In the experiment, the subjects hear a brief audio instruction to either do a task or simply not do anything (REST). So the model I used is
LR L R REST INS LR' L' R' REST' INS'
where ' denotes second timepoint scans and INS is the audio instruction that is just in the model but otherwise not used in any contrast.
yes, the model is for a single subject.
I have attached a realign parameters file if it helps!
Many thanks for your kind help,