Culling cards presents several problems beyond mastering the physical actions of the cull. The main one is contextual. You might have the world’s most technically invisible cull, but you still have to look through the cards. As Erdnase might have put it, the cull has yet to be invented that can be done without looking through the deck or, at least, spending too long so doing; although solutions (shuffle-culling, making use of edge marks, riffling the indices, and jogging cards) have been attempted with varying degrees of deceptivenesss. Indeed, The Expert at the Card Table (1902) describes two such “ordinary methods” for locating and securing (culling), but once again the matter of context comes up.
“If dalliance with the deck is allowed” (Erdnase again), as may happen impromptu, then any of the above may work. Open Culling—sorting cards under the guise of another effect, à la Victor Farelli et al.—is an excellent method. As Harry Lorayne was wont to say, if you can’t do something secretly, do it openly. The equivalent of Erdnase’s locating and securing on the discard is yet another approach. Thus we can cull target cards while openly removing others at the start of an effect, or while gathering cards up at the end of one. Erdnase’s concept of spotting top cards and culling cards of like value (so as to cut down on the number of cull actions) is worth noting.