Journal of the Society of Motion Picture and Television Engineers (1950-1954)

Mr. Zambuto: Sometimes measures of increased signal-to-noise ratio which can be made on certain frequencies will not tell the whole story, particularly when you consider recording of speech. This is due to the effect you were just mentioning, that is, spectral distribution of energy in speech. I happen to have made some experiments on this and we had some rather interesting results, particularly regarding the spectral distribution of speech in different languages. For instance, in Italian we found that we have more middle frequencies around 2500 to 4500 than you have in English, the latter being richer in higher frequencies. Therefore, it often is the experience of Italian mixers that if you boost that particular range of middle frequencies, you have to pull down your average level in order not to overmodulate, and therefore you reach 100% modulation at a lower power level; as a result you lower your expected signal-to-noise ratio considerably. I think this effect should be kept in mind in evaluating the advantages of a new standard. Mr. Hilliard: It has been my opinion, and I think it is being confirmed currently, that the amount of equalization proposed is as drastic as was used in the NARTB curve and it has gradually been dropped. The AES disk recording and reproducing curve indicates the trend, and I know from experience in motion-picture sound recording that this amount of equalization was tried originally and has been dropped because of the fact that you would penalize yourself in overall level or overload, or a combination of both. Mr. Ryder: At Paramount, from the inception of magnetic recording, we have used pre-equalization in the amount of between 10 or 12 db at 8000 cycles. We have had good success recording that way and I should say far less trouble than we had in other recording activity. At Ryder Services, when I do outside work I comply with the standards of the industry, I do not use pre-equalization. So I have one plant, namely Paramount, with preequalization; another plant, Ryder Services, without pre-equalization, and believe me, I prefer the pre-equalization. I'm on the side of Ampex. If there is varying contact in recording and reproduction, the effect is not as bothersome in my work with pre-equalization as in my work without pre-equalization. I am very much in favor of preand post-equalization, especially as we diminish the width of the sound track. Dr. Frayne: I believe since we have supplied some equipment to Paramount Studios I can elaborate a little on Mr. Ryder's remarks. I'm quite sure that Mr. Ryder continued to use the same preand post-equalization on magnetic that Paramount has used on photographic recording and which we had supplied to him. That equalization is not nearly so drastic as Ampex proposes. It flattens off, as a matter of fact, around 6500 cycles, and at least as originally supplied it does not exceed 12 db. Now, also bear in mind that the experience at Paramount is with original recording, not on release. As far as I know, Paramount has not released any film with that degree of preequalization, because the theaters couldn't play it back. You'll notice that the Research Council is proposing some preequalization. The recording characteristic was not shown here today, but it suggests about 3 to 4 db pre-equalization at 8000 cycles only so that we can get a fairly flat overall response with the proposed reproduction characteristic of the preamplifier. Mr. Ryder: We have actually tried this out in the theaters, and the first magnetic installation for theater reproduction at the Chinese in Hollywood was made for Paramount and under our supervision. The Chinese happens to be one of the key theaters in Hollywood. We have done a fair amount of experimenting in this field, and we still think that we're right, but this is the type of thing that makes for good horse racing and we like that, too. Dr. Frayne: I'd like to add, that in the Chinese installation they were 200-mil tracks which have a slight advantage over the coated 50-mil track. Mr. Athey: Apparently the spectrum problem of low-frequency pre-emphasis does not seem to worry anybody very much. I assume that something is supposed to give in the system at high levels. You may notice that I pointed out that in a sense we take advantage of it in order to reduce some of the requirements on our amplifiers. It seems to me that the real reason for low-frequency pre-emphasis? 236 March 1954 Journal of the SMPTE Vol. 62