For animals to navigate a given space, it is essential to perceive and memorize the external allocentric frame with landmarks. In the previous Part I (Arisaka 2022), by applying causality and locality at every single synapse, we proposed a new concept of MePMoS. This model allowed constructing the complete space-time dynamic connectomes of C. elegans and human brains. This Part II proposes a generic principle of space-to-time conversion for perceiving 3D space and 3D shape in a frequency-time domain, named Neural Holographic Tomography (NHT). In this general model, three strings of one-dimensional neurons can encode 3D space, utilizing the spike timings expressed by the phases of brainwaves. Therefore, dynamic shifting of phases can accomplish natural translation from an ever-changing egocentric frame acquired from sensory input to the perception of a stable allocentric frame, resulting in the steady perception of allocentric 3D space. We further propose a universal unit of memory, an engram named HAL (Holographic Ring Attractor Lattice). Facilitated by Hebbian plasticity, the phases of brainwaves, representing 3D space and shape, are registered by the 2D matrix of synaptic networks to form long-term static memories. This two-step conversion of sensory stimulation, first to a brainwave phase by NHT and second to synaptic memory by HAL, provides a general principle for perception and memory of allocentric 3D space and semantic 3D shape.