T 1385/18 (Audio signal decoder / Fraunhofer-Gesellschaft) 08-11-2022
AUDIO SIGNAL ENCODER, AUDIO SIGNAL DECODER, METHOD FOR ENCODING OR DECODING AN AUDIO SIGNAL USING AN ALIASING-CANCELLATION
Fraunhofer-Gesellschaft zur Förderung
der angewandten Forschung e.V.
Koninklijke Philips N.V.
Dolby International AB
Priority - main request (partial)
Priority - auxiliary request 0 (yes)
Documents - availability to the public (yes)
Inventive step - main request (no)
Amendment after summons - taken into account (yes)
Amendments - auxiliary request 0
Amendments - allowable (yes)
Remittal - (yes)
I. An appeal was lodged against the decision of the Examining Division to refuse the European patent application.
II. In its decision, the Examining Division held that the claims of all claim requests did not enjoy the priority of the US priority application
P1: US 61/253,468.
III. The Examining Division also established that documents D3, D4 and D6 were made available to the public before the filing date of the application.
IV. According to the decision, the Main Request was not allowable, because the subject-matter of the independent claims was anticipated by D3, D4 and D6 (see next point). D6 also anticipated the subject-matter of the independent claims of Auxiliary Requests 1, 2, 3, 4 and 5, which, for this reason, were also not allowable. In addition, claims 1, 12 and 13 of Auxiliary Request 2 had no basis in the application as filed.
V. The following documents were mentioned in the Examining Division's decision:
D1: B. Bessette et al. "Alternatives for windowing in USAC", 89. MPEG Meeting 29 June 2009 - 3 July 2009, document M16688
D2: B. Bessette et al. "Universal Speech/Audio Coding Using Hybrid ACELP/TCX Techniques", 2005 IEEE International Conference on Acoustics, Speech and Signal Processing, Vol. 3, 18 March 2005, pages 301-304
D3: M. Neuendorf et al. "Completion of Core Experiment on unification of USAC Windowing and Frame Transitions", 91. MPEG Meeting; 18 January 2010 - 22 January 2010, document M17167
D4: M. Neuendorf et al. "Proposal for Unification of USAC Windowing and Frame Transitions", 90. MPEG Meeting, 26 October 2009 - 30 October 2009; document M17020
D5: European Patent Office: "EPO and IEC agree to cooperate on standards and patents", 17 April 2013
D6: M. Neuendorf "WD7 of USAC", 92. MPEG Meeting, 19 April 2010 - 23 April 2010, document N11299
E1: ISO/IEC JTC 1/SC 29 "ISO Technical Management Board Resolutions on 1) Press Attendance at ISO Meetings and 2) Tape Recording at ISO Meetings", 5 September 1996, document N1756
E2: ISO/IEC JTC 1/SC 29 "ISO/IEC JTC 1 Policy on Electronic Document Distribution using World Wide Web", 12 March 1997, document N1963
E3: ISO/IEC JTC 1/SC 29/WG 11 "MPEG 101", 18 April 2005, document N9162
E4: "ISO/IEC JTC 1 Directives, 5th Edition, Version 3.0", 5 April 2007, document N8557
E5: "About MPEG", Extract from the Internet, 7 October 2010
E6: ISO/IEC "ISO/IEC JTC 1 Standing Document N 12", First Edition, 2010
E7: ISO/IEC "ISO/IEC JTC 1 Standing Document N 12", Second Edition, 2013
E8: ISO/IEC "ISO/IEC JTC 1 Standing Document N 23", First Edition, 2015
E9: ISO/IEC "ISO/IEC JTC 1 Standing Document N 19", Second Edition, 2016
E10: Sworn statement by Schuyler Quackenbush, December 2016
E11: "Abut MPEG", Extract from the Internet, 21 February 2010
E12: extracts from E11 and E4
E13: ISO/IEC JTC 1/SC 29/WG 11 "Resolutions of 92nd meeting", 92. MPEG Meeting, 19 April 2010 - 23 April 2010, document N11241
E14: "MPEG Basics", Extracted from the Internet on 24.04.2015
VI. An appeal was lodged against this decision by the applicant. In their statement of grounds of appeal, the appellant requested that the decision under appeal be set aside, and that a patent be granted on the basis of the Main Request or one of five Auxiliary Requests, all as underlying the decision.
VII. In a communication annexed to a summons to oral proceedings, the Board informed the appellant on its provisional opinion, according to which:
(a) The independent claims of the Main Request had to be understood as generic "OR"-claims, which enjoyed a partial priority.
(b) Documents D3, D4 and D6 were made available to the public before the priority date.
(c) The part of the subject-matter of claim 1 of the Main Request that did not enjoy the priority of P1 was not inventive over each of D3, D4 and D6.
(d) Auxiliary Requests 1, 2 and 3 were likely to fail for the same reason as the Main Request.
