Heat treating is both a science and an art. The first step is to harden the steel. To harden, one takes the annealed, formed piece and heats it to what is termed the critical point. This is often described as a "cherry red heat" but it differs for various steel compositions. The best way to determine the critical point is to heat the steel until a magnet is no longer attracted to the hot steel. Then, you quickly quench the piece (here we are talking about knife blades) in a liquid.
The best liquid to use for quenching is often a matter of dispute. At one time, sperm whale-oil was highly regarded. I use SAE 30 motor oil. Others use a stiff brine. Basically, the quicker the steel is cooled, the harder it is. Brine is used for files, it produces a super hard steel. Using oil will slow the cooling somewhat and will produce a quenched steel that has a bit of toughness. When quenching, it is important to plunge the knife quickly and straight into the quenching liquid. Hesitation or lack of a straight stab can cause the blade to warp.
Once you have quenched (and hardened) the blade, you are ready to heat treat. There are two ways to do this. One is to carefully heat the hardened blade to the temperature stated on the specification sheet to produce a give Rockwell Hardness. This requires an oven and accurate temperature control. The entire blade will be the same hardness and toughness using this procedure. Most commercial knives are done using this process and it works pretty well.
The method I use is called differential hardening or differential tempering and is done by "running the colors". You don't need an oven or thermometer; I have a coal forge and use that. You first polish the blade to a good shine. Then, you heat a bar of iron in a forge (or charcoal fire) to a red heat. You then place the back of the polished blade on the hot iron and watch the colors develop. What you are looking for is for the first straw color to just reach the cutting edge along the full length of the blade. This takes a bit of practice because you have to move the knife blade around to push the straw color ahead in places and hold it back in others. You do not heat and let the straw color reach the edge and then quench again, you let the colors move slowly across the blade and stop by themselves. Ideally, when you are finished, the bulk of the blade will be blue, about 1/8" of the edge will be straw and the other colors will be compressed between the blue and straw. When done this way, most of the blade is springy; springs are tempered to blue. But, the edge with its straw color is just a bit softened from the hardened state with a bit of toughness. Steel with 1% carbon heated to a faint straw will hold a fine edge for a long time.
I probably used up a dozen forged knife blades before I got adequately good at differential hardening. I still screw up on about one out of five blades. That is one reason I like carbon steel like 1095 or good saw steel. Unless you get it hot enough to burn the carbon out you can re-harden and try again. I would sure hate to have to do a sword this way though.
May I walk in beauty.