The MicroEye was reviewed in PCW in October 1983 and Stephen and I wrote an article which was published in Practical Computing the same month. Although there were no software listings, there were comprehensive and detailed descriptions of the algorithms we used and how we compressed, stored and analysed images in tiny 32K microcomputers. We held back on discussing vectorising edges of objects, although hinted at how we did this and suggested the MicroEye had applications in CAD. We regretted this later. However, it was a series of advertisements in laboratory equipment magazines, for the Sirius version of the MicroEye and MicroScale, that gave us a foothold in the scientific image processing market.

First to knock on our door was Wellcome Research, who needed a ‘low-cost’ system to measure the growth of tissue cultures. They wanted the MicroEye interfaced to a HP 9816 68000 processor-based PC (a snip at over £5,000). We developed a custom version of MicroScale for Wellcome, then a second version that more closely resembled our own product. The MicroEye needed a small modification to make it work with the HP 9816: a resistor to pull down one of the control lines. This increased the production cost of the MicroEye from £32 to £32.01. However, to bring the retail price into line with the price of the HP 9816, we increased it from £295 to £1,295. Aside from our MicroScale software, the HP version of the MicroEye was our highest margin product.

We had pricing credibility issues with the Sirius 9000, as no one would take a £295 peripheral seriously in the 16-bit computing market. So, the price of the MicroEye increased to £495 across the board. Stephen wrote a version of MicroScale for the Sirius using the C programming language, and it became clear that the performance of the software now outstripped that of the MicroEye.

We were repeatedly asked if the MicroEye could be modified to capture images in real time. Given that we had only just mastered image scanning, image grabbing seemed a distant dream. However, British Telecom had developed a 512 x 512, 256 greyscale framestore. The company planned to use the device, which could store two images, in CCTV systems and was hoping to find someone prepared to manufacture it under licence. So, one bright, big sky, East Anglian Monday morning, I drove to Ipswich to meet Tony Lancaster in the sprawling fortress that was BT’s Martlesham Heath laboratory. Had someone warned me about Tony, we might have got the licence. He had a habit of laughing out loud prior to delivering bad news, which made negotiating with him almost as impossible as it was surreal ...

(An extract from The Ghost in the Labyrinth by Peter Kruger)

Next chapter ...