{"id":"8fcef294-4a69-4e87-b7f7-09de67cb8b1f","arxiv_id":"gr-qc/0604032","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Amplitude-corrected post-Newtonian inspiral waveforms yield signal-to-noise ratios up to 25% lower than restricted waveforms, implying detection rates up to a factor of two lower than earlier simulations suggested.","lead":"This paper compares two ways of modeling the gravitational waves from merging black holes and neutron stars and works out how the expected detection rates change. It finds that using the more complete wave models would lower expected signal strengths enough to cut predicted detection rates by up to a factor of two.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The robustness claim for the up-to-25% SNR reduction hinges on Eq. (3.6), asserted without proof, and on the deferred proof of Eq. (3.1); if those fail, only unexplained numerical tables remain.","rationale":"The reader's weakest_assumption identifies the same load-bearing spot. Eq. (3.1) is equivalent to ||h|| < ||h0||; the paper's analytic route to this is through the sign-definite cross terms (3.6) and a deferred proof in [10]. Section III B states (3.6) without derivation, and Section IV explicitly sends the direct proof to [10]; these are manuscript-internal acknowledgments of the omission. Without (3.6) or [10], the claim that the result is independent of detection issues rests on unexamined algebra, and the 'up to 25%' generality is an interpolation from Table I and Fig. 1. The concern is not an external disagreement with consensus; it is an internally missing step in the argument. The numerical evidence is direct and reproducible in principle, so the appropriate status is conditional rather than rejection: if the inequality is verified or the companion proof appears, the central claim stands; if not, it is an unexplained numerical trend. Hence verdict_should_be is UNCHANGED relative to the reader's CONDITIONAL assessment.","tokens_in":7890,"tokens_out":12142,"duration_ms":118208,"concrete_test":"Using the explicit 2.5PN amplitude-coefficient expressions from Blanchet, Iyer, Will & Wiseman (1996) and Arun et al. (2004) as used in Eqs. (2.1)-(2.5), form C_(m/2)(theta,phi,psi,iota,eta) and evaluate Re[C*_(0)C_(m/2)] on a dense grid over iota in [0,pi], eta in [0.01,0.25], and all sign combinations of the beam-pattern factors F_+,F_x (or over theta,phi,psi with the Initial LIGO response). If any m=1,...,5 gives a positive value, Eq. (3.6) is false and the analytic basis for Eq. (3.1) fails. If the inequality survives, independently recompute the NS-BH row of Table I and the delta/gamma contours of Fig. 1; a match would confirm the numerical phenomenon even though the full proof remains deferred.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central inequality, Eq. (3.1), is presented as robust and independent of detection caveats, with analytic support in Section III B. The key step is Eq. (3.6): Re[C*_(0)C_(m/2)] <= 0 for m=1,...,5, stated as 'not difficult to show' but not demonstrated. The proof of the more relevant Eq. (3.1) is explicitly deferred to reference [10] ('The direct proof ... will be given in [10]', Section III B; also Section IV). If (3.6) or the deferred proof fails, the analytic explanation and the paper's claim that no further caveats are needed collapse; the numerical Table I and Fig. 1 would still show a trend, but without an understood cause and without support for the 'up to 25%' generality. The numerical computations may well be correct; the concern is that the central argument for robustness is currently unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compares signal-to-noise ratios (SNRs) for quasi-circular compact-binary inspirals computed with restricted post-Newtonian (PN) waveforms and with amplitude-corrected (2.5,3.5)PN waveforms. It defines the faithfulness SNR rho_h1[h2] and numerically computes, for Initial LIGO noise, three quantities: rho_h0[h0] (restricted template and signal), rho_h0[h] (restricted template, amplitude-corrected signal), and rho_h[h] (amplitude-corrected template and signal). Table I and Fig. 1 show that rho_h0[h] can be 20-30% below rho_h0[h0] for asymmetric systems, while rho_h[h] is only slightly above rho_h0[h], leading to the central inequality rho_h[h] < rho_h0[h0] with differences up to about 25% for total masses below ~30 solar masses. Since detection rates scale as the cube of SNR, the paper concludes that rate estimates based on restricted waveforms may be too high by up to a factor of two. The analytic explanation in Sec. III B uses the stationary-phase approximation and the coefficient inequality Re[C*_0 C_(m/2)] <= 0, stated without proof, while the direct proof of the central inequality (3.1) is deferred to a forthcoming paper [10].","tokens_in":8064,"tokens_out":5145,"duration_ms":50927,"significance":"If