International projectionist (Jan-Dec 1947)

Addendum: Quality' vs. Pleasing' Sound Reproduction By G. R. REDGRAVE Goodman Industries, Ltd. Readers will remember the provocative article under the above title in I. P. for February. This article originally appeared in Electronic Engineering (London), a contribution of J. Moir, acoustical engineer for the British Thomson-Houston Co., Ltd., of Rugby, England. Excerpts from letters of several correspondents to the British technical journal are appended hereto, together with Mr. Moir's comments on the various suggestions advanced. MR. MOIR concludes that the average listening public does not like "top" in: reproduced music. This is a debatable point . . . Far more evidence must be produced before it can seriously be suggested that the mass of ordinary listeners might prefer to have less "top" ;u the live performance. . . . Members of a concert audience do not complain of excessive high-frequency content in the music, nor, for example, would the reception of a "break" on the solo trumpet by an acknowledged master of that instrument suggest that the hearers are in any way dissatisfied. As for the development of the violin, cited by Mr. Moir in support of his suggestion, it is indisputable that the best violins ever produced are those made by Stradivari, who died over 200 years ago. Surely, if improvement of this instrument was possible or desirable, we should have observed some indication of it in the meantime, particularly in view of the greater knowledge of acoustics and wider choice of materiels available in latter years. . . . Mr. Moir admits his findings are based on the results of listening to reproduced music, and we can only conclude, therefore, that there was something present in the reproduction which would not have appeared in the original; or, in other words, that the apparatus suffered from a form of distortion. Equipment Procedure Queried We are particularly surprised that no mention was made of the transient response characteristic of the apparatus used in the experiments. It is our experience that equipment designed purely from the standpoint of achieving a good frequency response characteristic can sound most unsatisfactory; but, for example, a loudspeaker that has been designed for a good transient response curve is always a good loudspeaker, and, incidentally, is found also to possess a good frequency response curve. Finally, anent the engineer, surely there can be no question that his job ideally is to provide perfect reproduction of the studio performance. In practice he should aim to achieve results as near to this as the development of his art will allow. Deliberately to attempt anything else is to encroach on the territory of musicians and of craftsmen of musical instruments, but insofar as the engineer is handicapped — whether by limitations of his technique, by imperfections of the materiels, or by shortcomings of the media of his calling — he may be compelled to abandon his ideal. 'Highs' Are More Distressing He must also be guided by the fact that the general public cannot be interested in wide-range reproduction if this entails the usual disadvantages, as the human ear is more readily distressed by distortion products in the higher frequencies than elsewhere. This being so, a limited frequency range reproduced well will always be more generally acceptable than a wider range involving imperfections, and is therefore permissible as a temporary expedient until such time as the imperfections can be eliminated. By F. SLATER Mr. Moir's article, while indicating new and interesting lines of thought, left one unconvinced that the old ideal of perfect response had been fully explored in his investigations. Much more is neded to be known about the equipment used. If, for example, the higher notes were allowed to radiate from one or more baffle-loaded loudspeakers, the following points could be raised: baffle-loaded speakers focus the higher audio frequencies, an effect which is unnoticeable with the normal domestic receiver, but which immediately assumes a major note when the frequency response of the amplifier is extended. This effect surely becomes more and more obvious as the response is opened up. Speaker Type Important Equipment with baffle-loaded speakers, even where three or four are arranged to cover a wide angle, suffer from this effect and the consequent sharply-defined multiple reflections which result in reduced realism. With two similar extended reponse amplifiers— one coupled to a baffle-loaded speaker, the other to a horn-loaded type — it will be found that the high-frequency beam from the former has a much greater relative strength, and a greater annoyance value. It also intensifies the "point source" impression, being stereophonically less satisfactory than a baffle speaker coupled to an amplifier having a moderately restricted high-frequency response. The horn-loaded type, due to absence of focusing, shows none of these effects when the higher frequencies are reproduced. To me the reduced stereophony due to "point source" effect, and the interference patterns due to high-strength reflected beams, are as much a substantial form of distortion as those more easily measured. By B. C. SEWELL Gainsborough Pictures, Ltd. These attempts . . . have met with difficulties often ascribed to the prevalence of high-frequency distortion factors in film recording, but which probably have a more fundamental basis, as Mr. Moir's evidence suggests. The requirement of speech intelligibility demands a certain minimum high-frequency response, but this requirement is fairly easily met; and beyond this point the criterion is that of "naturalness." The evaluation of data regarding "preference" assumes a clear-cut distinction between the reproduced and directly heard performance. It must be assumed of critical listeners, (including, naturally, musicians) that they automatically adopt a different set of fundamental values when assessing a reproduced performance as opposed to a "live" one. This, I believe, is because the acoustic coloration is so greatly modified that the performance has not the illusion of reality and cannot be judged by the same standards. Stereophonic Reproduction One is tempted to think that the table of preferences might have been greatly modified if the reproduction had been stereophonic, thus permitting the functioning of the normal auditory facility of directional perception. In orchestral recording it is often found that delicate passages — from woodwinds, for instance — have insufficient value when reproduced through (Continued on page 28) 16 INTERNATIONAL PROJECTIONIST May 1947