The LIGO and Virgo detectors have repeatedly observed binary black hole mergers involving components heavier than 30 solar masses, a finding that startled astrophysicists because stellar evolution theory, particularly the pair-instability supernova process, was expected to limit black hole masses to lower values. An alternative proposal from Broadhurst, Diego, and Smoot (BDS) suggested these high masses could be an observational artifact: gravitational lensing magnifies signals from intrinsically lighter black holes at higher redshifts, making them appear more massive and nearer than they truly are.

A preprint posted to arXiv on 15 April 2026 (not yet peer-reviewed) by Ritesh Harshe and collaborators puts this idea to a comprehensive test using Monte Carlo simulations of thousands of lensed binary black hole (BBH) systems. The authors adopted the specific intrinsic mass and redshift distributions advocated by BDS, layered on realistic lensing optical depth derived from galaxy mass functions, applied random magnifications, and then filtered the simulated events through a model of LIGO/Virgo sensitivity across observing runs O1–O3. They evaluated consistency against four independent observational constraints: (1) the total number of confidently detected BBH events (roughly 80–90 in current catalogs), (2) the two-dimensional distribution of redshifted total mass versus apparent luminosity distance, (3) the complete absence of any events showing strong-lensing signatures such as repeated waveforms with matching time delays and amplitude ratios, and (4) upper limits on the stochastic gravitational-wave background that would arise from a vastly larger population of unresolved high-redshift mergers.

No choice of model parameters allowed the lensed population to simultaneously reproduce all four datasets. Matching the high-mass tail required either too many total detections, an undetected stochastic background exceeding current limits, or an implausibly high fraction of strongly lensed events that should have been identifiable. The study’s limitations include dependence on specific assumptions about lens galaxy populations and the exact shape of the intrinsic mass function; the real observational sample, while the largest to date, remains modest in size and is subject to selection biases. Even so, the multi-dimensional inconsistency is robust.

This preprint corrects an important oversight in the original BDS coverage and in much subsequent popular reporting: those accounts focused narrowly on reproducing the marginal mass distribution while ignoring global rate statistics and background constraints. By synthesizing the new simulations with the LIGO-Virgo-KAGRA population paper (Abbott et al. 2023, arXiv:2111.03634, which used hierarchical Bayesian inference on 76 BBH events to favor a broken power-law mass spectrum extending above 30 solar masses with no sharp cutoff) and with dedicated strong-lensing searches (Abbott et al. 2021, arXiv:2105.06360, reporting no confirmed lensed candidates in O1–O3), a clearer picture emerges.

If lensing had explained the high-mass events, the underlying stellar-evolution models would have required wholesale revision because the intrinsic population would resemble the familiar ~10-solar-mass galactic black holes; merger-rate predictions would shift dramatically toward higher redshifts, and cosmological parameter estimation (standard-siren measurements of the Hubble constant from GW catalogs) would need large magnification-bias corrections. Because lensing is ruled out, confidence increases that nature produces black holes in the 30–60 solar-mass range, likely from metal-poor massive stars or dynamical channels in dense clusters. The pair-instability mass gap appears narrower or more permeable than many models assumed.

This result connects to a broader pattern visible since GW150914: each new population surprise has ultimately expanded astrophysical understanding rather than exposed hidden systematics. Future O4 and O5 runs, with expected hundreds more events, will further tighten these constraints. The non-detection of a stochastic background is especially diagnostic; it functions as a powerful null result against any scenario invoking a large hidden high-redshift population.

Far from a niche debate, the rejection of lensing therefore anchors both stellar astrophysics and gravitational-wave cosmology on firmer empirical ground while highlighting where theoretical work on massive-star winds, binary interactions, and pair-instability physics must improve.