the central claim holds, it has direct practical consequences: simulated searches and detection-rate projections based on restricted waveforms would systematically overestimate expected SNRs, and the effect is large enough to change rate predictions by a factor of two. The paper's numerical construction is transparent and reproducible in structure: the SNR ratios are computed directly from published post-Newtonian waveform expressions with no parameters fitted to the outcome, and the main numerical statement is falsifiable. The paper also makes the useful observation that amplitude corrections to the dominant second harmonic alone reproduce the behavior of the full amplitude-corrected waveform for restricted-template searches (Table II). The main weakness is that the analytic robustness argument, which is used to elevate the numerical trend into a parameter-independent claim, is not actually demonstrated in the manuscript.","major_comments":[{"comment":"The inequality Re[C*_(0)C_(m/2)] <= 0 for m = 1,...,5 is stated as 'not difficult to show' but no proof is provided. This inequality is the key step in the analytic explanation of Eq. (3.2) and in the claim that the result is insensitive to detection details. Since the coefficients C_(m/2) are complex functions of the angles and mass ratio and are not written out in this paper, the assertion that the inequality holds 'irrespective of parameter values' is not verifiable from the text. Please supply a full derivation, or at least an explicit listing of the coefficients together with the signs of their real parts, and justify the statements that the m = 1/2 and m = 3/2 terms vanish identically while the others are negative definite.","section":"Sec. III B, Eq. (3.6)"},{"comment":"The proof of the central inequality rho_h[h] < rho_h0[h0], Eq. (3.1), is explicitly deferred to reference [10], which is listed as 'in preparation'. This inequality is the basis of the paper's headline conclusion and of the statement that 'no caveats are needed here' beyond the difference between a genuine signal and a (2.5,3.5)PN waveform. Deferring the proof of the main result to an unpublished manuscript leaves the central claim unsupported within the paper. The numerical tables and figures provide empirical evidence, but the claimed robustness depends on an analytic argument that is not present. The proof, or a detailed and checkable sketch of it, should be included in the paper, or the conclusion should be explicitly conditioned on the validity of the unpublished proof.","section":"Sec. III B and Sec. IV"},{"comment":"The comparison in Eq. (1.3) and the related discussion hold for fixed intrinsic parameters, while real searches maximize the SNR over intrinsic parameters as well as extrinsic ones. The paper acknowledges this caveat for the restricted-template case, but then presents the rho_h[h] < rho_h0[h0] result as independent of 'any other detection issues' without discussing how parameter maximization affects it. Since the proof of (3.1) is not given, the reader cannot check whether the inequality survives maximization over the template-bank parameters. Please clarify how maximization over intrinsic parameters enters the main claim, or restrict the statement to the fixed-parameter faithfulness comparison that is actually computed.","section":"Sec. I and Sec. IV"}],"minor_comments":[{"comment":"The 'up to 25%' statement is obtained by combining the left and right panels of Fig. 1, which display delta and gamma separately. Please state explicitly that the same points in (eta, M_c) parameter space are used for both quantities, or add a contour plot of the combined ratio (1+gamma)/(1+delta) so that the headline number can be checked directly.","section":"Sec. III A"},{"comment":"The stationary-phase expressions are written 'leaving out the step function in frequency', but the inner product integral extends to 7 f_LSO and the higher harmonics are cut off. Please specify how the cutoffs are imposed in the analytic integrand that underlies Eq. (3.5), since the step functions affect the integration limits and hence the sign of the comparison (3.7).","section":"Sec. II, Eq. (2.5) and Eq. (3.4)"},{"comment":"In the discussion of the right panel of Fig. 1, the phrase 'astrophysically less interesting' is used for large-M_c and small-eta systems. This is a subjective characterization; please give a quantitative selection criterion for the mass and mass-ratio range considered astrophysically relevant.","section":"Sec. III A"},{"comment":"Reference [10] is described as 'in preparation'. If it remains unpublished at the time of publication, the manuscript should state that the proof is unavailable and should not rely on it for a central claim. Consider citing a publicly available preprint or including the proof in an appendix.","section":"References"},{"comment":"Several inequalities are typeset with what appears to be 'greaterorsimilar' symbols; please use standard notation (for example, rho_h[h] >= rho_h0[h] with a qualification of the typical size of the difference) so that the quantitative comparisons are unambiguous.