The following latency tests were conducted after hearing a significant increase in round trip latency compared to what was reported in Live Professor. After taking latency measurements I discovered that real world round trip latency changed depending on the USB Audio in/out selection on the WING. The interesting part through all of this is that the reported round trip latency in Live Professor remained constant. As shown in the table below 32x32 was the best performing.
| Test Name | Latency (ms) | Notes |
|---|---|---|
| Local in to out | .54 | Measurement ref #1 |
| Ch. (no processing) | 1.29 | Measurement ref #2 |
| Ch. (processing) | 1.29 | Measurement ref #3 - HPF, DESSER, EQ, COMP |
| USB 2x2 EXT Insert | 26.73 | WING channel with processing and external USB insert. USB 2x2 -> Intel Macbook Pro -> Live Professor (direct in to out patch) @64 samples -> USB |
| USB 8x8 EXT Insert | 13.40 | Same as previous with USB 8x8 |
| USB 16x16 EXT Insert | 11.54 | Same as previous with USB 16x16 |
| USB 32x32 EXT Insert | 9.79 | Same as previous with USB 32x32 |
| USB 48x48 EXT Insert | 9.85 | Same as previous with USB 48x48 |
| WDante 2x2 EXT Insert | 14.58 | Same as previous but using Dante Card on the wing with DVS set to 2x2 @ 4ms |
| M1 Mac Full Chain | 9.67 | WING channel with processing and external USB insert. USB 48x48 -> M1 Macbook Pro -> Live Professor @64 samples -> Waves Sibilence -> Waves F6 -> USB |
| M1 Mac Direct in to out | 9.64 | Same as previous but direct input to out to bypass plugin chain |
- All testing was done in Open Sound Meter
- Measurements can be found in OSM folder
- Measurement rig - MOTU M2 with looped output 2 into input 2 for signal ref
- LP buffer size is constant in all tests that inlcude it - 64 samples
- Console FW 3.05