A still picture needed constant work
A desktop CRT made light by firing electrons at a phosphor coating inside the glass. Magnetic deflection steered the beam across the screen, line by line, from the top towards the bottom. The video signal varied its intensity to draw the picture. Even when a document stayed perfectly still, the beam had to keep drawing it.
The coating continued glowing briefly after the beam moved on. Engineers call that property persistence. The light then faded until the next visit from the beam. If those brightening and fading cycles remained visible, the apparently steady desktop flickered. The changing brightness was built into the way the display produced an image; it did not require an animation or a faulty application.
Two different kinds of speed
Refresh rate describes how often the picture is renewed. In a non-interlaced mode, a higher rate brings the beam back to each part of the image sooner and generally reduces visible flicker. Resolution describes how many picture elements make up the image. Raising one setting is therefore different from raising the other.
There was another complication: interlacing. An interlaced display drew alternating sets of lines in separate passes. Philips’s manual gives a 50 Hz example: interlacing produces 50 fields, but only 25 complete frames, each second. A non-interlaced 50 Hz mode draws 50 complete frames. The same-looking frequency label can describe different scanning schedules.
Why the phosphor mattered
Keeping the phosphor glowing longer could bridge the interval between scans. That also meant an old position might remain visible when text scrolled or an object moved. GRAD’s technical guide describes this trade-off as smearing and tailing. Shorter persistence reduced that residue but made frequent refresh more important.
Philips also singled out the light backgrounds of graphical interfaces and desktop publishing software as places where frame flicker was noticeable. A blank-looking white page still involved repeated light pulses across a large area.
The result depended on the phosphor, scan mode, image and observer. There was no single refresh number that made every CRT look identical to every person. Higher refresh and a suitable phosphor worked together to make a stationary page appear stationary.