Tungsten from hot filaments evaporates and adheres to the glass envelope, blackening and dimming them in conventional tungsten lamps.
Gasses are added on tungsten halogen lamps to chemically combine with the evaporated tungsten for re-depostion on cooler parts of the filament... not always exactly in the places that it is lost, but that's another story.
https://www.electrical4u.com/tungsten-halogen-lamps/ has quite a good explanation of the process. It does suggest that TV studio lighting where camera line up fader 7 was 2700 Kelvin (and allowed +/- 3 on the faders for level control with CT of +/- 500 K) would mean no 'recombination' of W back on to the filaments at lower fader levels. But the glass was unlikely to be cool enough to deposit the Iodide, either (especially quartz glass as the lamps were invariably smaller/closer to the hot filaments.
In practice there was no issue with dimming any T-H lamps, whether quartz or hard glass when I was involved in the 'game'.
Most common failure of quartz envelopes was at the pinch seal(s), frequently due to overheating of the contacts. That's why the lamp makers had specifications wrt the lamp holder heat dissipation required. Pinch seal failure of course meant the halogens escaped and air entered. The other common user error was touching the quartz envelope with bare hands.