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The numerical content appears sound and the paper is within the scope of the journal. The main issue is structural: the paper's most robust claim rests on an inequality stated without proof (Eq. 3.6) and on a proof deferred to an unpublished manuscript (Eq. 3.1). If the author can provide those proofs or suitably restrict the claims, the paper would be a useful contribution. The current version, however, asks the reader to accept the central robustness result on faith."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Chris — quick take.\n\nThe genuinely new thing here is showing that the 0.5PN SNR boost seen by Sintes and Vecchio reverses by 2.5PN order, so the best available amplitude-corrected waveforms give lower SNR than restricted ones — up to 25% lower, hence roughly a factor of two in detection rates. That is a concrete, useful correction for anyone planning LIGO injections or interpreting rate upper limits. The numerical work is honest and reproducible: the waveforms come from published PN expressions, the SPA treatment is straightforward, and there are no fitted parameters. The mass dependence plots are informative.\n\nThe soft spots are exactly where the reader flagged them. The analytic argument for the central inequality (3.1) is deferred to a companion paper, and the key step (3.6) is asserted without proof. On its own the paper is therefore a numerical observation with a plausibility argument, not a fully supported analytical result. The author admits as much by pointing to [10]. The faithfulness comparison in Eq. (1.3) ignores maximization over intrinsic parameters; the author notes this, so it's a stated limitation, but it means the 'up to 25%' for real searches is not yet established. That said, the numerical tables stand on their own and the trend is consistent across mass ratios. Nothing here looks circular or fitted.\n\nBottom line: this is a good, important-in-subfield paper that deserves refereeing. The referee should push for the proof of (3.6) or a clear statement that it is elementary and supplied in the follow-up, plus an effectualness study for the (1.3) comparison. I'd accept it as a solid contribution to the search-template literature, with the caveats kept explicit.","headline":"New numerical result on amplitude-corrected templates that likely cuts predicted LIGO rates by up to 2x, but the analytic proof is deferred and the robustness claim overreaches.","tokens_in":8574,"tokens_out":2033,"would_cite":true,"duration_ms":18608,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["04.25.Nx","04.30.-w","04.80.Nn","95.55.Ym"],"model":"deepseek-v4-flash","headline":"The paper claims that replacing restricted post-Newtonian waveforms with amplitude-corrected ones as search templates lowers expected inspiral signal-to-noise ratios by up to 25 percent, which translates into detection rates up to a…","keywords":["gravitational waves","inspiral","post-Newtonian waveforms","amplitude corrections","signal-to-noise ratio","matched filtering","detection rates","binary black holes"],"falsifier":"One could settle the claim by recomputing $\\rho_h[h]$ and $\\rho_{h_0}[h_0]$ with an independent code that uses direct fast Fourier transforms of time-domain (2.5,3.5)PN waveforms, rather than the stationary phase approximation, over a grid of masses, mass ratios, and inclination angles in the same noise curve, and checking whether the fractional deficit $\\delta = (\\rho_{h_0}[h_0]-\\rho_h[h])/\\rho_h[h]$ reaches 20 to 30 percent in the predicted asymmetric-mass region. If the deficit is absent, changes sign, or only appears under SPA, the central claim is refuted; separately, evaluating the coefficients $\\mathrm{Re}[C^*_{(0)}C_{(m/2)}]$ symbolically from the published amplitudes would falsify the analytic mechanism if any is positive for allowed parameters.","tokens_in":7648,"feed_emoji":"📉","tokens_out":11606,"duration_ms":92221,"temperature":0.7,"pith_summary":"This paper claims that gravitational-wave searches for inspiraling black-hole binaries have been overestimating the signal-to-noise ratios they can expect, because the waveforms used as templates and as simulated signals are usually 'restricted' post-Newtonian waveforms that keep only the dominant harmonic and drop amplitude corrections. Comparing matched-filter SNRs for the best available amplitude-corrected waveforms against restricted ones, the paper finds the amplitude-corrected SNR is up to 25 percent lower for astrophysically relevant binaries in a first-generation ground-based detector noise curve. Since detection rates scale as