So, let’s look at all the options for how to create this golem =)
Green lines indicate a DI signal, red - processed. To calculate the latency, you need to sum the delay every time a signal passes through a block that involves buffering.
Option number one, the most unfortunate one, it will not only add a slight latency, but also make the Apollo purchase completely pointless, in whatever order the interfaces are connected.
Option number two will allow you to make a high-quality recording of DI and processed sound. And also monitoring through Apollo high-quality outputs, but this will be the biggest latency - 13ms.
And option three, a hybrid, will allow you to record high-quality DI and processed sound, and get a fairly low latency when monitoring, only 6 ms, but through less high-quality outputs of Quad Cortex. In general, this will delay the monitoring line by only 3 ms than when using only the Quad Cortex, in my opinion it is quite acceptable.
