circuitry which is measured in terms of millions of operations per second, the benchmarks that define a sound card are very much analog. If you are an audiophile, many of the following terms may already be familiar. If most of your experience has been with logic systems, however, these concepts will appear very different than many of the other discussions in this book.

No discussion of sound concepts is complete without an understanding of the decibel (or dB). Decibels are used because they are logarithmic - you see, human hearing is not a linear response. If you increase the power of your stereo output from 4W to 16W, the sound is NOT 4 times louder - in fact, it is only twice as loud. If you increase the power from 4W to 64W, the sound is only three times as loud. In human terms, amplitude perception is measured logarithmically. As a result, very small decibel values actually relate to substantial amounts of power. The accepted formula for decibels is:

Pout

gain (in dB) = 10 log10 ----------

Pin

Don't worry if this formula looks intimidating - chances are that you will not need to use it, but consider what happens when output power is greater than input power. Suppose a 1mW signal is applied to a circuit, and a 2mW signal leaves, the circuit provides a gain of +3dB. Suppose the situation was reversed where a 2mW signal were applied to the circuit, and a 1mW signal left it. The circuit would then have a gain of -3dB. Negative gain is a loss, also called attenuation. As you see, a small dB number represents a large change in signal levels.

Expressed simply, the frequency response of a sound board is the range of frequencies that the board will handle uniformly. Examine the sample graph of Fig. 47-7. ideally, a sound board should be able to produce the same amount of power (0dB) across the entire working frequency range (usually 20Hz to 20kHz). This would show up as a flat line across the graph. In actual practice, however, this is not practical, and there will invariably be a rolloff of signal strength at both ends of the operating range. A good-quality sound board will demonstrate sharp, steep rolloffs. As the rolloffs get longer and more shallow at high and low frequencies, the board has difficulty producing sound power at those frequencies. The result is that bass and treble ranges may sound weak, and this effects the sound’s overall fidelity. By looking at a frequency response curve, you can anticipate the frequency ranges where a sound board may sound weak.

The signal-to-noise ratio (SNR) of a sound board is basically the ratio of maximum undistorted signal power to the accompanying electronic noise being generated by the board (primarily hum and hiss) expressed in decibels. Ideally, this will be a very large dB number which would indicate that the output signal is so much stronger than the noise signal, that for all intents and purposes, the noise is imperceptible. In actual practice, a good-quality sound board will enjoy an SNR of 85dB or higher - but these are difficult to find. For most current sound boards with SNR levels below 75dB, there may be audible hum and hiss present during silent periods, as well as a certain amount of sound "grit" underlying sound and music reproduction. Some very inexpensive sound boards are on the market with SNR levels as low as 41dB (noise may be noticeable, and actually annoying).

You may also find the SNR value expressed as an "A-weighted" decibel number. The reason for this is that human hearing is not equal at all frequencies, so we cannot hear all noise equally. The process of "A-weighting" emphasizes the noise levels at frequencies we are most sensitive to. Resulting SNR values are often several dB higher (better) than non-weighted SNR values. Be careful here, a sound board with a low SNR may use the A-weighted value in the specification sheet. If this is the case, subtract about 3 or 4dB for the actual SNR figure.

Sound and music are rich in harmonics (overtones) which are basically integer multiples of an original frequency signal (although at much lower levels). As a consequence, harmonics are a valuable attribute of sound. The number and amplitude of harmonics provide the sound characteristics that allow you to distinguish between a guitar, flute, piano, or any other musical instrument played at the same note - without harmonics, every instrument would just produce flat tones, and every instrument would sound exactly the same.

However, when sound is produced in an electronic circuit, other unwanted harmonics are generated which can alter the sound of the music being produced (thus the term harmonic distortion). The total harmonic distortion (THD) of a sound board is the root-mean squared (RMS) sum of all unwanted harmonic frequencies produced, expressed as a percentage of the total undistorted output signal level. In many cases, the RMS value of noise is added to THD (expressed as THD+N). The lower this percentage is, the better. THD+N values over 0.1% can often be heard, and suggest a less than adequate sound board design.

This figure is related to harmonics. When two or more tones are generated together, amplifiers create harmonics, as well as tone combinations. For example, if a 1kHz and 60Hz tone are mixed together, intermodulation harmonics will be generated (i.e. 940Hz, 880Hz, 1060Hz, 1120Hz, and so on). It is this intermodulation which gives sound a harsh overtone. Since intermodulation is not related to sound quality, it is a form of distortion which should be kept to a very low level. Like THD, intermodulation distortion (IMD) is the RMS sum of all unwanted harmonic frequencies expressed as a percentage of the total undistorted output signal level. IMD should be under 0.1% on a well-designed board.