the cube of SNR, rate predictions are correspondingly too high by up to a factor of two. The paper traces the effect to amplitude corrections of the dominant second harmonic, which systematically reduce the overlap integral rather than adding useful signal power.","feed_headline":"Full amplitude-corrected templates cut inspiral SNRs by up to 25%","feed_subtitle":"Detection rates scale with the cube of SNR, so predicted rates could drop by a factor of two.","key_machinery":"The load-bearing object is the normalized matched-filter SNR $\\rho_{h_1}[h_2] = (h_1|h_2)/\\sqrt{(h_1|h_1)}$, with $(x|y)$ the noise-weighted inner product over frequency. The full signal model $h$ is a sum of seven harmonics of the orbital phase, built in the stationary phase approximation from post-Newtonian amplitude coefficients; the restricted model $h_0$ is the single second harmonic at leading amplitude. The analytic mechanism is the expansion of the cross-spectrum for the dominant harmonic, $\\mathrm{Re}[\\tilde h_0^* \\tilde h^{(2)}] = C^2 f^{-7/3}\\sum_{m=0}^{5}\\mathrm{Re}[C^*_{(0)}C_{(m/2)}](2\\pi M f)^{m/3}$, combined with the paper's assertion that each correction coefficient satisfies $\\mathrm{Re}[C^*_{(0)}C_{(m/2)}]\\le 0$ for $m=1,\\ldots,5$. That makes the overlap of the restricted template with the amplitude-corrected dominant harmonic strictly smaller than the self-overlap, which is why the SNR deficit is systematic rather than a random fluctuation of waveform differences.","core_discovery":"The paper's central claim is the ordering $\\rho_h[h] < \\rho_{h_0}[h_0]$: when both the signal and the template are the best available amplitude-corrected post-Newtonian waveform $h$ (amplitude order 2.5PN, phase order 3.5PN), the matched-filter SNR is smaller than when both are the restricted waveform $h_0$, by up to 25 percent for binaries with total mass below about $30\\,M_\\odot$. Because detection rate scales as $\\rho^3$, this means rate estimates from restricted-waveform simulations are high by up to a factor of two. The paper also finds that using restricted templates against amplitude-corrected signals gives an even larger deficit of up to about 30 percent, but that comparison carries caveats about maximization over intrinsic parameters; the main inequality is presented as free of those caveats. Importantly, switching to amplitude-corrected templates recovers only a few percent of the lost SNR, so the deficit is not compensated by better template models.","pith_inferences":["The paper leaves implicit that detector search summaries and sensitivity statements should quote benchmarks for both restricted and amplitude-corrected waveforms, since the choice changes predicted rates by up to a factor of two.","A testable extension is to use an independent waveform family with amplitude corrections: the same mechanism predicts the largest deficits for asymmetric-mass, higher-total-mass binaries, so a separate implementation reproducing a deficit above 20 percent there would strengthen the conclusion.","Mapping the fractional deficit onto proposed next-generation detector noise curves would be a direct extension, since the frequency weighting in the SNR integral will shift and the up-to-25 percent figure is not expected to be universal."],"forward_implications":["A search that uses the best available amplitude-corrected (2.5,3.5)PN waveforms for both templates and simulated signals will see SNRs up to 25 percent lower than restricted-waveform simulations for sources with total mass below about 30 solar masses.","Because detection rates scale as $\\rho^3$, the rate predictions from restricted-waveform simulations are too high by up to a factor of two.","Switching from restricted to amplitude-corrected templates recovers only a few percent of SNR over most of parameter space, so the deficit is not a template-choice artifact.","For the restricted-template case, the paper finds an even larger SNR overestimate, up to about 30 percent, but subject to caveats about maximizing over intrinsic parameters; the main inequality is claimed to be robust to those caveats.","Amplitude corrections to the dominant second harmonic alone reproduce the behavior of the full multi-harmonic waveform for restricted-template overlaps, so higher harmonics contribute little to this effect."],"supporting_citations":[{"why":"Supplies the explicit post-Newtonian amplitude coefficients and harmonic decomposition from which the amplitude-corrected waveforms are built.","marker":"[2]"},{"why":"Supplies the 3.5PN phasing and energy flux used in the stationary-phase frequency sweep and phase.","marker":"[3]"},{"why":"Supplies the detector noise spectral density used in all SNR integrals and the faithfulness/effectualness distinction that frames the caveats.","marker":"[4]"},{"why":"Introduces the number of useful cycles and supports the choice of frequency