Fourteen nights on Synaptolepis kirkii, the woody South African shrub sold as African dream herb, taken on an empty stomach in the morning, followed by a seven-day washout.
By the seven-night mark the effect was subtle but real, similar in size to Silene capensis. I went from occasional dreams and patchy recall to having dreams in roughly every sleep segment and remembering them with reasonable detail more than half the time. Vividness felt qualitatively stronger than Silene, maybe by a quarter, but no lucid dreams turned up despite some rough attempts.
One honest confound: the protocol only fixes the morning empty-stomach dose and does not rule out other consumption later in the day, so the picture is not perfectly clean.
By day 14 the conclusions held. Sleep quality was marginally better than baseline, vividness and intensity were noticeable but subtle, and there was still no lucidity. The most interesting part came in the washout: the vividness and intensity carried on through day 7 off the compound, which hints at a slow-clearing or build-up effect worth a closer look.