Ever since I first read about the quantum eraser experiment, I was wondering: if after the "measurement" the wave function collapses and we can observe a particle - does that particle remain a particle forever? How far is the screen from the measuring device? What if you put it 100km away? What about 1ly away? Or 1pc away? Will we still see the two distinct lines of particle impact, or will it at some point become an interference pattern?
These are different questions, but I will adress the core issue here, which I think is a missconception about what is really the Wave-particle duality about. Quantum Mechanics doesn't say that objects are both particle and waves at the same time (wavicles some people say), nor that they are waves in some circumstances and particles in others (albeit it could depened on the intrepretation you chose). The idea is that quantum objects have features that can be explained partially taking aspects of classical models of particles and waves, but these are neither nor both. Quantum objects are a completely different kind of object that we can't model uniformely as particles nor waves, but can only describe by using the behaviour of the Schrödinger equation (which by the way is not in the same form of a D'Alembert's wave equation so it is not propagating lika a classical wave either, another generalised missconception). There is no transformation between the particle bahviour and the wave behaviour, it is just that the "amount of existence" (the quantum probability) of the "particle" in each point in space is characterized by a map which behaves almost like a wave throught time (the Shrödinger equation), but the probability collapses to a certain point in space when uncertainty is lowered, by an actual observation, then its "amount of existence" would shrink to a tiny volume, so we tend to say "it is located here", as we would say about a particle. Is just that the particle reference is an analogy, it really isn't a particle. There is no conversion between the two behaviours, just bad language applied to something we can't model with everyday life analogs or comparative models.