The plaintiffs appeal because of its inadequacy from a judgment in their favor in an action brought upon a written contract with the defendants, entered into on March 20, 1928, by which they sold their business as publishers of musical compositions. Part of the property sold was the copyrights of 29 musical compositions: 3 piano solos, 1 violin solo, 3 organ solos and 22 songs. That part of the contract out of which this action arose we quote in the margin.1 The plaintiff, Pallma, had been in the defendants' employ, but had left them in 1925 to form the plaintiff firm. However, by 1928 he and Warner were deeply in debt and wished to sell out, and, after they had, Pallma went back into the defendants' employ where he remained until 1933. The defendants rendered accounts to the plaintiffs under the contract until the year 1939, covering some of the matters mentioned in the contract which are no longer in dispute; but for the "bulk revenues," so called, they accounted only in part, and in a very small amount. The plaintiffs filed their complaint on October 26, 1939, demanding, among other matters now not involved, an account of these revenues; and Judge Mandelbaum directed the defendants to account by an interlocutory judgment, entered June 30, 1943, in which he appointed a master to state the account. The master reported on September 25, 1946, and Judge Goddard on July 3, 1947, confirmed the report in part and rejected it in part, remanding some of the issues to the master on October 14, 1947. The master made a second report on April 12, 1949, and on July 26 of that year Judge Goddard also modified this report in part and confirmed it as modified. It is from this judgment that the plaintiffs appeal upon the ground that they have not been allowed enough as their share of the "bulk revenues."

These were from two sources: payments made by the American Society of Composers, Authors and Publishers, known as ASCAP; and payments made by moving picture producers who "synchronized" music with their films. ASCAP's practice, with which the plaintiffs were familiar, was for music publishers, who were members, to grant it the right to exploit a catalogue of their compositions, for the performance of which ASCAP in turn granted licenses to radio broadcasters, hotels, dance halls, theatres, cafes and other public performances. These licenses gave the licensees the privilege of performing at will the compositions of all publishers which ASCAP controlled, in consideration of an annual lump sum, which was not dependent upon the extent to which the licensee might use the compositions. The gross, or "bulk," revenue so received ASCAP divided in half, giving one part to authors and composers and the other to publishers: we are here concerned only with the publishers' half. That was divided among all the publishers during the period with which this case is concerned by two different methods; one, from 1928 to 1936; the other, from 1936 to 1945. During the first period the publishers' half was divided by assigning to each publisher a "class" — A, B, C, etc. — which entitled him to a prescribed percentage of three-fourths of the half. His "class" was determined periodically by a "Classification Committee" which appraised his catalogue, taken as a whole, upon the basis of what was called the "availability" of the compositions in it. This committee did not try to appraise "availability" by any quantitative measure, but by an overall estimate of the popularity of the compositions. It did not, and would not, disclose the factors which determined its decision, except by vague variants on "popularity," such as "importance," "character" or "vogue." As we have just said, a publisher's "class" determined his percentage of only three-fourths of the publishers' half; the other fourth was distributed on the basis of what was called his "seniority," which was the sum of "credits" that had been allotted to him in past years, the "credit" for each such year being in its turn determined by his class in that year. After 1936 the following method was used. A record was kept of the performance over four radio networks of each composition in the catalogue of each publisher, and the composition was credited with a "point" computed by the aggregate of its performances. One-half of the publishers' half was divided by crediting to each publisher the percentage that the sum of his "points" bore to the total "points" of all member publishers. The other half was divided as the whole had been before 1936, except that "availability" then counted for only three-fifths instead of three-fourths of a publisher's dividend, and in consequence "seniority" counted for two-fifths instead of one-fourth.

