The CdS LDRs I have used have a resistance of about 50K in the dark but apparently they need to have a higher resistance to work in this circuit unmodified. According to the schematic, it's not a potential divider but apparently lowering R3 (to e.g. 120K) helps compensate in some way I don't understand? I tried desoldering R3 but somehow damaged the board because even though the pads don't appear to be shorted, they're shorted. So I'd rather use appropriate LDRs if possible. In the photograph below, the R3 resistor is the one closest to the orange and black component top right.
