Several micro Henry is small enough value - they could be even air-core if you want - it depends on task, construction, current, ets.
For that picture (that schematics) - quite often we take large type (2W-5W) resistors (R37, R41), and use them as inductor cores.
You can use some online calculator to calc turns number. You can chose between one layer winding and two (or multylayer) winding.
For example, for 12-14 mm resistor core it could be about 30-40 turns of wire - it's very approximately (I took numbers from 20-years ago memory).
Task is very easy if you have inductance meter (you can buy any cheap chinese one for this purpose), but if you don't have - it would be ok, if you'll have 6-8 uH real inductance (+-10..-+20% have to be ok).
If you dont't have a coil winder (there are reasonably cheap hand-crank one's on eBay), then a cordless drill can be used as a winder. I'd find a suitable cyindrical tube and wind excess turns, you can can measure inductance and remove turns to fine-tune the value that way, and it will be neat.
For high current audio inductors the windings should be supported to prevent vibration, so wind tightly on a former and apply a little superglue or lock the windings in place with epoxy or similar if an open coil. This will reduce the tendancy for them to "sing".
It is very strange the location of such chokes. In any case, which is said below, say, to make them over a resistor is a common task. take a piece of wire 0.5 to 1mm diameter (Enameled) and wire 10-12 turns over the specified resistor ans then solder the extremes to resistor wires.
It also depends on wire diameter (Thinner wire has high inductance per turn than wide ones), coil diameter (Increasing diameter increase inductance non-linealy) turns space (closer has higher inductance than spaced), coil length, number of wire layers and core properties. As a first attempt, try the Wheeler (or Weeler, I am not sure) formula in any Radioamateur's handbook or in the web.
For audio frequencies I'd not be too worried about reducing losses, that's going to get important at higher frequencies where the skin-depth of copper becomes significantly less than the wire diameter. Distortion only happens if the core is magnetic, magnetic materials are non-linear, often strongly so. Air/plastic/ceramic are basically free-space and Maxwell's equations (which are linear) apply directly.
Thiele merely applied the appropriate formula that he found in Wheeler's 1928 paper. Thiele cites Wheeler's paper as his bibliographic reference . Screen capture image from Thiele's AES paper, below.
Besides the wonderful web-based "calculator" linked in post #9 of this thread, there are a bunch of other websites that present and discuss Wheeler's formulas. Here are a couple nice ones:
A) Wheeler (And many other, see the Langford-Smith book) formula is an approximation. There is no exact formula for real world inductors because their inductance is affected by a multiplicity of factors;
Wikipedia comfirms my previous knowledge:
As a practical matter, longer wires have more inductance, and thicker wires have less, analogous to their electrical resistance (although the relationships aren't linear, and are different in kind from the relationships that length and diameter bear to resistance).
B) DC inductors without iron cores or with open magnetic circuit, are usually based in maximum resistance to DC, or sometimes, self rigidity to vibration and movements. In RF large copper diameters are used and in extreme cases, they are silver plated at UHF frequencies to reduce skin depth effects.
Wire diameter is not included in the Wheeler inductance formulae for multi-turn coils; because its effect is extremely small. The formulae are accurate to 1-2% even when neglecting wire diameter entirely. Thus the effects of wire diameter amount to less than 1-2% of the total inductance of a multi-turn coil. Wheeler, an engineer, considered it negligible. And he neglected it.
There are charts available that show wire gauge vs current capacity. Use a gauge that can handle what you need with a little headroom for safety. I usually go up at least one gauge from what is needed.