The defendants dealt with each moving picture producer by a separate contract, but these fell into three groups. In the first group were contracts in which the producer paid an annual lump sum for the privilege of using all the compositions of a publisher's catalogue as he might choose; thus these were like the contracts which ASCAP made with its licensees. In the second group were the contracts of producers who paid a prescribed sum for each use of a composition, but with an annual minimum for the whole catalogue. In cases in which the sum of the use payments was less than the minimum this group would be the same as the first group; but the master made no distinction to provide for the possibility that the minimum might be less than the sum of the use payments. However, as there were only two of these contracts whose "revenues" together amounted to only $7,000 out of over $560,000, and as neither party has apparently objected to this omission, we shall treat the second group as identical with the first. The third group of contracts was of those in which the producer credited the publisher with a prescribed amount for each use of a composition with a minimum which could however be carried over from year to year. These afforded a ready means of separating the return upon each composition, and the master did not include them as "bulk revenue"; and the plaintiffs do not assert that they constitute a separate unsatisfied liability. Nor do we find it necessary to consider that part of the defendants' receipts which they received under what went by the name of the ERPI contracts, for the master found that in this sum — $175,000 — the plaintiffs were entitled to share only on the "use basis" on which the payments had been computed, and that the defendants had "already accounted for" that share.

The master in his first report, following Judge Mandelbaum, held that, although the defendants were justified in putting the plaintiffs' 29 compositions into their catalogue and allowing ASCAP and the moving picture producers to make payments upon the catalogue as a whole, they were wrongdoers in failing to keep that part of the "bulk revenues" which came from the plaintiffs' compositions separate from the part that came from their own compositions. He did not, however, apply the extreme doctrine that, since because of this wrong the defendants were under a duty to disentangle the two contributions, all doubts should be taken against them. Indeed, if he had, it would have been difficult to avoid the conclusion that the plaintiffs could claim all the "bulk revenues"; and even the plaintiffs themselves do not assert that. The master appears to have taken the intermediate position of somewhat weighting the scales against the defendants, when he was in doubt. He found that the change of the defendants' "class" from C to B in April of 1928, was owing to the addition of the Pallma compositions to their catalogue, which they put forward to ASCAP as a reason in a letter of February, 1928. He was confirmed in this conclusion because of a letter written in March, 1932, in which they asked for a still further advance in "class," and mentioned 33 songs in their catalogue as especially deserving, of which two were Pallma songs. He also relied upon a survey, made shortly before the contract, based upon the number of performances of Pallma songs compared with performances of the defendants'. He concluded from these facts that the plaintiffs had contributed six percent to the ASCAP "bulk revenues" over the entire period, and this amount he allowed them.

Judge Goddard reversed this finding because it was without adequate support in the evidence, and, as was inevitable, he also reversed the master's finding that the same percentage should be applied to the moving picture producers' "bulk revenues." He held that the defendants' ASCAP revenues after 1936 should be divided in the proportion which the performances of the Pallma compositions bore to the defendants', performance being the basis of half the ASCAP award to the defendants after 1936. In so doing he made no allowance for possible differences between a performance basis, and an "availability" or "seniority" basis on which ASCAP had computed the other half of these receipts. He gave no instructions about the proper division of the moving picture producers' "bulk revenues." On the second hearing the master divided the ASCAP revenues before 1936 in the proportion of the sheet sales of Pallma compositions to the whole catalogue; and he divided the moving picture producers' revenues over the whole period in accordance with the sheet sales. Judge Goddard confirmed these findings as to ASCAP revenues, and to this the plaintiffs have not objected. He did not, however, accept the master's division of the moving picture producers' revenues; nor would he accept the defendants' argument that they should be divided in the proportion of performance. Sheet sales he thought did not allow for the use of the compositions as "background," which was part of the "bulk" contracts with moving picture producers. "Performance" percentages he discredited because, although the defendants in the ERPI contract had bargained on the basis of performances, they had not used them in the "bulk" contracts. He therefore took half the difference between the sheet sales and the performance percentages; and the final judgment was framed accordingly, the expenses of the reference being divided between the parties.

