Ensuring the quality of cellulose nanofibers (CNFs), particularly in terms of dispersibility and defect density, is crucial for fully utilizing their potential. While CNF dispersibility can now be adjusted by selecting appropriate raw material species or employing specific chemical pretreatments, controlling defects remains a challenge. In our previous study, we found that the dent defects exist on the CNFs, and these defects could not be reduced by any mechanical disintegration method [1]. Therefore, in this study, we aim to control the defects through the selection of raw material species and the application of optimized pretreatment methods. Two raw materials of softwood bleached kraft pulp and softwood holocellulose were chosen in this study from the perspective of the nanoscale supramolecular structure of the materials. These two materials were subjected to a regioselective oxidation reaction using 4-acetamido-2,2,6,6-tetramethylpiperidinyl 1-oxyl as a catalyst at pH 5 or 10 to prepare four types of CNFs. Based on a statistical analysis of CNF defects, such as dents and fragmentation, using image processing for atomic force microscopy (AFM), we show a process that produces CNF dispersions comprising truly uniform single crystallite units of the highest quality to date. The wavelet transform was further applied to the AFM height profiles to analyze the location and periodicity of the right-handed twisting of the CNFs. We also demonstrate that the defectless CNFs possess a regulated twist periodicity for a greater part of the entire lengths.