Audio Latency Measurements 14Jun23

Differing time delays from processing in varying gear in audio reproduction is a huge issue, as it is therefore in designing the NestT CC system. I don't have full pro gear so a synchronizing timestamp isn't passed amongst the pieces.  Processing time delay differences among the miniDSP DDRC22, the TV, the Denafrips DAC, and the Dolby DP564 to separately drive L/R mains through one chain and CC/LS/RS through another chain are a problem. This wouldn't be the case in a common single 5.1 AVR (receiver) as it is responsible for everything and can adjust timings internally, even accounting for differing distances of the speakers to the listener with a calibration mic. I researched if the companies whose gear I'm using specified what their gears' delays were and this answer is "rarely yes." The specs, or more often the communities in which gear is discussed, described latency but it was rarely definitive, due to differences in what specific processing you were gonna be doing in each piece. If one is just doing passthru that's one thing but if applying filters or doing decodes then that's often longer, and by varying amounts depending on the number of filters, and even sampling rates & bit depth of material changes a given device's latency. First off I reduced the source to just the miniDSP DDRC22 out, and then adjusting the time delays by ear in experiment 3 helped immensely; almost 10msec was required! I realized a more expansive and coherent front sound stage, even with mis-matched speakers and a digital amp. However, my experience begged to now actually measure the latency of my components and guide the delay adjustments more scientifically. How? Enter RTL, round trip latency, a measurement I can make closed loop with a mac or PC driving an impulse signal through each piece of gear separately and then feeding it back into the computer via a USB audio interface. A number of software apps make this measurement fairly simply (most often in support of balancing a subwoofer) and accurately and I chose Oblique Audio's RTL Utility to do it. I use a Behringer UM22 audio interface to get mics into the Mac so used that to get the analog stereo signals from some gear looped back into the mac USB for the RTL Utility to analyze. I have cables that convert 1/4" instrument plugs to standard RCA.

I isolated each of the pieces of gear for RTL measurement, including the USB audio interface itself as well as the mac itself. For whatever reason, the impulse measurement technique failed through the miniDSP DDRC-22D, so I didn't get delays through it.

For the digi in - digi out special case of of the miniDSP boxes, I simply used a fiber-to-USB converter to isolate it alone, or used the USB audio interface ~ I measured both ways to double check results for these key pieces. I'm hoping a display port to HDMI converter will allow me to ultimately measure the latency of my Sony TV from HDMI in to its optical audio out ~ though that's not a variable for center channel quality when using strictly the miniDSP DDRC22 output as a source for the DP564 decoder, I just want to know. Here are the results...

What do you notice here besides my scientific calc of the delay I need to add to the CC for processing and proximity? Well, most of my stuff is quiet enough but signal loss through the miniDSP gear is high, but maybe that is from processing and not true through loss, and it's concerning through the USB interface as well. Second, while the Pontius R2R DAC sounds awesome, it is slow as heck and attenuates to boot. Anyway... I have my 12.2msec net number now and will design delays and filters in the miniDSP to compensate... done.

UPDATE 26Jun23... I received the mini-display port to HDMI dongle (with audio) but could not get the app to pass the test signal through it cleanly and as a result I could not measure the TVs latency.