Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by MediaGary

  1. Thank you Jim [ I promise *not* to call you Mr. Roseberry again :-) ] for your pithy summation. You have prevented me from digging too deep into the weeds on this topic to develop an answer that fits my neurotic approach to technical questions. I was reading through this document in my non-existent 'spare' time [ https://www.mindshare.com/files/ebooks/pci express system architecture.pdf ] In its 1038 pages, it covers all the stuff about PCIe function, but of course leaves it to the reader to infer how an "aggressive" driver would cause citizenship problems within the overall system. This document may still be handy for my project to get a GC-Titan-Ridge Thunderbolt working properly in my 2010 Mac Pro. We'll see. Okay, back to the original topic! Thanks to all!
  2. Your instinct is right in that a change of driver is necessary, but that aspect of that driver behavior pertinent to audio performance is to find a video card/driver that's "agile" in letting go of the PCIe slot. The video card driver of the R7 250 is possibly "hogging" the PCIe slot, but that behavior is not specifically related of the number of active lanes within that slot. There is an audio forum that I frequent that has good things to say about AMD RX580 cards regarding their compatibility with audio performance. I have an Nvidia GTX 1070 card in my machine which is supposed to be less good in that respect, but it's working fine with no audio issues. Many people have done well for audio work using the native Intel graphics functions. Here is a link to a long thread on the issue. In short, there are confirmations and counter-cases, so it's nothing absolute in either direction: [https://www.gearslutz.com/board/music-computers/1212416-dpc-latency-better-amd-graphic-cards-3-card-comparison.html ]
  3. As Mr. Roseberry points, out, their ADAT A/D conversion is typically under 1-millisecond of delay. I have collected data about ADAT conversion latency for a couple of products that I own: - The ART TubeOpto-8 is 32-samples A/D latency at 44.1k (ref: ART User Manual). - The Audient ASP800 is 39 samples (ref: Audient tech support). From that point onward, it's traditional to ignore any propagation delays in the fiber optics or wires because of the small distances involved. To offer some pedantic info, signals propagate at about (Velocity Factor = light speed in a vacuum) 0.65-VF within fiber optic transport, and at about 0.95-VF for signals in copper wire. Data rate is often confused with propagation speed and gets us into confusing discussions about latency for audio. My best thought experiment that helps to clear the air is to walk through the propagation delay issues that the Mars missions must endure. No matter the data rate in kilobits or megabits, the propagation delays from Earth to Mars are an average of about 14 minutes each way; dependent on the relative positions of the two planet in their respective orbits. Best case is about 4 minutes, and worst case is about 24 minutes. Okay, pedantic itch is scratched! Carry on!
  • Create New...