Sound Card Basics

Although there are a huge number of sound cards around, there are certain standard features common to all. These are discussed here, along with an explanation of what a basic sound card is capable of and how it works. This is a good place to start if you want to understand what your card is capable of but if you bought your card some time since 1998 you will almost certainly have more features than listed here. These will be covered in your card’s Specs entry. Even if you do have a new card you could still benefit from reading this article, if only because you’ll appreciate the latest advancements on your model.

Connectors

The standard connectors include a joystick connector that doubles as a MIDI interface. This is where you’d attach a MIDI keyboard and any external modules you have, although you’ll need to use the special cable usually supplied with the card and available from most computer shops.

There are also four 35mm jacks, one each for a set of stereo speakers, a microphone and line-level input and outputs. Some cards also have small, square onboard sockets that connect to standard CD-ROM drives. Audio CDs can be played by the drive, the signal running through the card and out to the speakers. You may also see one or two other onboard connectors which, on older cards, may be an IDE CD-ROM connector (handy if your ancient PC only has one IDE port) or a wavetable expansion connector to which an additional MIDI synthesiser board can be attached. External boards are often better sounding than built-in MIDI synthesisers and its samples may be easier to edit.

Types Of Synthesiser Sounds

The average, basic sound card will provide an FM MIDI synthesiser and since mid-1996 most popular models feature a wavetable synthesiser too. The difference is that FM (frequency modulation) synthesis creates approximations of instrument sounds by generating and combining electronic tones whereas wavetable synthesisers use recordings, or samples, of real instruments. These samples are played faster or slower to get different pitches. Samples have to be stored somewhere and older cards use onboard memory to hold their sounds. Some have sockets for extra memory chips that make more space for larger, more realistic samples. Later cards, such as Creative Labs’ SBLive! range and the Aureal Sonic Vortex-based cards load wavetable samples into the PC’s system memory. This is more flexible and can be much cheaper.

Numbers of Synthesiser Sounds

The number of MIDI sounds your card can play simultaneously is limited. If a synthesiser can play more than one sound at the same time it is a polyphonic instrument. A card with 32 voice polyphony will be able to play 32 sounds at the same time and playing 33 notes on an attached keyboard will force the synthesiser to drop one automatically.

Another limitation is the number of types of sounds the card can play simultaneously. Most standard sound cards have 16 MIDI channels, best imagined as instruments such as piano, drum kit, violin etc. It doesn’t follow that every piece of music you write is limited to 16 instruments, though, because most sequencer programs can change the instrument associated with a channel mid-song.

The polyphony rule applies to the total number of sounds played, not to each channel so you don’t get 16 lots of 32 voices. For example, if at any one point in a song the piano part is playing eight notes as a chord the other 15 possible instruments must share the remaining 24 notes (assuming your card only has 32 note polyphony).

Digital Audio

Sound cards also have the ability to record and play digital audio. A common example of digital audio would be a .WAV file. The explosions you hear in modern games are also digital audio samples. Sound cards can record through either the microphone or the line-level input (Line In). Recording in this way, straight to memory and then hard disk, is just like recording to tape although the quality of sound will be better on the PC than with most home tape decks. Sounds stored on a hard disk will not degrade over time as they will on a decaying tape. Old cards can only record to memory, not directly to the hard disk, but if your model is post 1996 you should be OK.

Sample Rates and Resolutions

Also unlike tape decks, sound cards let you choose the quality of sound you record - up to a point. This quality is defined by sampling rates and resolution. Sampling, or recording, involves regularly checking and copying the incoming analogue signal. The more often it checks the more faithful the final recording will be. In addition to the sampling rate, another factor called resolution has a lot to do with the recorded sound quality. An 8 bit resolution means that when sampling a sound at any one time the computer can only choose from 256 possible values. A 16 bit resolution increases the number of possible values to 32,000. Recording at 8 bit will introduce many inaccuracies to the sound which will manifest themselves as background noise. Imagine the sound you get from a badly-tuned radio and you won’t be far off.

The original PC sound cards could only record and play at 11kHz or 22kHz at a resolution of 8 bits, while more modern cards can manage at least 44.1kHz at 16 bits. If you have one of these later cards (and the chances are that you do) try recording a sound into Windows’ Sound Recorder program and save it first as a 44.1kHz, 16 bit sound then as an 11kHz, 8 bit sound. As suggested by the program, the difference is similar to CD quality compared to telephone quality.

Some very expensive sound cards offer 20 bit resolution. As CDs can only hold 16 bit audio tracks you might wonder why more highly specified equipment exists. The main reason is that when a PC applies effects to a recording it is doing lots of maths and the more numbers it gets to play with the more accurate the final results will be. Even if the finished tracks will be downgraded to 16 bit for CD use they will still be more accurate than those created from original samples recorded at 16 bit.

Noise

If cheap, modern cards can record high-quality audio straight to disk you might wonder why professional musicians spend hundreds of pounds on special cards that seem to offer less features. One reason is that usually the cheaper the card, the more noisy it is. The measurement of a card’s noisiness is its signal-to-noise ratio. A high ratio is good because a loud signal will overcome a small amount of noise. Recordings made with a low signal-to-noise ratio will feature conspicuous hissing in the background because the sounds you want to hear cannot compete effectively with the noise you wish to eliminate.

Noise can come from a number of sources including the card’s own onboard amplifier and the electrical noise that exists inside every PC case. If a card has an amplifier, as most old ones do, it will boost both the sounds you want and the noise picked up from elsewhere. More modern cards lack an amplifier for this reason.

Anatomy Of A Basic Sound Card

A standard Creative Labs Sound Blaster 16.

1. Line-In Input
2. Microphone Input
3. Line-Out Output
4. Stereo Speaker Output
5. Joystick/MIDI Port
6. CD Input
7. IDE Connector
8. OPL-3 FM Synthesiser

Author: Simon PG Edwards 20 July 1999

Have you found your way here from a search engine or link from another site and cannot see the site navigation tools at the top of the page?
Visit www.PCMusic.Org to experience this site properly.