You could do it if you didn't need to limit the overall to 1000ms. You could also do it if you do not expect valid responses with the first 200ms window. In that case, you could simply list 200ms as the time and then take the RT to the mask.

If none of this is getting you anywhere, another possibility might be to simply have the initial stimuli and the mask as subsequent trials. You could probably get by with reducing your inter-trial interval to 10ms or so and it would appear relatively seamless. Just a thought..

The *real* answer to this, btw, is for us to add the capability to define multiple min-max values to a single trial. FWIW, that is a goal for the next version.