Yeah, I have no intuition with this. The telescope they used was 8.2m and they used 300-400 second exposures (with spectral filters). I guess it comes down to

Yeah, I have no intuition with this. The telescope they used was 8.2m and they used 300-400 second exposures (with spectral filters). I guess it comes down to space being really really dark and cyrogenically-cooled CCDs having pretty good noise floors. The authors had this to say about it: Surprisingly, like 2012 VP113, 2015 TG387 is relatively bright compared to the limiting magnitude of most of the survey fields. The Subaru observations generally go deeper than 25.5 mags, making 2015 TG387 about 1.5 mags brighter than the limiting magnitude of the disocovery survey fields it was found in.