We cannot agree that the defendants are subject to the burden which the law places upon a wrongdoer — usually a fiduciary — which resolves all doubts against him in separating the beneficiary's interest from his own. The contract of 1928 expressly permitted the defendants to combine the plaintiffs' compositions with their own in a single catalogue; and, indeed, the master and both district judges conceded that to this extent the defendants did no wrong. But, that once being conceded, the defendants were justified in exploiting the combined catalogue as they would have exploited their own, and in collecting the revenues in "bulk" whenever that was customary in the publishing business. So far as concerns the ASCAP revenues, there can be no possible doubt of this; the plaintiffs were not members of ASCAP, which would surely not have consented to treat them as such: the Classification Committee would not have given them a classification based on any such catalogue. The difficulties of procuring separate allowances for Pallma compositions were perhaps not so insuperable in dealing with the moving picture producers', but it would be entirely gratuitous and unwarranted to assume that these producers would have separated the "bulk revenues." Yet without some such separation in the case of both ASCAP and the producers, the plaintiffs' share had to remain the subject of compromise, or of ascertainment by what in candor can be no better than the honest guessing, which has pervaded, and was bound to pervade, this litigation. Therefore, although we recognize that this interpretation of the contract puts the plaintiffs at a disadvantage, that is inherent in the contract itself, and is far more just than to impose any burden of proof upon the defendants for performing the contract in the way which the plaintiffs either intended, or were chargeable with notice that it would be performed. Nor does the defendants' failure to account for the "bulk revenues" throw any burden of proof upon them; an accounting, when not voluntary, always presupposes that the accounting party has been in default; he can properly be charged with the expenses of the proceeding, but with no more. Upon any accounting, he does indeed have the burden of proving any credits, if they are challenged; but the other party has the burden of proving all "surcharges"2 and in the case at bar we are concerned only with those. Therefore the plaintiffs had the burden throughout.

Coming then to the details, we agree with Judge Goddard that the master's finding was "clearly erroneous" that the plaintiffs contributed six percent of the ASCAP revenues over the whole period. We do not forget what we have so often said, and what indeed Rule 53(e) (2), Federal Rules of Civil Procedure, 28 U.S.C.A., makes peremptory; i. e., that a master's findings are as conclusive upon the district court as that court's findings are conclusive upon us. Nevertheless some review is always open; and it extends as much to untenable inferences from conceded evidences as to the credibility of testimony. The defendants in their letter of February 28, 1928, did indeed notify ASCAP that, beginning the next day, they would exploit the Pallma compositions as part of their catalogue. Moreover, that was while the defendants' application for a change in "class" was pending, and it was put forward as one ground for a change of "class." On the other hand, not only did ASCAP's answer give no indication that the Pallma compositions had anything to do with the result, but it explicitly stated that "the addition of the Pallma credits to yours would have no effect upon your standing." The master was in error in striking this out; the plaintiffs had put in the letter as evidence of ASCAP's action upon it; the answer was part of that action. Again, the defendants' letter, four years later — March 22, 1932 — gave no support to the master's finding. It did indeed mention 33 "numbers," which were "popular as well as standard successes," and of these two were Pallma songs; but there is as little ground in that for supposing that the Pallma compositions contributed six percent to the revenues, as there would be for assuming that they contributed two percent because in the same letter the defendants mentioned 52 "outstanding composers" in their catalogue, of whom only one was among Pallma composers. Nor can we accept the master's exclusion of "theatre performances" from the so-called survey of 1927. It is true that Mills testified that of the music played at motion picture performances (which is to be distinguished from that "synchronized" in films), "much" was not "regularly published, copyrighted and available in some form to all licensees"; and that such music ASCAP did not count; but he also said that "much of it was" so published. It was utterly unjustified therefore to eliminate from Mills' letter of March 5, 1928, all theatre credits. If we were to use one-half of them, the plaintiffs' contribution would be less than one percent; and, even if we were to take only one-fifth, it would be little more than one and a half. Finally, the notion that Mills' opinion was not reliable evidence, because he was not a member of the "Classification Committee," is equally untenable. His duties constantly brought him in contact with classifications; he was the chairman of the Administrative Committee; and if there was an expert, certainly he was one.

For these reasons we hold that Judge Goddard was right in declaring the first report "clearly erroneous"; and no question arises as to whether he was right also in accepting the second report as to the ASCAP revenues, for the plaintiffs rest their appeal as to these solely upon the argument that the first report was right. There remains therefore only the division of the moving picture producers' revenues. The master's first allowance was made on the assumption that the producers' contribution to these could properly be measured by their contribution to the ASCAP revenues; and, as we have said, it fell with his finding as to them. Was he right in using sheet sales as a proper standard of division in his second report? If that was not right, was Judge Goddard right in substituting the mean between the sheet sales percentage and the performance percentage? The master used the sheet sales as a standard, exclusive of any performance factor, because he understood Judge Goddard to have ruled that the performance standard was not applicable to the "bulk" contracts of moving picture producers. It turned out that he had misapprehended the judge, and his actual finding was therefore irrelevant. The judge had, therefore, either to decide the issue for himself on the evidence, or to send the case back a third time; and he very wisely decided to decide it himself. It is his finding, not the master's, which is before us.

