1: \begin{abstract}

2:   We present numerical relativity simulations of nine-orbit 

3:   equal-mass binary neutron star covering the quasicircular late

4:   inspiral and merger. 

5:   The extracted gravitational waveforms are analyzed for convergence

6:   and accuracy. Second order convergence is observed up to contact,

7:   i.e.~about 3-4 cycles to merger; error estimates can be made up to

8:   this point.

9:   The uncertainties on the phase and the amplitude are dominated by

10:   truncation errors and can be minimized to $0.13$~$\rad$   

11:   and $\lesssim1~\%$, respectively, by using several simulations

12:   and extrapolating in resolution. 

13:   In the latter case finite-radius extraction uncertainties become 

14:   a source of error of the same order and have to be

15:   taken into account.

16:   The waveforms are tested against accuracy standards for data analysis.

17:   The uncertainties on the waveforms are such that accuracy standards 

18:   are generically not met for signal-to-noise ratios relevant for

19:   detection, except for some best cases using extrapolation from several runs. 

20:   A detailed analysis of the errors is thus imperative for the

21:   use of numerical relativity waveforms from binary neutron stars

22:   in quantitative studies.

23:   The waveforms are compared with the post-Newtonian Taylor T4

24:   approximants both for point-particle and including the analytically

25:   known tidal corrections. 

26:   The T4 approximants accumulate significant phase differences of 

27:   $2$~$\rad$ at contact and $4$~$\rad$ at merger, 

28:   underestimating the influence of finite size effects.

29:   Tidal signatures in the waveforms are thus important at least during the last six orbits of the merger process.

30: \end{abstract}

31: