Mullard used a few dodges to ensure that blemished face plates were not used in tube manufacture such as when fitting the EHT connection in the cone of the 12 inch tube, this would always be located immediately above the gather or shear marks so that when the tube was fitted in the set, the tube mask would hide these marks - pretty clever.
Otherwise, it was a matter of 100% inspection and selection and this is how it was done. Firstly, only blemishes which could be seen from a distance of 3 feet 6 inches distance either under external lighting or with a raster on the tube were considered.
Any blemish which would be hidden by the tube mask could be disregarded. The planar surface of the faceplate was divided into a number of zones with the specification for the central area being more stringent than for the edges. Consideration was given to size location and distribution of blemishes such that they would have no obvious effect on picture viewing. In this way, Mullard enjoyed a les than 1.5% failure rate on screen blemishes. We'll close today's blog entry with this photograph of a Mullard man going 'bog eyed' looking at screen blemishes and phosphor coating faults: -