You have a 4 dimensional soundspace to work with.
Left-right placement, using pan/balance, M/S, and phase/timing (experimental).
Top to bottom, frequency dependant, this is (arguably) the most sure-fire way to separate instruments.
Front to back, using reverberation and delay, with a side order of EQ.
Timing...not, or deliberately overlapping parts that compete in other areas listed above.
So...you could carve out a space for vocals, or the reverse approach, fitting all else around them, by centering them LR, (common) or panning them to one side (less common), adjusting vocal EQ or other parts EQ or a combi of both, moving them front (less delay or reverb, common) or back (more delay or verb, rare), or by arrangement, backing off or eliminating other parts during different parts of the piece.
There is another variable related to front/back placement, namely amplitude, where you have two possibilities, fader adjustment, and compression.
If I were asking myself this question, I'd start by isolating ALL
the effects described above, one at a time, to see what my tradeoffs were, then experiment with combining two at once, till I had a solid understanding of how they inter-relate, then identify specific problems in specific songs, and use single or multiple techniques to resolve them.
One very common approach is to use some EQ, panning, and reverb to separate guitars and certain keyboard patches (including orchestral backing) from your vocals, reverb and EQ to separate drums from vocals since both are usually centered left to right, and frequency separation between bass and vocals, where you have to be careful since some of the bass definition comes from higher resonances in the vocal range.
Beyond this, there are infinite possibilities, which form a large part of the art of mixing. Experiment and define your own priorities once you have the basics in hand, and use
reference material, songs you like that are recorded by others, if you value potential commercial success.