Investigating an issue with the DT168.
Below are SMAART traces where we fed the DT168 outputs pink noise via a console and directly from known good analog to Dante sources.
What’s interesting is that at 48k the artifacts (phase and HF rolloff) occur consistently, regardless of who is Dante clock master (including the DT168), and regardless of if the source comes from local SQ inputs, a MR8R (directly or via dLive), or a Sound Devices MD-4, or pink from SMAART directly. When in 48k mode, the artifact is always there, the HF phase might vary slightly but the magnitude is always about 2.5dB down at 19.5kHz.
Tested both DT168s that I have here and both came up identical.
When run in 96k mode from the same sources (the ones that can do 96k natively) or via 48k sources converted to 96k via an SQ or dLive doing the conversion - the response traces are flat - flat enough to not question
This doesn’t appear to affect the DT168 inputs - they seem to do SRC just fine and look flat.
The first two lines are at 96k, the bottom three are at 48k. I just spaced them out on the graph so they’d be easy to view.
Thoughts? We are almost entirely Dante - and we pretty much live in the 48k world as do a lot of Dante networks because of compatibility. SQ for example I can never run at 96k and use most of the wireless we have since its 48k only, so the console Dante card is stuck at 48k and there’s only one card. I’m leaning toward not using the outputs on the DT168 for FOH or other important items, unless/until AH can fix this - as this seems like something that should be fixed.
Note DT168 shows as up to date in Dante Updater. Other devices are generally up to date as well - but the issue seems specific to the DT168.
Additionally looping a DT168 input back to a DT168 output on the same unit, seems to additionally affect the HF response, which is a valid configuration for a Dante Brooklyn/Broadway chipset
SMAART graph is 3dB per line