While it does not directly effect the fidelity of sound reproduction, sensitivity can be an important specification. Sensitivity is basically the amplitude of an input signal (such as a microphone signal) that will produce the maximum undistorted signal at the output(s) with volume at maximum.

By itself, sensitivity is hard to apply to a sound board, but if you consider the board’s output power versus its input signal power and express the ratio as a decibel, you would have the gain of the sound board. Many sound boards offer a potential gain of up to 6dB. However, it is important to note that not all sound boards provide positive gain - some boards actually attenuate the signal even with the volume at maximum. In practical terms, this usually forces you to keep the volume control at maximum.

An ever-growing number of sound card owners are using their sound cards to record sound, or broadcast sound over the Internet through such applications as WebPhone. Sound recording demands the use of microphones, and not all microphones work properly with every sound board. Often, the user mistakes a poor microphone response as being a problem with the sound card. This part of the chapter looks at some important considerations for choosing and using a microphone.

There are three types of microphones: dynamic, condenser, and electret condenser. You will find all three microphone types available for sound boards:

• Dynamic - Dynamic microphones are typically hand-held or desktop units. They have a larger response range and typically sound better than condenser microphones. A dynamic microphone does not require phantom power because the diaphragm element in the microphone can create enough electric current for the sound board to use.
• Condenser - Condenser microphones are the small multimedia microphones that typically come with computers. When you open a new sound board and take the microphone out of the box, it is almost always a condenser microphone. They do not have as good a response range as dynamic microphones, and they also have a smaller diaphragm - this demands phantom power for the sound board.
• Electret Condenser - Electret Condenser microphones are basically condenser microphones with a built-in battery for power. They have the same response as a condenser microphone, but they do not require phantom power to operate. Some electret condenser microphones will allow you to remove this internal battery. With the battery not installed, phantom power would be required.

So the next question is "what IS phantom power?" Phantom power is simply a small, low-current power supply on the sound board which is used to power some microphones. Devices like dynamic microphones can produce enough current on their own to avoid the use of phantom power, but condenser microphones demand phantom power as a current source.

Here’s the main problem with today’s sound boards - not all of them provide switchable phantom power. Ideally, sound boards (like the Ensoniq Soundscape) would provide phantom power, and allow you to jumper the phantom power on or off depending on which microphone you plan to use. If you use a dynamic microphone, you’d switch phantom power off. If you use a condenser microphone, you’d switch phantom power on. When a sound board does not provide phantom power at all, you’re stuck using a dynamic microphone, or a powered electret condenser microphone. If a sound board provides full-time phantom power (and you cannot turn it off), you’ll need to stay with a condenser microphone.

You can probably see the potential for trouble here. If you use a condenser microphone on an unpowered sound board, the microphone will not work at all (or generate little more than faint noise). On the other hand, plugging a dynamic or electret microphone into a powered sound board will usually result in severe clipping - once again, you’ll capture little more than noise.

Whether you’re choosing a microphone for yourself, or recommending one to someone else, there are some considerations to keep in mind. Perhaps the most important issue is the application. If you just need a basic, inexpensive microphone to record a few simple voice notes, a condenser or electret microphone would work just fine, and your sound board will require a phantom power supply. If you want to record more professional vocals, or preparing a presentation, a dynamic microphone will generally provide the best results, and no phantom power is needed.

Traditionally, sound boards use many of the same chipsets and basic components, but since each board is designed a bit differently, it is very difficult for commercial diagnostic products to identify failed IC functions. For the most part, commercial and shareware diagnostics can only identify whether a brand-compatible board is responding or not. As a result, this chapter will take the sub-assembly replacement approach. When a sound board is judged to be defective, it should be replaced outright. This part of the chapter reviews the problems and solutions for sound boards under both DOS and Windows. The following tips may help you nail down a sound problem most efficiently:

• Check to see that your speakers are connected, powered, and turned on.
• Check that the speaker volume and sound card master volume are turned up.
• Check to see that the mixer volume and master volume are set properly.
• Make sure that the music or sound file(s) are installed properly.
• Check that all sound card and multimedia drivers are installed.
• Make sure that the drivers are up to date.
• Check for resource conflicts between the sound card and other devices in the system.
• Make sure that the sound card is selected and configured properly (especially for DOS apps).
• The sound device should be enabled and configured under CMOS (for sound functions incorporated on the motherboard).

