AES Journal Forum Comments

AES Journal Forum: Comment by Kevin Gross on "The Loudness War: Do Louder, Hypercompressed Recordings Sell Better?"

Mon, 23 Apr 2012 00:38:37 -0600

Title: The Loudness War: Do Louder, Hypercompressed Recordings Sell Better?
JAES Volume 59 Issue 5 pp. 346-351; May 2011
Comment by: Kevin Gross

Loudness is the least of it as far as I'm concerned. Here in the US, moviegoers are treated to 30 minutes of back-to-back television advertisements then 15 minutes of movie advertisements (previews) before the feature starts. It is difficult to get me to the theater these days. Others are not deterred. Engineers apparently have a different way of looking at things. This should be kept in persepective in these discussions.

AES Journal Forum: Comment by Robert Auld on "The Loudness War: Do Louder, Hypercompressed Recordings Sell Better?"

Fri, 20 Apr 2012 11:35:30 -0600

Title: The Loudness War: Do Louder, Hypercompressed Recordings Sell Better?
JAES Volume 59 Issue 5 pp. 346-351; May 2011
Comment by: Robert Auld

My own feeling on the motion picture trailer loudness issue (this is not a scientific comment, but just based on my own observations) is that the producers who fueled the loudness race in picture trailers shot themselves in the foot. One reason I have largely stopped going out to theaters to see movies is because I do not enjoy being subjected to several minutes of loud, blasting trailers before getting to see the movie itself. Indeed, for awhile I was in the habit of bringing ear plugs and putting them in during the trailers, then removing them for the feature!

I cannot believe that no members of the public shared my reaction. The loud trailers may have been a factor in driving a segment of the public away from viewing movies in theaters. And there is no question that the movie going experience, over the years, has deteriorated—small theaters, poor sound systems, loud trailers and high ticket prices have turned movie going into an expensive, unpleasant experience. It is no wonder, what with the improvements in home video playback, that many people prefer to watch movies at home.

As for the main discussion, I agree that more research is needed to definitively settle the question of whether hyper-compression affects record sales one way or the other. But my own bias is that content is the main driver of sales, and that trashing the sound of the content with excessive compression does not help sell that content.

AES Journal Forum: Comment by Mohamad Al-alaoui on "Improving the Magnitude Responses of Digital Filters for Loudspeaker Equalization"

Mon, 05 Mar 2012 09:34:59 -0600

Title: Improving the Magnitude Responses of Digital Filters for Loudspeaker Equalization
JAES Volume 58 Issue 12 pp. 1064-1082; December 2010
Comment by: Mohamad Al-alaoui

I am grateful for these comments that are basically correct. The comments concern the manuscript that was published in December 2011 in JAES [1]. In [1] the method introduced in [2] was applied to the bell filters examples that appeared in [3] together with a comparison with the MTZi method that was introduced in [3].

There was indeed an error in the MTZi code. However, after correcting the code the MTZi method worked better than the article method for cases 1 and 3. The article method is clearly better, after correction, for cases 4, 7, and 9 and rather similar for the other cases. In all the cases the article method gave lower order filters, order 2, the same as the analog filters for all the 10 cases while the MTZi filters are of order 4 for cases 1, 3, 9, and 10 and of order 3 for the remaining cases.

The second comment is also correct. Indeed the MZT for the bell filter give the same results as the article method. This was observed in two of the 6 examples presented in the original paper that presented the article method while in all the other examples the article method performed clearly better than the MZT [2]. In that paper all the examples were also compared with MZT method. Thus for some filters, and this obviously include the bell filter, the article method and the MZT method coincide. Due to the error in the MTZi code, indicated in the first comment, the article method yielded better results than the MTZi method and since the MTZi method was shown to be better than the MZT method this seemed to obviate the need for comparison with the MZT method.

It should be noted that the MTZi paper dealt with cases 1 and 2 and [1] added the other cases which show now after the correction that MTZi give better results than the MZT only in some cases, and with higher order filter, and this is indeed a contribution of the new article.

