We have found some errors in the noise source section of Table 1 in [1]. The latter should be replaced by the revised Table 1 presented here.

Table 1. Revision for Table 1 of [1]

View Table

The main contribution to the laser beatnote short-term stability described in [1] was the frequency noise (FM noise) of the external-cavity diode laser (ECDL) through the intermodulation effect, not the FM noise of the distributed-feedback (DFB) laser. Some sentences of [1] should be revised according to this revision. Revised parts in these sentences are highlighted in italic font below.

The sentence at line 6 of the abstract in [1] should read as follows:

The detection noise budget and absolute phase noise measurements are in good agreement with the measured short-term frequency stability of the laser beatnote, at the level of $1.1 \times {10^{- 12}}\, {\tau ^{- 1/2}}$ until 100 s, currently limited by the frequency noise from an external-cavity diode laser setup.

The sentences at the end of page 3257 of Ref. [1] should read as follows:

We calculate that the laser beatnote short-term stability should be limited at the level of about $1 \times {10^{- 12}}$ at 1 s by the ECDL setup. In the DFB setup, the main noise contribution to the total budget is the laser frequency noise at the modulation frequency and harmonics. The expected short-term stability at 1 s of the ECDL setup is estimated at the level of 1.0 × 10^–12 at 1 s, mainly limited by the intermodulation effect and the FM-AM noise conversion process.

The last two sentences of Section 3.B in Ref. [1] should read as follows:

In this case, we extract ${b_{- 2}} = 53 \;{{\rm dBrad}^2}/{\rm Hz}$, yielding an expected Allan deviation ${\sigma _y}(1s) = \sqrt {\frac{{{b_{- 2}}}}{{\nu _0^2}}} =$ 1.3 × 10^–12. This value is in agreement with the expected 1.0 × 10^–12 stability limit reported in Table 1.

A part of the second paragraph of Section 3.C in [1] should be revised:

This result is in good agreement with the short-term stability expected from the detection noise budget (1.0 × 10^–12) and absolute phase noise performances (1.3 × 10^–12), reported in Section 3.B. … As suggested by our detection noise budget, the individual stability of the DFSDS-based microcell ECDL should be at the level of 1.0 × 10^–12 at 1 s in current conditions…

In [1], the sentence at line 8 of the conclusion should read as follows:

The short-term stability of the laser beatnote is limited by the frequency noise of the ECDL.