Unlike most other expansion devices that are driven by system or supplemental BIOS, sound boards make use of small device drivers to set up their operations. These drivers are generally included in CONFIG.SYS and AUTOEXEC.BAT, and are called when the system is first initialized. Most sound board drivers are only used to initialize and set up the board, so they do not remain resident - this is good since it reduces the load on conventional and upper memory. However, these initialization routines vary from board to board. For example, the files installed for a Creative Labs Sound Blaster will not support a Turtle Beach MultiSound board. When you elect to replace a sound board, you must also disable any current sound board drivers, and include any new supporting driver files. The process is not difficult - just follow the installation instructions for the board - but the software consideration does add another wrinkle to the replacement process.

When there are problems installing or upgrading a sound board, one of the first issues to suspect is the driver loading order. Sound boards are typically multi-function devices which require several drivers in CONFIG.SYS and AUTOEXEC.BAT. If the drivers are installed in the wrong order, the sound board (or other features of the board) may not function. As a rule, the drivers should be loaded in the following order AFTER your memory managers:

• The sound board’s device driver:

• The CD-ROM port setup driver (if the sound board is equipped):

• The CD-ROM driver (if the sound board is equipped):

Many current sound board designs are compatible with "multimedia communication" technologies such as Internet Phone, Webphone, and communication tools. These tools require full-duplex operation - that is, sound is digitized with the microphone and received sound is played through the speakers simultaneously. This demands full-duplex drivers. If you plan to use communication tools, you’ll need to install full-duplex sound card drivers that are appropriate for your particular sound board and operating system. For example, the Creative Labs SB32, AWE32, and AWE64 require the windows 95 full-duplex driver file (SBW95UP.EXE) available from the Creative Labs web site (www.creaf.com). To use those same devices for full-duplex under Windows NT 4.0, you’d need the AWENT40.EXE driver file. As a rule, always check with the sound board maker for current full-duplex drivers.

NOTE: You may find that full-duplex drivers are NOT available for older sound boards, or sound boards running under OS/2 and Windows NT. In that case, you cannot support full-duplex applications.

Of all the sound board problems reported, perhaps the most common is the failure to play wave files (ordinary sound files with the .WAV extension) under Windows 95. This problem usually manifests itself during the Windows startup or shutdown when the accompanying sounds are not played. There are a variety of issues that can prevent .WAV files from playing.

Program-specific problems - If you cannot play .WAV files from a specific program that you use in Windows 95, check to see if the same problem occurs when you play the file from another program. If the problem occurs only with one particular program, the files associated with that program may be damaged, or that program may not be configured correctly under Windows 95. If you cannot get .WAV files to play under ANY application, chances are that another issue is responsible.

Sound device is not configured properly - If you cannot play any .WAV files in Windows 95 (or if .WAV files are not played at the proper volume), you may not have a sound device selected, or the sound device that you have selected may not be configured properly. To select and configure a sound device in Windows 95:

• Open the Control Panel and double-click the Multimedia icon.
• In the Playback area under the Audio tab (Fig. 47-8), click the playback device that you want to use in the Preferred Device list, and then move the Volume slider to the value you want (usually 50-75% volume is adequate).
• In the Recording area under the Audio tab, click the playback device that you want to use in the Preferred Device list, and then move the Volume slider to the value you want.
• Make sure that the speakers are properly connected to the sound card, and that the speakers are turned on.

Mixer settings are not configured properly - If you cannot play any .WAV files under Windows 95 (or if .WAV files are not played at the proper volume), the mixer control settings may not be configured properly. You can use the mixer control program included with Windows 95 to adjust the volume for playback, recording, and voice commands. To configure mixer control settings for Windows 95:

• Click the Start button, point to Programs, point to Accessories, point to Multimedia, and then click Volume Control (Fig. 47-9).
• Make sure that the "Mute All" check box below the Volume Control slider and the "Mute" check box below the Wave slider are not selected, and that the Balance sliders for Volume Control and Wave are in the center of the scale.
• Move the Volume Control and Wave sliders at least halfway to the top of the scale. You may need to adjust the current Volume Control or Wave settings to play .WAV files at the volume level you want.

NOTE: If the Volume Control and Wave sliders do not appear, click Properties on the Options menu, and then click the Volume Control and Wave check boxes in the "Show the following volume controls" box to select them.

The sound hardware is not configured properly - It is possible that your sound card may not be compatible with the type of .WAV file you are attempting to play, or there may be a resource conflict between your sound card and another device installed in your computer. Check the Device Manager to see if there are any resource conflicts with your sound board. To determine whether your sound card supports the .WAV file format you are attempting to play, contact the sound card’s manufacturer.

The sound files are damaged - If you cannot play certain .WAV files in Windows 95 (or if the .WAV files are not played properly), the .WAV files themselves may be damaged. To check if a .WAV file is damaged, use the right mouse button to click the .WAV file in Windows Explorer, click Properties on the menu, and then click the Details tab. The Audio Format line should contain information about the type of compression used to compress the file, the sound quality of the file, and whether or not the file is in stereo. If this information is missing, the .WAV file is probably damaged, and should be reinstalled or recopied to the drive.

NOTE: If you can play other .WAV files of a similar format, chances are good that the suspect file is indeed damaged. If you can play .WAV files of different formats, but NOT .WAV files of a particular format, it may be that your sound board does not support the particular format.

Compression-related problems - Windows 95 includes 32-bit versions of several common CODECS including Adaptive Delta Pulse Code Modulation (ADPCM), Interactive Multimedia Association (IMA) ADPCM, Group Special Mobile (GSM) 6.10, Consultative Committee for International Telephone and Telegraph (CCITT) G.711 A-Law and u-Law, and Truespeech from DSP. These 32-bit CODECS are installed by default during Windows 95 Setup, and are used by multimedia programs even if a 16-bit version of the same CODEC is available. Make sure that .WAV file format is supported by an available CODEC. Otherwise, you may need to install an appropriate CODEC.

Symptom 47-1. There is a noticeable buzz or hum being produced in one or both speakers. Low-cost speakers use unshielded cables. Unfortunately, strong signals from AC cords and other signal-carrying conductors can easily induce interference in the speaker wires. Try rerouting speaker cables clear of other cables in the system. If problems persist, try using higher-quality speakers with shielded cables and enclosures. In most cases, that should resolve everyday noise problems. If the noise continues regardless of what you do, there may be a fault in the sound board amplifier. Try moving the sound board to another bus slot away from other boards or the power supply. If that does not resolve the problem, try a new sound board.

Symptom 47-2. There is no sound from the speaker(s). The lack of sound from a sound board can be due to any one of a wide range of potential problems. If the sound board works with some applications, but not with others, it is likely that the problem is due to an improperly installed or configured application. See that the offending application is set up properly (and make sure it is even capable of using the sound card). Also check that the proper sound driver files (if any) are loaded into CONFIG.SYS and AUTOEXEC.BAT as required. In many cases, there are one or two sound-related environment variables that are set in AUTOEXEC.BAT. Make sure that your startup files are configured properly.

Check your speakers next. See that they are turned on and set to a normal volume level. The speakers should be receiving adequate power, and should be plugged properly into the correct output jack - if speakers have been plugged into the wrong jack, no sound will be produced. If the cable is broken or questionable, try a new set of speakers. Also see that the master volume control on the sound board is turned up most (or all) of the way.

If problems continue, there may be a resource conflict between the sound board and another device in the system. Examine the IRQ, DMA, and I/O settings of each device in the system. Make sure that no two devices are using the same resources. You might like to use the PC Configuration Form at the end of this book to record your settings. If problems persist, and no conflict is present, try another sound board.

Symptom 47-3. CD audio will not play through the sound card. This problem can occur under both DOS and Windows. First, make sure that the sound board is actually capable of playing CD-audio (older boards may not be compatible). If the sound card is playing sound files, but is not playing CD-audio, there are several things for you to check. First, open the PC and make sure that the CD-audio cable (a thin, 4-wire cable) is attached from the CD-ROM drive to the sound board. If this cable is broken, disconnected, or absent, CD-audio will not be passed to the sound board. If the cable is intact, make sure that the CD-audio player is configured properly for the sound board you are using, and check the startup files to see that any drivers and environment variable needed by CONFIG.SYS and AUTOEXEC.BAT are available. If the CD-audio fails to play under Windows, make sure that an MCI (multimedia control interface) CD Audio driver is included in the Drivers dialog box under you Windows Control Panel.

Symptom 47-4. You see an error such as; "No interrupt vector available". The DOS interrupt vectors used by the sound board’s setup drivers (usually INT 80h to BFh) are being used by one or more other drivers in the system. As a consequence, there is a software conflict. Try disabling other drivers in the system one at a time until you see the conflict disappear. Once you have isolated the offending driver(s), you can leave them disabled, or (if possible) alter their command line settings so that they no longer conflict with the sound board's software.