band for SNR estimates.","marker":"[7]"},{"why":"Defines effectualness versus faithfulness, the distinction that separates the robust main inequality from the more caveated restricted-template comparison.","marker":"[8]"},{"why":"Shows that stationary-phase and fast-Fourier-transform SNR computations agree to a few percent for the relevant mass range, justifying the SPA approach.","marker":"[9]"},{"why":"Contains the deferred detailed analytic proof of the main inequality $\\rho_h[h] < \\rho_{h_0}[h_0]$, on which the robustness claim directly leans.","marker":"[10]"},{"why":"Documents the earlier observation that SNR increases at 0.5PN amplitude order, which the paper reproduces and then shows reverses at higher order.","marker":"[11]"}],"fun_headline_variants":["Full amplitude-corrected templates reduce inspiral SNR by up to 25%","Restricted waveforms inflate binary black hole detection rates by 2x","Amplitude-corrected waveforms halve predicted merger rates","Inspiral search SNR drops up to 25% with complete templates","Amplitude-corrected templates cut predicted detection rates twofold"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the asserted but not demonstrated inequality $\\mathrm{Re}[C^*_{(0)}C_{(m/2)}]\\le 0$ for $m=1,\\ldots,5$ on the amplitude-correction coefficients of the dominant harmonic, together with the deferred proof of the main inequality; if any of those coefficients is positive for some masses or orientations, the analytic explanation collapses, even though the numerical trend might remain.","fun_headline_variants_meta":{"raw":{"variants":["Full amplitude-corrected templates reduce inspiral SNR by up to 25%","Restricted waveforms inflate binary black hole detection rates by 2x","Amplitude-corrected waveforms halve predicted merger rates","Inspiral search SNR drops up to 25% with complete templates","Amplitude-corrected templates cut predicted detection rates twofold"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000843,"raw_usage":{"total_tokens":3612,"prompt_tokens":827,"completion_tokens":2785,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":443,"completion_tokens_details":{"reasoning_tokens":2692}},"tokens_in":443,"tokens_out":2785,"duration_ms":20090,"temperature":1.0,"reasoning_tokens":2692,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-28T17:28:34.627433+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One could settle the claim by recomputing $\\rho_h[h]$ and $\\rho_{h_0}[h_0]$ with an independent code that uses direct fast Fourier transforms of time-domain (2.5,3.5)PN waveforms, rather than the stationary phase approximation, over a grid of masses, mass ratios, and inclination angles in the same noise curve, and checking whether the fractional deficit $\\delta = (\\rho_{h_0}[h_0]-\\rho_h[h])/\\rho_h[h]$ reaches 20 to 30 percent in the predicted asymmetric-mass region. If the deficit is absent, changes sign, or only appears under SPA, the central claim is refuted; separately, evaluating the coefficients $\\mathrm{Re}[C^*_{(0)}C_{(m/2)}]$ symbolically from the published amplitudes would falsify the analytic mechanism if any is positive for allowed parameters.","supporting_citations":[{"cited_title":"Damour, B.R","cited_arxiv_id":null,"evidence_quote":"Contains the deferred detailed analytic proof of the main inequality $\\rho_h[h] < \\rho_{h_0}[h_0]$, on which the robustness claim directly leans."},{"cited_title":"Newto- nian","cited_arxiv_id":null,"evidence_quote":"Supplies the explicit post-Newtonian amplitude coefficients and harmonic decomposition from which the amplitude-corrected waveforms are built."},{"cited_title":"Liv. Rev. Rel. 5, 3","cited_arxiv_id":null,"evidence_quote":"Supplies the 3.5PN phasing and energy flux used in the stationary-phase frequency sweep and phase."},{"cited_title":"Blanchet, B.R","cited_arxiv_id":null,"evidence_quote":"Supplies the detector noise spectral density used in all SNR integrals and the faithfulness/effectualness distinction that frames the caveats."},{"cited_title":"Buonanno and T","cited_arxiv_id":null,"evidence_quote":"Introduces the number of useful cycles and supports the choice of frequency band for SNR estimates."},{"cited_title":"Buonanno, Y","cited_arxiv_id":null,"evidence_quote":"Defines effectualness versus faithfulness, the distinction that separates the robust main inequality from the more caveated restricted-template comparison."},{"cited_title":"Damour, B.R","cited_arxiv_id":null,"evidence_quote":"Shows that stationary-phase and fast-Fourier-transform SNR computations agree to a few percent for the relevant mass range, justifying the SPA approach."},{"cited_title":"Droz, D.J","cited_arxiv_id":null,"evidence_quote":"Documents the earlier observation that SNR increases at 0.5PN amplitude order, which the paper reproduces and then shows reverses at higher order."}],"review_version":1}