DSSP does not match PDB header

Question

[lorenzo.fabbri]

Hi all. Looking at the DSSP of a protein, I noticed that it does not match the information in the header of the PDB file. I explain better: if a helix in the header of the PDB file starts at 30 and ends at 40, the same helix in the DSSP file starts, say, at 29 and ends at 41. I saw that basically every tool that does "secondary structure extraction" like ESPript or even Jalview do use DSSP information. Do I have to manually check the PDB header or can I "trust" DSSP? Thank you.