Symptom 47-5. There is no MIDI output. Make sure that the file you are trying to play is a valid MIDI file (usually with a .MID extension). In most cases, you will find that the MIDI Mapper under Windows is not set up properly for the sound board. Load the Windows MIDI Mapper applet from the Control Panel, and set it properly to accommodate your sound board.

Symptom 47-6. Sound play is jerky. Choppy or jerky sound playback is typically the result of a hard drive problem - more specifically, the drive cannot read the sound file to a buffer fast enough. In most cases, the reason for this slow drive performance is excessive disk fragmentation. Under DOS, the sound file(s) may be highly fragmented. Under Windows, the permanent or temporary swap files may be highly fragmented. In either case, use a reliable DOS defragmenter such as PC Tools or Norton Utilities (leave Windows before defragmenting the disk), and defragment the disk thoroughly.

Symptom 47-7. You see an error such as; "Out of environment space". The system is out of environment space. You will need to increase the system’s environment space by adding the following line to your CONFIG.SYS file:

shell=c:\command.com /E:512 /P

This command line sets the environment space to 512 bytes. If you still encounter the error message, change the E entry to 1024.

Symptom 47-8. Regular "clicks", "stutters", or "hiccups" occur during the playback of speech. This may also be heard as a "garbled" sound in speech or sound effects. In virtually all cases, the system CPU is simply not fast enough to permit buffering without dropping sound data. Systems with i286 and slower i386 CPUs typically suffer with this kind of problem. This is often compounded by insufficient memory (especially under Windows) which automatically resorts to virtual memory. Since virtual memory is delivered by the hard drive, and the hard drive is much slower than RAM anyway, the hard drive simply can't provide data fast enough. Unfortunately, there is little to be done in this kind of situation (aside from adding RAM, upgrading the CPU, or changing the motherboard). If it is possible to shut off various sound features (i.e. music, voice, effects, and so on), try shutting down any extra sound features that you can live without. Make sure that there are no TSRs or other applications running in the background.

Symptom 47-9. The joystick is not working, or not working properly on all systems. This problem only applies to sound boards with a multi-function MIDI/joystick port being used in the joystick mode. Chances are that the joystick is conflicting with another joystick port in the system. Disable the original joystick port or the new joystick port - only one joystick port (game adapter) can be active at any one time in the system. Since joystick performance is dependent on CPU speed, the CPU may actually be too fast for the joystick port. Disable the joystick port, or try slowing the CPU down.

Symptom 47-10. You install a sound board and everything works properly, but now the printer does not seem to work. There is an interrupt conflict between the sound board, and an IRQ line used by the printer. While parallel printers are often polled, they can also be driven by an IRQ line (IRQ5 or IRQ7). If the sound board is using either one of these interrupts, try changing to an alternative IRQ line. When changing an IRQ line, be sure to reflect the changes in any sound board files called by CONFIG.SYS or AUTOEXEC.BAT.

Symptom 47-11. You see a message; "Error MMSYSTEM 337: The specified MIDI device is already in use". This problem often occurs with high-end sound boards such as the Creative Labs AWE64. This error is often caused by having the sound board’s mixer display on with the wavetable synthesizer selected (i.e. the LED display in the Creative Mixer turned on and Creative Wave Synthesizer selected as the MIDI playback device). You can usually correct the problem by turning the mixer display off.

Symptom 47-12. You see a message; "Error: Wave device already in use when trying to play wave files while a MIDI file is playing". This problem often occurs with high-end sound boards such as the Creative Labs AWE64, and is usually the result of a device configuration problem. If "full-duplex" is turned on and you try to play a .WAV file and a MIDI file at the same time with the wavetable synthesizer (i.e. the Creative Wave Synthesizer) selected as the MIDI play back device, an error will occur. To resolve this problem, you need to turn off the full-duplex mode:

• Hold down <Alt> key and double-click on My Computer.
• Select the Device Manager tab. There should be a listing for "Sound, Video, Game Controllers" in the Device Manager, double-click on the listing to expand it.
• You should now see a listing for sound board (i.e. Creative AWE32 16-Bit Audio). Double-click on the listing, then select the Settings tab. Un-check the box labeled "Allow full-duplex operation". Click OK until you are back to the Control Panel.
• Now try to play a .WAV and MIDI file at the same time.

