but wouldn't we see the variability occur in both modes (record and play?) And why not friction if record mode (which also "plays") brings the record head in contact with the tape causing additional tension/friction -- could this not account for the variation happening in record mode only?As many tape decks can drop in and out of record by a button - there is no difference in tape tension when recording or playing. Friction has no impact on tape speed on a machine that is set up and operating correctly - as we've said. tape speed is controlled by the capstan and pinch wheel. If this is set so that the spool motors can pull it through, they do so at a crazy rate - maybe 30ips plus, not close to the normal speed. The capstan must be changing speed if the replay speed is not the same as the record speed - why is this? Presumably it's simply faulty and the speed control is not working properly - usually the servo monitors the speed it's turning and if it is going to fast slows it down and if too slow speeds it up - so quickly the speed is to all intents and purposes constant. The problem we are talking about here is NOT friction, or slippage, it's loss of capstan stability.