I'm interested in getting the distance between Cα for all residues in an alignment. I've noticed that the Cα RMSD values in the txt file obtained from the header of the sequence alignment created by matchmaker via  

sequence header rmsd save 

are different from the distances measured using the rmsd or distance command. The Cα distance measured is approx. 41% higher across the entire alignment than in the saved output. When selecting a pair of aligned residues from the alignment viewer, the distance shown does not match the one in the output file but the distance measurement. I've attached a file with the respective distances for the alignment of chains A in 3c4f and 3gqi as an example. 

Not sure if this is a misunderstanding from my side or a bug, yet I'd greatly appreciate a clarification of the differences.