Symptom 47-13. You hear "pops" and "clicks" when recording sound under Windows 95. There is insufficient cache to adequately support the recording process (or cache is improperly configured). Try the following procedure to alter the way cache is allocated:

• Open Notepad and load SYSTEM.INI
• Locate the area of SYSTEM.INI labeled [vcache].
• Add the following line below [vcache];

• Save your changes to the SYSTEM.INI file.
• From the desktop, right-click on My Computer, then select Properties.
• Select the Performance page, then click on File System.
• Find the slider marked "Read-ahead optimization", then pull the slider to None.
• Save your changes and restart Windows 95.

Symptom 47-14. You notice high frequency distortion in one or possibly in both channels. In many cases, the AT Bus Clock is set over 8MHz, and data is being randomly lost. This problem usually occurs in very fast systems using an ISA sound board. Enter the system’s CMOS Setup and check the AT Bus Clock under the Advanced Chipset Setup area. See that the bus clock is set as close as possible to 8MHz. If the bus clock is derived as a divisor of the CPU clock, you may see an entry such as /4. Make sure that divisor results in a clock speed as close to 8MHz as possible. If problems still persist, try increasing the divisor to drop the bus speed below 8MHz (note that this may have an adverse effect on other ISA peripherals).

Symptom 47-15. You hear "pops" and "clicks" when playing back pre-recorded files under Windows 95. There is an excessive processing load on the system which is often caused by virtual memory and/or 32-bit access. Start by disabling virtual memory; open the Control Panel and double-click on the System icon. Select the Performance page and click on Virtual Memory. Set the swap file to None and save your changes. Try the file playback again. If problems persist, try disabling 32-bit file access. If that still does not resolve the problem, try disabling 32-bit disk access.

Symptom 47-16. You hear "pops" and "clicks" on new recordings only, pre-existing files sound clean. This is often due to issues with software caching. If you are using DOS or Windows 3.1, disable SmartDrive from both CONFIG.SYS and AUTOEXEC.BAT, then restart the computer for your changes to take effect. If problems continue (or you are using Windows 95), there may be an excessive processing load on the system due to virtual memory or 32-bit access. Follow the recommendations under Symptom 47-15.

Symptom 47-17. You hear "pops" and "clicks" when playing back or recording any sound file. In most cases, there is a wiring problem with the speaker system. Check all of your cabling between the sound board and speakers. If the speakers are powered by AC, make sure that the power jack is inserted properly. If the speakers are powered by battery, make sure that the batteries are fresh. Check for loose connections. If you cannot resolve the problem, try some new speakers. If the problem persists, replace the sound board.

Symptom 47-18. The sound board will playback fine, but it will not record. The board probably records fine in DOS, but not in Windows. If the sound board is using 16-bit DMA transfer (typical under Windows), there are two DMA channels in use. Chances are that one of those two DMA channels are conflicting with another device in the system. Determine the DMA channels being used under Windows, then check other devices for DMA conflicts. If you are using Windows 95, check the Device Manager and look for entries marked with a yellow icon.

Symptom 47-19. A DMA error is produced when using a sound board with an Adaptec 2842 controller in the system. This is a known problem with the Digital Audio Labs "DOC" product and the Adaptec 2842. You will need to alter the controller’s FIFO buffer. Go in the controller’s Setup by hitting <Ctrl>+<A> when prompted during system startup. Select the advanced configuration option, then select the FIFO threshold - chances are that it will be set to 100%. Try setting the FIFO threshold to 0 % and see if this makes a difference.

Symptom 47-20. A DMA error is produced when using a sound board with an Adaptec 1542 controller in the system. This is a known problem with the Digital Audio Labs "DOC" sound product and the Adaptec 1542. The problem can usually be resolved by rearranging the DMA channels. Place the Adaptec controller on DMA 7, then place the sound board on DMA 5 for playback, and DMA 6 for recording.

Symptom 47-21. The sound card will not play or record - the system just locks up when either is attempted. The board will probably not play in either DOS or Windows, but may run fine on other systems. This is a problem that has been identified with some sound boards and ATI video boards. ATI video boards use unusual address ranges which sometimes overlap the I/O address used by the sound board. Change the sound board to another I/O address.

Symptom 47-22. The sound card will record, but will not playback. Assuming that the sound board and its drivers are installed and configured properly, chances are that a playback oscillator on the sound board has failed. Try replacing the sound board outright.

Symptom 47-23. The sound application or editor produces a significant number of DMA errors. This type of problem is known to occur frequently when using the standard VGA driver that accompanies Windows - the driver is poorly written and cannot keep up with screen draws. Try updating your video driver to a later, more efficient version. If the driver is known to contain bugs, try using a generic video driver written for the video board’s chipset.

