Do learners of a second language (L2) need frequent contact with native speakers of that language in order to master its phonology? What if they hear audio recordings of native speakers and receive immediate corrective feedback about their perception? We used a randomized controlled experiment with 135 Chinese speakers (with English as their L2) to examine whether a high-variability perceptual training (HVPT) paradigm might enhance the perception of challenging contrasts between English voiced and voiceless stop consonants. Learners in all the age groups tested -- middle childhood, early adolescence, and young adulthood -- showed enhanced perception of English stop consonants after 20 five-minute training sessions conducted across 4 to 6 weeks, based on audio-recorded input coupled with corrective feedback. The training benefits were maintained at the one-month follow-up. Our results suggest that HVPT using audio-recordings of native speakers can be an affordable and useful language enrichment to supplement live interaction with native speakers, for L2 learners of a wide age range.