Fields of technology as diverse as microwave filter construction, characterization of material interfaces with atomic precision, and detection of gravitational waves from astronomical sources employ mechanical resonators at their core. The utility of mechanical resonators can be explained by their sensitivity to a multitude of force fields and actuation mechanisms, and the existence of extremely precise position measurement techniques. The possibility of faithful, real-time reconstruction of mechanical motion, especially when the resonator is monitored interferometrically or with cavity spectroscopy, has enabled the creation and engineering of quantum states of macroscopic mechanical resonators, in the fields of quantum optomechanics, levitodynamics and quantum acoustodynamics. These experiments must however overcome the large thermal occupation of mechanical oscillators in typical conditions, due to contact with high-temperature baths. A fruitful path to counteract the thermal fluctuations governing the motion of mechanical oscillators has been the engineering of the quality factor (Q) of the mechanical mode of interest, which coincides with the control of dissipation and isolation from its environment. The class of methods termed dissipation dilution, implemented in strained, high-aspect-ratio nanoresonators, allow the control of the mechanical Q through mode shape engineering, and underpinned in the last decade an astounding growth in mechanical coherence levels. In this thesis, I report on the development of new types of ultracoherent mechanical resonators leveraging dissipation dilution, united by a a common goal of preparing quantum states of such resonators at room temperature. We improved the mechanical performance of dissipation-diluted resonators by utilizing single-crystal materials with high strain and aspect ratio, and investigated the performance of strained silicon as a material for nanomechanics. Combining soft clamping in phononic crystal (PnC) structures with the low friction of crystalline materials, we witnessed record-high mechanical quality factors beyond 10 billion at MHz frequencies, at temperatures around 7 K, corresponding to phonon lifetimes of several hours. We have also endeavoured towards observing quantum optomechanics at room temperature, by exploiting mechanical resonators with high dissipation dilution. We assembled a room temperature membrane-in-the-middle experiment, where a soft clamped PnC membrane interacts dispersively with the optical modes of a Fabry-Pérot cavity. In our first studies of this system, we witnessed an intensity noise mechanism vastly exceeding the magnitude of quantum noise of the employed light field, and understood it to be due to the sizeable thermal motion of the membrane resonator, distorted by the nonlinear cavity transduction. We termed this noise source ``Thermomechanical Intermodulation Noise'' (TIN), and devised methods to reduce its magnitude below the vacuum fluctuations of the light field. The limitations of our room temperature optomechanics experiment could also be overcome with new types of mechanical resonators. I report on the development of hierarchical membrane resonators with a partially soft clamped fundamental mode, exploiting the notion of wave amplitude suppression over a three-beam joint, and PnC membrane resonators with density modulation, which offer enticing performance for ground state cooling from room temperature in our experiment.