Which came first - Pirates of the Caribbean, or Gladiator?
Should you switch phantom power off if it isn't needed?
Why you will be beaten to success by people who are no better than you
How to become a better singer
Famous concert pianist plays a wrong note!
A very unusual tape recorder used for mastering
"Treasin Compilation Mixtape Vol.1" by 2G-GottiCapon
New vs. old guitar strings: Part 1 - The case for new guitar strings
Extraordinary stereo from your effects pedal
How to keep master processing to a minimum
Whereas once upon a time you could play a perfectly acceptable game with very simple equipment, now the recommended minimum is a Pentium P90 equipped PC with CD-ROM drive, a decent amount of RAM and hard disk space, graphics card, sound card and stereo speakers. You may have noticed that there is a special type of loudspeaker that you are apparently meant to use with a computer that are beige in colour, small, and particularly puny sounding. You don't have to use speakers like these, but they are almost forced on customers in PC hypermarkets and form yet another limitation that Electronic Arts', and other games producers', sound designers have to get over. In the early days of CD-ROM games, although the software would of course be loaded into RAM, the greater part of the graphics and sound data would be lifted as required from the CD. As you know, CD-ROM drives are slow compared to hard disk drives therefore the data rate is limited. To throw animated graphics up onto the screen at a reasonably fast frame rate gobbles up a lot of the speed available and therefore something in the audio is going to have to give. Possible options are to cut the sampling rate, reduce the resolution, or go from stereo to mono. The first thing that goes is the sample frequency, down from 44.1kHz to 22.05kHz, with the resulting halving of audio bandwidth. With typical computer speakers, this might go unnoticed. If this still doesn't allow sufficient bandwidth for the graphics, then sixteen bits can be cut down to eight, and stereo can become mono - which actually might be quite acceptable for speech or sound effects. All of these parameters can be tailored for each element of audio so fortunately it isn't necessary to make a global decision for the whole of the game. Also, don't forget that a mono signal can be panned by the software, and Q-Sound too can convert an assembly of single channel samples into a virtual 360 degree experience.
Squeezing audio into a narrow data channel isn't simply a matter of chopping out the redundant data and throwing it away. For instance, if speech has to be reduced to eight bits, then compression is vital, and some time would be spent working out the optimum degree of limiting, compression and EQ to squash the signal as much as possible while retaining maximum intelligibility and perceived quality. Reverb doesn't really work very well in eight bits since long tails inevitably fall apart and draw attention to the defects of the audio. As far as music is concerned, although synthesised instruments can sound quite good at reduced resolutions, real instruments, and any music that attempts a degree of subtlety will suffer greatly. To get around these problems to a certain extent, Electronic Arts are letting the computer's hard disk take some of the strain, because if material can be loaded from the CD-ROM onto the hard disk, then it can be accessed much faster, and it is likely that before long eight bit audio will be a thing of the past, thankfully. Also, since double speed CD-ROMs will undoubtedly be replaced by 4x speed drives over the next couple of years then this particular bottle neck will be widened. One difference between games makers and music and video producers is that where the latter will consider the limitations of the equipment people have at home, and tailor their product to work within these limitations, games producers tend to force the pace and almost demand that their customers buy faster PCs, and faster everything else, every eighteen months or so.