For those interested, I performed some roundtrip latency measurements in standalone mode. I measured this as follows:
- Use my Korg NTS-2 to generate a skewed square wave at a very low frequency (so it’s easy to match phase).
- Connect the NTS-2 output to its first input. This serves as the reference signal (t=0).
- Connect the NTS-2’s first thru signal (with a direct connection to its first input, i.e. no A/D + D/A conversion) to the Push’s first input.
- Set up the Push in standalone mode with only a single Audio device that listens to Input 1 and outputs to the external out.
- Connect the Push’s external out to the NTS-2’s second input.
- Estimate the latency as the time difference (delay) between the rise of the square wave of the baseline signal (input 1) and the signal returned by the Push (input 2).
This got me the following:
| Sample rate (kHz) | Buffer size (samples) | Roundtrip latency (ms) |
|---|---|---|
| 44.1 | 32 | 7.4 |
| 64 | 9.8 | |
| 128 | 15.6 | |
| 256 | 27 | |
| 512 | 39 | |
| 48 | 32 | 7.0 |
| 64 | 9.2 | |
| 128 | 14.4 | |
| 256 | 25 | |
| 512 | 36 | |
| 88.2 | 32 | 5.2 |
| 64 | 6.6 | |
| 128 | 9.6 | |
| 256 | 15.2 | |
| 512 | 27 | |
| 96 | 32 | 5.0 |
| 64 | 6.2 | |
| 128 | 9 | |
| 256 | 14.2 | |
| 512 | 25 |