Symptom 47-24. The sound board will not record in DOS. There are several possible problems that can account for this behavior. First, suspect a hardware conflict between the sound board and other devices in the system. Make sure that the IRQs, DMA channels, and I/O port addresses used by the sound board are not used by other devices.

If the hardware setup appears correct, suspect a problem between DOS drivers. Try a clean boot of the system (with no CONFIG.SYS or AUTOEXEC.BAT). If sound can be run properly now, there is a driver conflict. Examine your entries in CONFIG.SYS or AUTOEXEC.BAT for possible conflicts, or for older drivers that may still be loading to support hardware that is no longer in the system.

Finally, suspect the hard drive controller. Try setting up a RAM drive with RAMDRIVE.SYS. You can install a RAMdrive on your system by adding the line:

device=c:\dos\ramdrive.sys /e 8000

The 8000 is for 8MB worth of RAM - make sure there is enough RAM in the PC. Once the RAMdrive is setup, try recording and playing from the RAMdrive (you may have to specify a new path in the sound recorder program). If that works, the hard drive controller may simply be too slow to support the sound board, and you may need to consider upgrading the drive system.

Symptom 47-25. When recording sound, the system locks up if a key other than the recorder’s "hot-keys" are pushed. This is a frequent problem under Windows 3.1x. The system sounds (generated under Windows) may be interfering with the sound recorder. Try turning off system sounds. Go to the Main icon, choose the Control Panel, then select Sounds. There will be a box in the lower left corner marked "Enable system sounds". Click on this box to remove the check mark, then click OK.

Symptom 47-26. After the sound board driver is loaded, Windows locks up when starting or exiting. In virtually all cases, you have a hardware conflict between the sound board and another device in the system. Make sure that the IRQs, DMA channels, and I/O port addressed used by the sound board are not used by other devices.

Symptom 47-27. When using Windows sound editing software, the sound board refuses to enter the "digital" mode - always switching back to the analog mode. Generally speaking, this is a software configuration issue. Make sure that your editing (or other sound) software is set for the correct type of sound board (i.e. an AWE32 instead of a Sound Blaster 16/Pro). If problems persist, the issue is with your sound drivers. Check the [drivers] section of the Windows SYSTEM.INI file for your sound board driver entries. If there are more than one entry, you may need to disable the competing driver. This is a known problem with the Digital Audio Labs CardDplus, and is caused by incorrect driver listings. For example, the proper CardDplus driver must be entered as:

Wave=cardp.drv

and the companion driver must be listed as:

Wave1=tahiti.drv

You will need to make sure that the proper driver(s) for your sound board are entered in SYSTEM.INI. You may also need to restart the system after making any changes.

Symptom 47-28. The microphone records at very low levels (or not at all). Suspect your microphone itself. Most sound boards demand the use of a good-quality dynamic microphone. Also, Creative Labs and Labtec microphones are not always compatible with sound boards from other manufacturers. Try a generic dynamic microphone. If problems persist, chances are that your recording software is not configured properly for microphone input. Try the following procedure to set up the recording application properly under Windows 95:

• Open your Control Panel and double-click on the Multimedia icon.
• The Multimedia Properties dialog will open. Select the Audio page.
• In the "Recording" area, make sure to set the Volume slider all the way up.
• Also see that the Preferred device and Preferred quality settings are correct.
• Save your changes and try the microphone again.

Symptom 47-29. The sound card isn’t working in full-duplex mode. Virtually all current sound boards are capable of full-duplex operation for such applications as Internet phones. Check the specifications for your sound board and see that the board is in fact capable of full-duplex operation. If it is, and full-duplex isn’t working, your audio properties may be set up incorrectly:

• Open your Control Panel and double-click on the Multimedia icon.
• The Multimedia Properties dialog will open. Select the Audio page.
• If the Playback device and the Record device are set to the same I/O address, this is only Half Duplex.
• Change the Playback device I/O address so it is different from the Record device.
• Hit the Apply button, then hit the OK button.
• You should now be in Full Duplex mode.

NOTE: Some of the very latest sound boards (such as the Ensoniq SoundscapeVIVO 90) will carry full duplex operation with the same Playback and Record device selected.

Symptom 47-30. You encounter DMA errors using an older sound board and an Adaptec 1542. In many cases, you can clear DMA issues by slowing down the 1542 using the /n switch. Add the /n switch to the ASPI4DOS command line in CONFIG.SYS such as;

device=c:\aspi4dos.sys /n2

If slowing the 1542 down with an /n2 switch doesn't fix the problem, then you should strongly consider upgrading the sound board. This is a known problem with the older Digital Audio Labs CardD sound board.

