|
OK so why don't we take a couple of examples in detail then. A specific rev of a specific converter and a specific clock with a specific length of a specific cable. Have a few of these combinations in hand. All reproducible by any of us.
Let's develop a good test signal that has a combination of test tones...sines, squares, saws of both polarities, double sines (e.g. 17KHz and 19KHz), digital black, noise, impulses, phase interactions, and a variety of program material. Let's run it though each of these setups set both to internal clock and external clock via DAC/ADC loopback. Let's do several runs on each configuration, alternating in time to avoid e.g. thermal effects. Then let's take those signals and null them to the original signals sent in. Take a look with Audio Precision or Spectrafoo or whatever.
Is there a difference? What is it? What configuration is closest to the originals? What is furthest? What difference displays the lowest, most even noisefloor? What is the distribution of harmonic distortion, and how does it change at different frequencies?
Invite skeptics to contribute to the protocol. Maybe someone knows of particular test signals or program material that the converters have a hard time with, or that reveals differences best. We could use this protocol on everything and get a very good idea of what we're working with.
Give us something tangible. Where the rubber hits the road. You might convert me. If I'm $400 or $1300 or whatever from truly better audio, then you're not doing either of us a favor hiding proof of benefit. If you've got something you can point to that really works, I'll risk paying for the return shipping and the lost hours and buy the stuff and try to reproduce the test.
|