Using the concept of "outlaw western" and one of the best western films ever made, Tombstone, I compiled a file that would give that feel to with a little bit of a help from a LoRA.
This was trained using AOM3, specifically, the A1B model and has been fun to test out on different models as well. The weight should sit comfortable between either 0.8 and 0.6 for best results, but can go lower when mixed with other LoRas. The faces get a little funny if it sits at a full 1.0 weight.
Good trigger words to use would be cowboy, boots, wild west, dark atmosphere, fog, nun, cowboy hat, etc.