Copyright 2017 Robert Clark
Some possibilities for altitude compensating nozzles include actual adjustable-sized nozzles but using non-flexible materials such as ceramics. In such cases the high expansion ratios needed for optimized expansion at high altitude would give nozzles of impractical size, for instance they wouldn't fit within the width of the rocket stage.
A couple of possibilities for dealing with this eventuality:
Vacuum optimized nozzles, whether altitude compensating or not can take a great deal of mass of an engine or stage. See for example the specifications of the Star 48B here:
You see the weight of the nozzle assembly is nearly that of the case assembly. This would become even more of a weight problem with extreme expansion ratios of hundreds to one.
The lightweight space shuttle underside tiles may provide a solution. Their volume density is only 0.144 gm/cm^2. And according to this report their areal density is 1.19 gm/cm^2:
TPS Materials and Costs for Future Reusable Launch Vehicles.
The AETB-8/TUFI listed is a toughened tile material that has higher impact resistance while maintaining the same temperature resistance.
Judging from the size of the size of the Star 48B nozzle, the nozzle weight might be reduced from 90 lbs to ca. 20 lbs using the AETB-8/TUFI tiles.
A possibly even more lightweight material was developed by aerospace engineer/mathematian GW Johnson. He described it in this video presentation at the 16th Mars Society conference in 2013:
Reusable Ceramic Heat Shields - GW Johnson - 16th Mars Society Convention.
Johnson estimated the volume density as only 0.03 gm/cm^3, a third that of the shuttle tiles. Also interesting is that Johnson originally developed the material to act as insulation for ramjet combustion chambers.
Note that even the insulation in the Star 48B solid motor is a sizable weight at 60 lbs, compared to the 129 lbs casing weight. Then Johnson's ceramic might also be able to be used as a lightweight replacement for solid motor insulation.