Symptom 47-31. You encounter hard disk recording problems under Windows 95. Recorded audio is saved to your hard drive. For most systems, sound data can be transferred to the HDD fast enough to avoid any problems - if data transfer is interrupted, your recorded sound may "pop" or break up. There are many factors that effect HDD data transfer speed. The following sections outline a number of procedures that might help you optimize a system for sound recording. First, try disabling the CD Auto Insert Notification feature:

• Go to the Device Manager
• Open the CD-ROM entry
• Select your CD-ROM drive, and click Properties
• Go to the Settings page
• Uncheck the "Auto insert notification" box
• Select OK

Next, try turning down the level of graphics acceleration:

• Right-click on the My Computer icon
• Left-click on Properties
• Select the Performance page
• Select the button labeled Graphics
• Start by turning down the acceleration one notch (you can return later to turn it down further if more performance is required)
• Select OK

It may also be necessary to adjust the size of your virtual memory swap file:

• Right-click on the My Computer icon
• Left-click on Properties
• Select the Performance page
• Select the button labeled Virtual Memory
• Choose "Let me specify my own virtual memory settings"
• If your PC has 16MB of RAM, set the minimum and maximum at 40MB. If you have 32MB of RAM, set the minimum and maximum at 64MB.
• Select OK

Try removing any active items from your Startup group:

• Click the Start button
• Go to the Programs menu, then select Startup. If you see anything here, it may be hurting your system performance. Eliminate anything that is not absolutely necessary.
• To remove items, click the Start button, go to Settings, then select Taskbar
• Choose the page labeled Start Menu Programs
• Click the Remove button
• Open the Startup group by double-clicking it
• Remove any items which you feel are not necessary and are wasting resources
• Select the Close button when finished

Clear any indexes of the Find Fast utility:

• Click the Start button
• Go to Settings, then choose the Control Panel
• Open the Find Fast utility
• Go to the Index menu
• Select Delete Index…
• Select an index in the In and Below drop box
• Select the OK button
• Repeat steps 5 through 7 until all indexes are removed

Try defragmenting the hard drive:

• Click the Start button
• Go to Programs, Accessories, and System Tools
• Choose Disk Defragmenter
• Select the drive to defragment and click OK
• Click the Start button to begin defragmentation

Finally, you may want to suspend the System Agent (if installed):

• If System Agent is installed, open it by double clicking its icon in the taskbar
• Go to the "Advanced" menu
• Choose "Suspend System Agent"
• Close the System Agent window.

Symptom 47-32. The microphone records only at very low levels, or not at all. Check your phantom power settings first. In many cases, the microphone’s gain is set too low in the sound board’s mixer applet. Start the sound board’s mixer, make sure that the microphone input is turned on, then raise the microphone’s level control. Remember to save the mixer settings before exiting the mixer. You should not have to restart the system.

Symptom 47-33. Your dynamic microphone clips terribly, and recordings are noisy and faint. This is probably due to phantom power being switched on in your sound board. Try turning the phantom power off. If you cannot turn phantom power off, try plugging the dynamic microphone into the sound board'’ Line Input jack. Remember to start the sound board'’ mixer applet and set the Line Input level properly.

Symptom 47-34. You have trouble using Creative Labs or Labtec microphones with your (non-Creative Labs) sound board. This is a common complaint among Ensoniq sound board users. It turns out that Ensoniq sound boards are not compatible with Creative Labs or Labtec microphones. Try a generic microphone instead.

Symptom 47-35. There is static at the remote end when talking through a voice application such as WebPhone. Noise is occurring at the Line Input or Microphone Input which is being transmitted to the remote listener. Check the Line Input signal. You might try reducing or turning off the Line Input mixer level. If the problem persists, check your phantom power setting and your microphone. Try reducing the microphone level in the sound board’s mixer. Try a different microphone.

This concludes the material for Chapter 47. Be sure to review the glossary and chapter questions on the accompanying CD. If you have access to the Internet, take a look at some of the sound board resources listed below:

Creative Labs: http://www.creaf.com

Turtle Beach: http://www.tbeach.com/

Frontier Design Group: http://www.frontierdesign.com/

Ensoniq: http://www.ensoniq.com

SIC Resource: http://www.sicresource.com/

Star Multimedia: http://www.starusa.com/