Holographic storage has the potential to become a digital data storage technology with fast readout and high density. Computer users have come to expect, however, that data retrieved from their storage devices will be retrieved error- free (with a probability of error less than 10-12). In both conventional storage devices and holographic data storage, achieving this degree of reliability involves a good understanding of the data channel and a combination of careful hardware engineering, signal processing, and coding. At the IBM Almaden Research Center, we have leveraged the expertise acquired with 1-dimensional, time-dependent data channels found in magnetic and optical data storage systems, to develop unique and highly effective signal processing and coding algorithms to optimize the performance of the 2-dimensional, space-dependent digital holographic data storage channel. Crucial to our efforts has been the high-performance holographic data storage platform we built in 1996. This tool has allowed us to characterize and perturb a real holographic data channel, and implement and evaluate new data-coding and signal processing algorithms. This rapid feedback loop between ideas, implementation, and results both aids in selecting fruitful approaches and yields deeper understanding of the underlying data channel. In this paper, we discuss the holographic digital data storage channel as divided into five parts: the optical path, pre-processing (how the data gets into the holograms), post-processing (manipulation of raw data just after optical detection), conversion into binary 0's and 1's, and error-correction (using added redundancy). Optimizing the channel involves maximizing the system performance (density, speed) while minimizing complexity (and thus cost) and maintaining the required degree of reliability.