If efficiency is less important than noise, a custom linear regulator might be an option. Efficiency would be very terrible and a lot of heat would be developed, but you would be able to feed your 12V gear from your 24V battery bank. Far from ideal, but would work and be relatively simple to make.
If that's not an option (and if you're using precious, expensive RE then efficiency should be one of the most important things), I would go for a DC/DC PWM converter, either homebuilt or a commercial one. There are ways of making it quiet: shielding, grounding and filtering. Closed metal encasing, filter at the input, more filtering (filters completely shielded too), good bonding and grounding, lots of ferrites in the line, more filtering, feedthrough capacitors to keep things fully shielded when power enters another metal box in which the actual PWM sits. Then feedthrough capacitors to couple the power out of the box, more ferrites, more (shielded) low-pass filters. And a few electrolytics as well.
If you take a commercial PWM converter, build it in another (*completely*!) shielded and grounded metal box (and if you really want to go overboard, in another metal box), 60-70 dB suppression should be easy, 80-90 dB doable, 100-110 dB will be challenging but not impossible (likely involving a box-in-a-box).
There's no reason not to use PWM converters for noise reasons, as noise can be suppressed. The fact that most commercial equipment is lousily suppressed shouldn't deter you from adding your own shielding, IMO.
Not much more options, as I see it.... either tap the 24V battery at 12V, use a linear converter (insane), or use a properly built DC/DC PWM converter. Or get gear that works off 24V (or 28V). BTW, good radio equipment shouldn't be too sensitive to noise on the power lines.... it should have adequate filtering in the powerlines itself already. Noise picked up via antennas is a different matter of course, but that's where the shielding, grounding and filtering of a DC/DC PWM converter comes into play.
That's what I would do.
BTW, I assume most of the current is drawn by the final amplifier; it might be an idea to at least make that work off 24V. Transistors/FETs for those voltages (28V) usually have a little more gain and higher efficiency too than the 12V ones. With the bulk of power coming straight from 24V, the actual receiver/transmitter would probably only draw a few ampères at 12V and could be supplied with a much smaller linear regulator or DC/DC PWM converter (again, adequately shielded).