===========Reviews========================

======= Review 1 =======

> *** Technical quality and correctness: Rate the technical quality and correctness
Good (4)

> *** Novelty: Rate the novelty
Average (3)

> *** Clarity and written quality: Rate the clarity and written quality
Excellent (5)

> *** Relevance to the conference: Rate the relevance to the conference
Excellent (5)

> *** References: Rate the references
Good (4)

> *** Overall evaluation: Overall evaluation of this paper
Accept (4)

> *** Reviewer's comments on this paper: While adding your comments please strictly avoid the return button!

A method for idetifying and improving the best quality live audio recording is proposed. It seems to me that the paper is really well written and technically correct. On the other side, it just presents an application of well known procedures of signal processing. Nevertheless, a clever and interesting application that is relevant to the conference.

A minor comment. I would suggest not to use the fractions (\frac{}{}) in the inline equations.

======= Review 2 =======

> *** Technical quality and correctness: Rate the technical quality and correctness
Good (4)

> *** Novelty: Rate the novelty
Good (4)

> *** Clarity and written quality: Rate the clarity and written quality
Good (4)

> *** Relevance to the conference: Rate the relevance to the conference
Good (4)

> *** References: Rate the references
Average (3)

> *** Overall evaluation: Overall evaluation of this paper
Accept (4)

> *** Reviewer's comments on this paper: While adding your comments please strictly avoid the return button!

The paper is we written and easy to read. The only problem is to check the proposed examples that are not accessible through the proposed address in the paper (http://networks.cs.northwestern.edu/~aaron/fusion).
Hence the results is difficult to evaluate. The authors are kindly asked to put a link valid in the paper and permanently accessible to all even after the conference.

======= Meta review 3 =======

> *** Recommendation: Overall recommendation based upon the reviews
Definite Accept (5)

> *** Reviewer's comments on this paper: While adding your comments please strictly avoid the return button!

(1) The paper is well written and showed a methodology to enhance the signal quality by combining the audio records made in a noisy environment. They provided the possible signal to noise improvement one can otain in different condition. The first result is well known in array processing community and should list at least one or two references.
(2) Extension to frequency domain is given -- however, it is not clear if any advantage is there doing additional processing in prequency domain as they convert the signals back to time domain for ranking purposes. They should include an fft of the combined signal using the time domain approach and its extension to frequency domain to see the improvements.

> *** Summary of all reviewer comments: While adding your comments please strictly avoid the return button!

(1) Paper is well written
(2) the website networks.cs.northwestern.edu/~aaron/fusion provided does not work -- should include the correct website
(3) avoid using the fractions in in-line expressions
(3) give references from array processing to show the improved SNR for the case I.
(4) show the evidence of improvement by providing the fft plots of fused data using (a) time domain approach and (b) extension to freq. domain