I'd suggest a search.
You will get many answers.
Whilst some have reckoned isolators are not needed (eg hereon: Haemphyst, JMelton86), I see it as I do not want one faulty battery wrecking the other. (ie, I want maximum battery life.)
After that comes things like battery independence - ie, having cranking power irrespective of the state of the "other" battery(s).
[ And even if only a question of battery life, I could not be bothered with expensive and subjective monitoring when - in my case - a simple relay prevents all that crap! ]
For those with charge lights - ie, the alternator has a D+ else "L" or similar output for a dash charge lamp, else perhaps another equivalent "not-charging" lamp - I propose a mere relay sized for the current you require. That might be a $5 30A relay or a $25 250A relay - both far cheaper than (IMO inferior and less reliable) voltage sensing isolators.
If you do not have a charge lamp - ie, you have a stator type alternator - then you have little choice but a voltage sensing isolator or - better still - a current sensing isolator (but AFAIK they do not exist commercially).
And in either case, for high currents, you may need a separate high-current relay. EG - an 80A "smart" isolator or 30A "charge light" relay powers a 200A or 500A or 1kA relay.
Such isolators overcome the need to match battery types and conditions (age, temperature etc) and their connections. (IE - for maximum life, hard-paralleled batteries are the same type, batch, age, history, temperature and are connected "cross-diagonally" to ensure charge & discharge symmetry.)
If those limitations and complications aren't enough, then just realise that two parallel
identical batteries will have double the failure rate of the single battery (where failure rate pertains to self-failure - eg, manufacturing defects and tolerances, history etc). And if one is in the boot/trunk and one in the engine bay.... ha!
Anyhow, isolation overcomes that. You can then parallel any battery (wet, AGM, old, new) whilst being charged without any concerns "
out of the ordinary". (FYI - and parallel them whilst being discharged...)
Irrespective of isolation, the inter-battery connection needs to be fused at each end as close as practicable to each battery. (Common "fusing" sense prevails - the fusing is to protect the inter-cable; if batteries are adjacent, either or both fuses may be unnecessary.)
The fuses and cable size (and isolator rating) need to handle the "end" load plus the battery recharge current, but the fuses must protect the cable and (if fitted, the) isolator.
I prefer self-resetting circuit breakers, but that is where my battery recharge current can be much higher than my typical load (eg, a 5-10A load, but recharge commonly 50A or higher using 50A breakers).
And gone are the days of glass fuses.
If you search you should my other posts that outline keeping normal cranking batteries for the engine and near-amp batteries (AGM else in sealed vented enclosure), usually with a big AGM and heavy "local" wiring, else maybe heavier wiring from "the front" with a smaller local AGM as a cheaper and superior alternative to
stiffening capacitors.
But the interconnection must handle at least the
average power (current) draw of the amp. If exceeded (eg, 110% for 1 hour etc), the fuses will blow else cable & relay melt.
The battery recharge current is additional, but with the amp lower than the max "allowable" draw and a copious alternator, the remote batteries will eventually fully recharge given enough time (eg, alternator at 13.8 to 14.4V with negligible interconnection voltage drop).
As to actual sizing.... nah - search my recents - especially perhaps rfhvhtoo's
relay or isolator as an example of wasted time, or his
charging agm battery] as an example of what NOT to buy. As I recall, I comment on the incredulous specs given by certain amp manufacturers (like over 100% efficiency, and output powers that vary (linearly!) with input voltage) though I suppose the later will kill any suggestion for the need for stiffening cap and big batteries!!!
(Those advertising "tricks" are IMO brainless. Check the fuses and some basic research. Not that I know of your systems specs, and RMS verses "other" ratings. The obvious misprints I see - like RMS power is 0.7 times peak power - it's 0.5 x peak assuming peak is traditional "Music" peak - ie, peak magnitude.)
BTW - solid grounding. Unfused. Grounding should at least equal "hot" power cables. Extra grounding cable (gauge) is usually the cheapest way of reducing distribution resistance (and noise) especially when the chassis/body is used - ie, doubling 1m of grounding wire may have the same effect as doubling 5m of hot +12V wire.
As to the audio connections, I'll leave you to the experts hereon.
They can also give real figures and suggestions to my generalised power & battery approach.