Early solid state sound recording courtesy of Captain Bob

The late Robert Maxwell was thinking ahead of the game with solid state memory recording.

In 1987 Toshiba’s engineers in Japan announced that they had developed a portable player that looked like a Sony Walkman but used solid state memory cards instead of tapes.

The system was called IC-Voice and Robert Maxwell backed it to the tune of £2 million through a company called International Learning Systems Japan Ltd. ILS ordered 50 000 IC-Voice units to be delivered over three years. ILS planned to use the gadgets for langauge tuition.

Education courses, to be produced by the BBC, would be stored in Read Only Memory chips. Each ROM stored eight minutes of pre-recorded speech, half used for 80 master sentences in one language (e.g. English) and the other half for the same sentences in another language (e.g. German).

The card also contained a RAM chip which could store 32 seconds of speech. So students would listen to a key phrase, speak it into RAM and then compare their own pronunciation with the permanently-stored example in the ROM.

Ten years ago, audio data compression was in its infancy and memory was expensive. So Toshiba economised by using different coding standards for the ROM and RAM chips.The ROM had a capacity of 16 Megabits, and the pre-recorded speech streamed at 33 kilobits/sec to give reasonable sound quality.

A 256k RAM chip recorded at 8 kbps, which seriously coarsened the sound. Everyone kept predicting that memory prices would fall. But it did not happen, largely because the computer industry created a vicious circle of bloat, with new software needing more and more memory to run. Later Maxwell used CD-ROMs to store data.

Technology later caught up with the solid state idea, and memory prices spiralled downwards. Solid state recording is now the way of the world. But all this came too late for Cap’n Bob’s foresighted investment in IC-Voice to pay off.