It's always tricky working out a way to execute a stimulus presentation while an RT is still being taken. In the case you are describing, you might try the following.

First, create a new version of that subsequent sound--one that has 300ms of quiet at the start of the sound file. Then, play that file immediately *before* the target with a time value of 0. That will cause the sound file and the subsequent target to execute simultaneously, but the RT will be taken to the target. Meanwhile, the sound will continue playing and you would thus hear the tone at 300ms into the RT. Do you think that would work for you?

-Blair