Maybe you compose what you what to say mentally or "under your breath" and the UT then feeds it back to you phonetically and you repeat the sounds.
Get good enough at it and you can compose, speak, listen to the native word, listen to your translation, all at the same time.
Just takes practice.