Performance is a better standard of contribution than the sheet sales, and for that reason the master and the judge did use it for the second period of the ASCAP revenues, during which that standard was available. Indeed, if the plaintiffs had so argued, it might have been a closer approximation to the truth, to carry the performance standard back over the first period. The master thought, and the plaintiffs argue, that there were no reliable statistics on which to base a performance percentage applicable to the moving picture revenues, and that therefore only the sheet sales standard was available. We think that this is not true; that there was reliable evidence; and that Judge Goddard was right in overruling the master, so far as he can be said to have made any finding to the contrary. The percentages appear on Exhibit R-A, consisting of two sheets purporting to contain the number of performances of the defendants' whole catalogue and that of the Pallma compositions. This exhibit was typed in the office of the defendants' attorneys from information furnished them by one of the defendants, Sam Fox, which he in turn took from books kept in the defendants' offices. The method of keeping the books Fox explained. The moving picture producers regularly sent to the defendants "cue sheets," containing the numbers of performances of all compositions in the defendants' catalogue, including any Pallma performances. These were filed and preserved, and their contents were from time to time transcribed into the defendants' books by employees, called "auditors." Two or three omissions were apparently discovered in the books at the earlier hearings, but even if these made the books not conclusive, they did not make them incompetent. Fox's testimony made them competent for the following reasons, under the federal statute,3 as well as under § 374-a of the New York Civil Practice Act. The "cue sheets" themselves were records kept by the producers and regularly delivered to the defendants; presumably they were made up by persons authorized to make them. It was part of the duties of the "auditors" to enter their contents in the books, and, although this was not done at once, but only when it was found convenient, that was within a "reasonable time," for there was no need for haste, since the "cue sheets" were preserved. True, Exhibit R-A was a transcript from the books, and the plaintiffs objected to it on the ground that the books would be "the best evidence" which we will accept as an objection to secondary evidence of their contents. However, to produce the books themselves would have been cumbersome and dilatory, and Fox consented to let the plaintiffs examine them at length in his office. Therefore the master took this course. He admitted the exhibit, and twice adjourned the hearings so as to give the plaintiffs a chance to compare the exhibit with the books, and to adduce any testimony to show that it was not an accurate transcript; and they never did introduce any such testimony. This was certainly a valid substitute for the production of the books themselves, and justified the use of secondary evidence.4

As we have said, performance was a much better standard to test contribution than sheet sales. The producers performed those compositions of the whole catalogue which they thought would best please their audiences, and their judgment was the best index available of the relative drawing power of "synchronized" music. Thus, even though we disregard the factor of "background" — a term which appears in the moving picture contracts, but whose meaning is by no means clear — the contribution of a composition to the "bulk revenues" is better so appraised than by the number of people who wished to buy copies of it for individual, or even orchestral, performance. If Judge Goddard erred at all, it was in taking a mean between the sheet sales and the performance, instead of using only the standard of performance.

There is a disputed item of $310.77, as to whose proper allocation we cannot be sure. The judge deducted it from the amount of the percentages, because he supposed it had been already paid; but the plaintiffs say that it had been computed upon the total — $562,944.70 — of all the "bulk revenues" from the moving picture producers, before deducting the ERPI payments — $175,000. As we understand it, this would increase the recovery by 175/563 of $310.77 — somewhat less than $97. It is hard to say that the plaintiffs have proved their right to this, but since the defendants' brief does not answer the claim, we will allow so much of the item.

My brothers think that, since the defendants were only willing to concede that they owed $313.83 to the plaintiffs, and the plaintiffs have now recovered nearly $8,500, the defendants should pay all the expenses of the reference. Although I feel the force of that argument, for myself I should have ruled otherwise. The efforts of the plaintiffs to swell their claim seem to me quite unwarranted; and if they had kept it within bounds, I doubt that the expenses would have been more than half what they were. However, my view is not to prevail, and the defendants will pay all expenses of the references together with an addition of $97 to the recovery.

Judgment modified as above, and as modified, affirmed.