First printed in the December 1935 issue of Gramophone.

According to accounts in the technical Press this is a High Fidelity year. That was the slogan of many firms at the Radio Exhibition, but so far as one can make out very few of them have any real conception of what the term means, or should mean. Presumably they associate it with instruments giving a. higher standard of quality of reproduction, but there appears to be much confusion on the two fundamental questions: ‘Higher than what?’ and ‘On what standard?’

During the past few months it has been my fortune to pass under review a large number of wireless receivers and radiogramophones, of many makes and prices.

From this review two clear and significant conclusions have emerged. The first is that gramophone reproduction has improved considerably during the past year or two. The improvement has been partly due to very welcome advances in recording, but also to substantial developments at the reproducing end. The former shows itself more definitely in acoustic reproduction whereas both are in evidence in the radio-gramophones. There is no doubt, for example, that the general run of pick-ups to-day is much superior to that of a few years ago.

The second conclusion is that, by and large, broadcast reproduction is definitely worse with this year's models than it was before. There are exceptions, notably, for example, at Haves, where the standard has in no case been worsened and in some respects improved. My colleagues with whom I was conducting the review were indeed charmed with the new sixteen-guinea radio-gramophone (Model 370), which is really quite a wonder at its price, and even that has to give points to the special schools model which has just been produced. The new Hartley-Turner, again, is a great improvement on last year's model, good as that was; the cold mercilessness has given place to a warm benevolence which is just as analytical but is more comfortable. Birmingham Sound Reproducers have also taken a step forward in their 1935 models.

The specialist firms with whom we are more familiar in The Gramophone—DaVey Radio, Expert Radio, Godfrey Radio, and now Rimington Radio—were not included in the general review to which I have referred, but from other evidence I have no doubt that they, too, have more than maintained the high standard with which their names have been associated.

These, however, are exceptions to the general rule. It is true that the bass preponderance, about which so many of us have complained, is fast disappearing. But it has been succeeded by a woolliness and a lack of articulation which previous bass boom tended to make less apparent. In some cases there is a peculiar nasality as well. In short, we seem to have come to the soporific stage which one remembers in connection with gramophone reproduction as the Peridulce era: the days of 'mellow tone' and what the Editor called 'romanticism'.

In what sense, then, can this be said to be a high fidelity year? The answer, I think, is that at last there is coming to be a general impression that those people who might be prepared to open their purses are dissatisfied with the quality of reproduction which has recently been offered to them. Hypnotic advertising may suppress sane judgment for a while, but the awakening is apt to be uncomfortable.

All this, however, does not get us much nearer to an understanding of what the term 'high fidelity' should mean. It only tells us one or two things that it does not mean.

When we look critically on the course of progress in sound reproduction, one or two interesting points begin to emerge. The first is that empiricism, that is, the hit and miss method with nothing more, never gets very far. The other is that pure intellectualism, that is, theorising divorced from practice, may go a long way but there is little guarantee that it is the right way. Real progress is only made by a combination of the inductive and deductive processes, that is by the science which theorises on the basis of observation, and tests the theory by a further- resort to experience.

Principle of Uniform Frequency Response

The greatest advance in sound reproduction was made when the principle of uniform frequency response was applied, by means of electrical impedance methods of calculation, to establish wave filter theory. For some time every technician talked about the necessity for impedance matching in order to secure uniform frequency response. The idea soon became prevalent that in order to secure improvements in reproduction all that was necessary was to extend the range of uniform frequency response to embrace as much of the musical scale (which extends from 16 to about 16,000 cycles per second) as possible. At first a range of 100 to 5,000 cycles per second was aimed at and gradually it has been achieved, though there are still very few commercial instruments which reproduce throughout that range with anything like uniformity. The term " high fidelity " came into use, originally in America like most of these terms, to signify a frequency response range from about 50 to about 10,000 cycles per second, and later it came to be applied only to those instruments which also had a so-called undistorted output of about 10 watts.

So soon, however, as the waggon had been hitched to this new star, a number of doubts even as to the principle itself began to take shape. Fuller experience has begun to modify the theorising, and a number of new factors have assumed a measure of importance. To the philosophic technicians the development has seemed quite natural; the blurb writer has invented new clichés to cover his embarrassment at being stripped naked of all his old claims; only the man-in-the-street is left confused and mistrustful.

Unhappily, there is as yet no simple, clear-cut technical standard to take the place of the old one; we are still suffering from growing pains. A number of the considerations which must be taken into account in forming such a standard are, however, becoming quite clear. A discussion of these I must leave to a further article. Here I will content myself with the declaration that although a uniform frequency response, amongst other things, is a necessity for perfect reproduction that is an exact simulation of the original, this perfect reproduction is impossible of achievement save in conditions which render it valueless. The important practical problems are, first, to determine the conditions that must be fulfilled to achieve a good illusion of reality in reproduction in particular surroundings, and second, to discover what departures from these conditions are permissible before the difference in quality becomes appreciable.

For both problems, it should be noticed, the final arbiter is not a mathematical formula, but a consensus of opinion amongst people of normal hearing.