It should be noted that comments started as a private correspondence and I proposed that the discussion should be public. I am grateful indeed for the valuable comments that gave the opportunity to improve the results in [2]. With these improvements better results than MTZi in all cases will be obtained and the improved method does not coincide with the MZT Method. I will be submitting the manuscript which addresses also the error in [1] to JAES in the near future.

[1]M. A. Al-Alaoui, “Improving the Magnitude Response of Digital Filters for Loudspeaker Equalization ”, Journal Audio Engineering Society, Volume 58, Number 12, pp. 1064-1082, December 2010.

[2] M. A. Al-Alaoui, “ Novel Approach to Analog to Digital Transforms,” IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications, . Vol. 54, No. 2, pp.338-350 , February, 2007.

[3] D. W. Gunness, O. S. Chauhan, “Optimizing the Magnitude Response of Matched z-Transform Filters (“MZTi”) for Loudspeaker Equalization”, AES 32^nd International Conference Proceedings, pp. 1-10, Hillerod, Denmark, September 2007.

AES Journal Forum: Comment by Nay Oo on "Perceptually-Motivated Objective Grading of Nonlinear Processing in Virtual-Bass Systems"

Sat, 11 Feb 2012 10:12:44 -0600

Title: Perceptually-Motivated Objective Grading of Nonlinear Processing in Virtual-Bass Systems
JAES Volume 59 Issue 11 pp. 804-824; November 2011
Comment by: Nay Oo

Thank you again for your interest, research efforts, and comments.

I would like to add on a few things. The physical bass-boasting nature of NLDs is related to the derivative analysis of the nonlinear IO curves. Please see Table 14, Table 15 and Figure 11 of our paper. Your findings indicate that the good NLDs have physical bass boasting effect together with virtual bass effect caused by missing fundamental. Bad NLDs suppress the physical bass or lack of the physical bass. Our derivative analysis of NLD curves in the paper shows the same thing. In my humble opinion, it is not feasible to normalize the physical bass component (i.e., fundamental frequency or first harmonic) without modifying the nonlinear IO curves.

Moreover, bass intensity is only one half of the issue; the other half is the distortion artifacts caused by NLDs and deployment of Rnonlin model [1] to predict the scores.

Ref:
[1] C. T. Tan, Moore, B. C. J., Zacharov, N., and Matilla, V.-V., "Predicting the perceived quality of nonlinearly distorted music and speech signals," J. Audio Eng. Soc., vol. 52, pp. 699-711, 2004.

AES Journal Forum: Comment by Christophe Macours on "Perceptually-Motivated Objective Grading of Nonlinear Processing in Virtual-Bass Systems"

Fri, 10 Feb 2012 10:17:22 -0600

Title: Perceptually-Motivated Objective Grading of Nonlinear Processing in Virtual-Bass Systems
JAES Volume 59 Issue 11 pp. 804-824; November 2011
Comment by: Christophe Macours

Thanks again for your answers.

It would have been wise to ensure all NLDs have a unity gain on the physical bass components. That would have helped the assessment of the true virtual bass enhancement, the one resulting from the generated harmonics only without boost of the physical bass components. Indeed, an FFT comparison of the NLDs in terms of fundamental frequency gain yields the following results (from the highest to the lowest gain, sinewave input) :

ATSR: +5.65 dB
EXP2: +5.05 dB
FEXP1: +0.60 dB
NTANH: +0.55 dB
NSIG: +0.15 dB

REF: +0.00 dB

FEXP2: -0.70 dB
SQS: -1.40 dB
CUBE: -2.50 dB
CLP: -4.30 dB
HWR: -6.00 dB
ASQRT: no fundamental
FWR: no fundamental
SQL: no fundamental

Interestingly, the "Good" bass enhancers are precisely those with a positive gain on the fundamental frequency. It seems fair to state that the author's conclusions would have been quite different if all NLDs would have had their fundamental frequency gain normalized to unity, such that they would only have differed by their non-linear behaviour.