Hi James,
I have two audio recordings of lectures from the 129th AES convention last year in San Francisco that do an excellent job summarizing the technical difficulties that currently exist for both high resolution audio and computer audio playback.
The first was a Master Class presentation on High Resolution Audio delivered by Keith O. Johnson. The second was a panel seminar called Progress in Computer-Based Playback of High Resolution Audio, chaired by Vicki R. Melchior. The panelists were Bob Bauman of Lynx Studio Technology, James Johnston, DTS Inc, Andy McHarg, dCS Ltd and Daniel Weiss, Weiss Engineering Ltd.
Both of these events were extremely informative and helped shape my thoughts on computer-based audio playback while reflecting on how the information relates to the Bryston BDP-1 player. Please let me know if you'd like a copy of the audio files in order to listen to the lecture's yourself.
Here are summaries of both lectures from the AES archives:
MASTER CLASS PRESENTATION ON HIGH RESOLUTION AUDIO BY KEITH O. JOHNSON:
Computers, televisions, and mobile devices are functionally merging and integrating to become easy to set up fun to run systems. Right now, server-like versions can play high-resolution multichannel files from workstations, function as a control center, create loudspeaker crossovers and equalization, as well as perform room correction using all loudspeakers. Data management and processing create these advanced features but systems like these can present issues with processing activity, sample rate conversion, jitter, noise propagation, digital conversion, and interfaces. One encounters discussion about perceptual differences from technical changes that should not affect accuracy nor produce large differences in timing spectra. The class will briefly introduce systems and components, then show potential behavioral artifacts within system parts and describe test methods that might reveal explanations. Then we'll explore process monitoring, buffers, quick locking low jitter clocks, floating conversion environments, jitter from OP Amps and a test methodology that targets process related intrinsic behavioral problems. Cluster, subtraction and DSP overload tests will be described along with projected human perceptual and mental load that might have audibility or hinder playback involvement. This presentation of an overall background knowledge should encourage studies that are more detailed and it should be helpful with creation and development of very high quality systems.
Progress in Computer-Based Playback of High Resolution Audio:
With the continuing decline in discs as music sources and concurrent growth of electronic distribution, computers and network attached storage (NAS) are now rapidly evolving as front end components in place of traditional transports and players. Computers have long been useful within mastering workflows, though not always loved, and their introduction into high quality music systems raises a new range of engineering challenges.
Intrinsic to computers are problems of EMC, switching noise, dirty power, jittered clocks, crosstalk, driver and operating system variability, protocol incompatibilities, and software errors, to name a few. These may directly influence audio quality. Of special importance, for example, are the design as well as system configuration of digital audio interfaces (USB, Firewire, S/PDIF, WiFi, Ethernet etc), D/A conversion, and data processing, along with clocks and power sourcing.
The panel in this workshop are active in the design of these systems and will discuss some of their results and thoughts regarding the most salient factors for optimization of sonic performance in this area.
Regards,
David
