It’s a nested loop for each cone trace, so yes I assume it’s the cone trace taking the bulk performance.