Sorry, no – you have to expect a new utterance to begin after a continuous live utterance that is around 30 seconds. It actually used to be a much longer gap between live utterances and has become a lot shorter in the last few RapidEars updates so I’m a bit surprised to hear it’s an issue now. Where in the logs are you seeing that it’s a second? Keep in mind that .7 seconds is the detected pause length so that duration is always going to be a factor in the separation between two utterances.
What is unusual about this case is that a normal user would rarely be capable of speaking for around 30 continuous seconds without a notable pause. People don’t really speak that way in practice unless they are reading from a text and intentionally trying not to pause at all. If you want to prompt a faster separation between utterances, pick a slightly shorter secondsOfSilenceToDetect like .5 so that it happens more regularly and not at the point that RapidEars has detected what it thinks is a stuck utterance, which is what happens when one single live utterance exceeds 30 seconds, since that is a rarity in normal practice and generally indicates that something is wrong (like a loud continuous noise in the background that isn’t expected to stop).