R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Best Transport: FireWire, USB, or AES/EBU ?  (Read 4999 times)


  • Newbie
  • *
  • Offline Offline
  • Posts: 4
Best Transport: FireWire, USB, or AES/EBU ?
« on: March 14, 2006, 07:11:22 PM »

I would really like to see an audiophile-quality product with a self-clocked DAC using FireWire (or USB) transport for input. I'd like to explain why I think this would be the ultimate DAC, provided it has enough outputs for the application.

First of all, within the constraints of a DAC fed by AES/EBU or S/PDIF, I believe that technologies like CrystalLock are probably the best that can be done in a system where the data is "pushed" towards the DAC. Given the design of most CD players, along with many other pieces of professional gear, this is a necessary evil (just like word clock is a necessary evil when multiple independent converter boxes need to work together for large numbers of outputs).  But pushing audio data is not the only possible design.

Computers are capable of working with a "pull" model, where the output convertor is in control of the data flow. In fact, Apple's Mac OS X has CoreAudio technology which has turned the audio software industry upside down, almost literally, by changing the old push model to a pull model in the modern DAW. Even when "playing" an audio CD, typical problems of buffer control can be managed by the software because it can read the data from the CD using an outside source for flow control, rather than trying to generate a clock from the spinning disc.

Because the data flow control problems are solved when using something bidirectional like FireWire or USB, it seems that all of the Lavry Engineering expertise with designing ultra-low jitter onboard clocks would result in the ultimate reference (outboard) audio convertor for computer sources.

Other feature ideas:

* Adding ADC, also based on the same internal clock, would allow Pro Tools and other DAW software to bounce through external analog processing with minimal distortion from jitter.
* A design with low-latency DAC and ADC chips would be met with massive approval. In this case, I am referring to today's complex conversion chips which introduce many samples of delay for digital processing, as compared to older converters with only about one sample delay. Typical interfaces today use convertor chips with many samples of delay, but a high-end design avoiding these chips would have an advantage in the market.
* A modular design allowing for expandability from a basic stereo DAC to multichannel outputs might be even more important than combining ADC. Expandability would allow additional outputs without requiring that a word clock be distributed outside the unit.

I welcome comments, especially if I am missing something. Are the existing FireWire interfaces with internal clocking already performing as well as a Lavry Engineering DAC? Is it only the task of removing jitter from external clocks that requires the black magic? Smile  I imagine there are a lot of circuits out there, even self-clocked, which do a bad job of designing the clock circuits, leaving room in the market for a Lavry Engineering interface which excels in quality.

I see a few other topics in this forum that are basically saying the same thing, but they seem to have fizzled out. I'd like to know whether Lavry Engineering has any improvements to offer in this area.

Brian Willoughby
Sound Consulting
Pages: [1]   Go Up

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.024 seconds with 19 queries.