Try our new info resource - "Aladdin's Cave" (Main menu)Just added a separate link to Ash's Dropbox thread (shortcut)
Some thoughts on syncing that you could consider as it's one of sequentially imposed conditions that I believe can lead the wrong way. The cylinder is drawing vacuum against the carburettor slide as restriction/resistance, this is indicated to you on the dial as to how much the piston going down the bore can lower the pressure below atmospheric when measured against a fixed constant (probably a spring and plunger in the gauge) and so give you a number to aim for and adjust the slides to give parity across the cylinders. BUT, you've got one cylinder that aggregated measures lower than the other 3 because it doesn't seal correctly. By extension, that cylinder will most likely impart LESS vacuum to be measured through it's stroke as it has effectively less restrictions placed on it (carburettor slide as restriction/resistance PLUS it's leaking) which in pure vacuum measurements forces you to lower the slide to observe parity on the gauge. Which of course is correct for absolute vacuum numbers, but not for combustion. It effectively causes you to set that throttle slide lower than the others to get the readings correct.This is wrong for the combustion equalization of the cylinders (some of the rattling is because each cylinder doesn't input the same energy into the crankshaft so that it's rotation is more staccato) what it needs is the throttle slide to be more open than the others to make that cylinder equal the other three. Worse than that, the error as i'd view it is doubled. The cylinder with low compression already inputs less energy into the crankshaft, setting the throttle slide lower (to achieve parity of vacuum) also adds to that error making the idle more erratic, not less. To compensate you'd need to set the other three to parity on the gauges, and LESS (this is what modern sequentially fired ecu injection systems do to maintain smooth idle) vacuum on this compromised cylinder, which should improve idle smoothness, but could compromise higher rpm performance. An aggregated carb setting for synchronisation would be accurate bench synchronisation and not use the gauges so as to avoid the prompted wrong settings of absolute vacuum readings. It wouldn't fully compensate at idle, but would leave more accurate higher rpm performance alone.
Going on to the noises made while idling. When new, the primary drive chain has fairly limited slack which keeps the driven components (clutch and gearbox internals) at more or less the same speed. They are however a "remote" flywheel as they impart their own own constraint on speeding up the crankshaft when unloaded (idling and not driving the bike) and by nature fairly unsophisticated for noise insulation. If the crank speed oscillation frequency arrives through a worn primary chain then these oscillations effectively keep transferring their difference in inertia by variously putting the slack in the opposing chain runs to keep then exiting the whole system which we hear obviously as unwanted noise. If the engine running doesn't deliver equal pulses for each firing stroke it simply gets heard as more noise as it ricochets through the mechanical system.As you raise crankshaft speed toward 1500rpm the primary chain centrifugal forces (effectively tensioning the chain) overtake the inertia from firing pulses, in addition the the firing pulses frequency rising above a point at which the flywheel natural frequency is exceeded cause the whole system to go quite. None of the faults disappear but just become harmonious.
Personally I find it hard to believe you'd have to replace the chain with 17.7k miles total.