The TTS-1 problem at 96k is a known issue.
If you want to use the TTS-1:
Work at 48k (with TTS-1)... and bounce down the resultant audio.
Convert the audio to 96k and import into the desired 96k project.
You may want to bounce the tracks from the top of the tune (to make it easy to align) you can trim away silence after the audio is imported into the 96k project.