I don't really forge the chisels, which would be a high heat setup with significant working. Without the hammering in forging in that type of setup, the grain size in anything would grow large (if one were just to take bar stock and heat it to yellow heat and leave it at that to soak). The whole idea of temperature control in normalizing before heat treatment is to heat the metal but not high enough that grains enlarge. Temperature for this varies by starting grain size, steel type, etc, and the soak times are usually in an hour to hours.
Moderately overheating something that's already fine and heating it (if it's quick) doesn't usually lead to that much damage if the heating isn't done repeatedly (the grain size doesn't increase that much).
So, anyway, this is the same file after the large grain picture from triple heating just heated to somewhere in the ballpark of 1700 (which would be enough for the grain size to grow) and then hammered to about half thickness. This is what I do to shape my chisels and give myself a break hand finishing them. The objective after this, I guess, is not to allow the steel to sit in the forge too long at too high of a temperature.
Sorry, the break is curved, so you can't see too much (the focused area is right down the center). It's not as fine as the first picture (which is a bit puzzling -that was just quick heat of a file that had already been heated several times before), but it does look like some improvement over the triple heat picture. I'll have to repeat this another time or two with another file and see what I can come up with. This broke off much more easily than the large grain picture (which is puzzling, but these haven't been tempered, so it may not amount to much and it could just be the odd shape from hammer marks made the back side unsupported and easy to break).
(and a repost of the triple heated picture for comparison - again, the picture above was taken on the same file, same general spot *after* the piece below was broken off prior. the forged bit is only about 3/16" further down the file length than the prior piece.