We perceive pitch differently at different loudness levels. That's one of the first problems. That's one of the reasons why we tell people in the studio to remove one ear of their headphones. This way they hear themselves, with bone conduction & the amplified signal in the other ear. And people just like to turn up their headphones way too loud which makes them all sing flat.
Tracking vocals utilizing speaker fold back is also quite doable. But when you have the speaker up to a comfortable level, of course it leaks/bleeds into the solo vocal microphone. This is certainly problematic. The way you correct this is to cut another vocal track the same way, without the vocalist. You must not move the microphone. You must not touch the microphone. It must remain precisely in the same position it was for the vocal track along with the fold back speaker also not being changed at all. You then take both of those tracks and combine them together while taking the second vocal track (without the vocal) and phase inverting that track. When you balance the two tracks carefully together in Mono, the speaker fold back & the acoustics will electrically cancel out. This leaves only the difference between the two intact, which is the vocalist. I've utilize the same technique while cutting brass, strings & woodwinds in this studio. This way, they never have to be over encumbered with headphone nonsense. And they all seem to play much better that way when listening to the blaring rhythm tracks through this studio fold back speakers. So, for instance, if I have 6 string players, we'll make a first pass. For the second pass, I have the string players play musical chairs i.e., I've set up six additional chairs behind the first six chairs. Those two tracks then get combined with one being phase inverted. And voilą! The speakers and the room acoustics actually disappear. It's like magic. It is magic. Audio Magic. Then we'll do another pair just like that and what do you get? A section of 24 string musicians that sound like a real orchestral section, in stereo no less. Repeat the same for the brass. Repeat the same for the woodwinds. Repeat the same for the backup vocalists. And then don't have the solo vocalist repeat anything. I've been utilizing this technique since 1979 quite successfully. I snicker through a lot of this since I'm one of the few people who has done this for so many years. It's actually also a great way with digital recording today (no extra analog tape noise) to deal with crappy small room acoustics. If you don't bump or touch any of the microphones the crappy acoustics will also have a tendency to cancel. But you have to have those fold back speakers on for both passes for this to work. Any initial movement of any microphone will completely zap the timing differential between any microphones that should be canceling.
You can call me Capt. Can Do
Mx. Remy Ann David