Understand that there are two noise sources in resistors:
1. Johnson noise- this depends on the resistor value and has nothing to do with how the resistor is made. It is a fundamental property of the universe. You can only reduce it by lowering the temperature.
2. Excess noise- this does depend on resistor construction. Wirewound are lowest, metal film slightly higher, carbon composition much higher. For most values of resistance, the Johnson noise will swamp excess noise for metal film and wirewound.
Your noise will hardly be affected if you switch between a 4 cent metal film and a $40 bulk foil (or whatever fancy resistor suits you). Best strategy is to buy a name-brand (e.g., Vishay, Ohmite) metal film and use a higher wattage than necessary to avoid heating. If matching and long term stability are important (e.g., equalization resistors, plate loads for LTP), your resistors can be pre-baked.
I've used Vishay dale (mexico plant) Vishay (Thailand plant)multicomp (Thailand and india source) Welwyn and International (chinese source) metal film resistors. They are a lot quieter than the 1961 to 1970 age carbon comp resistors I have been replacing. All have had accurate resistance. I don't know if any brand is quieter. I don't really notice noise differences below about 100kohms, but there and above it is quite obvious. I use a lot above 100k on the tube amps in the organs. High resistance carbon comp resistors tend to drift up in value.
Sure, just stick them in an oven- for most resistors, 110 degrees C will be fine. Measure the resistances, bake for several hours, then after they cool, recheck the resistances. This is useful if you need a long term match.