(e) The priority might be valid for the claims of Auxiliary Requests 4 and 5.
(f) A remittal to the Examining Division was likely for an examination of those aspects that had not been addressed by the Examining Division.
VIII. In response to the Board's communication, the appellant submitted a new Auxiliary Request 0, to be ranked between the Main Request and Auxiliary Request 1. This request was replaced, during oral proceedings, by a new Auxiliary Request 0. The final requests of the appellant were therefore to set aside the decision under appeal and to grant a patent upon the basis of the Main Request, or (new) Auxiliary Request 0, or one of Auxiliary Requests 1 to 5.
IX. Of the seven claim requests on file, only the Main Request and the (new) Auxiliary Request 0 are relevant for the present decision.
X. Claim 1 of the Main Request reads (reference signs removed):
A multi-mode audio signal decoder for providing a decoded representation of an audio content on the basis of an encoded representation of the audio content, the audio signal decoder comprising:
a transform domain path configured to obtain a time domain representation, in the form of an aliasing-reduced time-domain signal, of a portion of the audio content encoded in a transform domain mode on the basis of a first set of spectral coefficients, on the basis of a representation of an aliasing-cancellation stimulus signal and on the basis of a plurality of linear-prediction-domain parameters,
wherein the transform domain path comprises a spectrum processor configured to apply a spectral shaping to the first set of spectral coefficients in dependence on at least a subset of the linear-prediction-domain parameters, to obtain a spectrally-shaped version of the first set of spectral coefficients,
wherein the transform domain path comprises a first frequency-domain-to-time-domain converter configured to obtain a time-domain representation of the audio content on the basis of the spectrally-shaped version of the first set of spectral coefficients;
wherein the transform domain path comprises an aliasing-cancellation stimulus filter configured to filter the aliasing-cancellation stimulus signal in dependence on at least a subset of the linear-prediction-domain parameters, to derive an aliasing-cancellation synthesis signal for cancelling aliasing artifacts from the aliasing-cancellation stimulus signal; and
wherein the transform domain path also comprises a combiner configured to combine the time-domain representation of the audio content with the aliasing-cancellation synthesis signal, or a post-processed version thereof, to obtain the aliasing-reduced time-domain signal as the decoded representation of the audio content.
Claim 14 defines a corresponding "method for providing a decoded representation of an audio content". Further independent claims 13 and 15 define a related encoder and method of encoding, respectively. These claims are not relevant for the present decision. A further independent claim 16 defines a "computer program for performing the method according to claims 14 or 15". Also this claim is not relevant for the present decision.
XI. Claim 1 of Auxiliary Request 0 corresponds to claim 1 of the Main Request, but adds, as a first feature, the further definition (reference signs removed):
wherein the multi-mode audio signal decoder configured to switch between three modes, a frequency-domain mode, which uses a spectral coefficient information and a scale factor information, a transform-coded-excitation linear-prediction-domain mode, which uses a transform-coded-excitation information and a linear-prediction-domain parameter information, and an algebraic-code-excited-linear-prediction mode, which uses an algebraic-code-excitation-information and a linear-prediction-domain-parameter information.
Method claim 2 corresponds to claim 14 of the Main Request and is amended correspondingly.
Main Request - priority right
1. P1 discloses a proposed change to the working draft 4 (WD4) of the Unified Speech and Audio Coding (USAC) standard. The change consists of two modifications:
(a) Adopting forward aliasing cancellation (FAC) windows to simplify the transitions between the algebraic-code-excited linear-prediction (ACELP) frames (in the time-domain) and the transform-coding frames.
(b) Using frequency-domain noise shaping (FDNS) in the transform-coded-excitation linear-predictive-domain (TCX-LPD; also known as "weighted linear-prediction-domain" or wLPD) path to simplify the transitions between the different transform-coding frames (i.e., the frequency-domain (FD) frames and the TCX-LPD or wLPD frames).
The modifications have the effects of reducing the complexity of the codec and of reducing the over-coding that is necessary when using USAC according to the unchanged WD4.
2. The application claims priority from P1. In their decision, the Examining Division found the priority not to be valid. In the communication containing its preliminary opinion on the case, the Board found the priority to be only partly valid, for slightly different reasons.
3. The appellant considers this to be wrong and argues as follows.
(a) The aliasing-cancellation stimulus filter defined in claims 1 and 14 was a linear-predictive-coding (LPC) synthesis filter identical to the one shown in Figure 2 of P1, simply by the fact that it was configured to filter the aliasing cancelation stimulus signal in dependence on linear prediction domain parameters. Further, it was not necessary for claims 1 and 14 to include a definition of the specific processing steps for deriving the aliasing-cancellation stimulus or synthesis signal as described in D1 (to which reference is made in P1), or a definition of the specific properties of that signal. The derivation of the synthesis signal only happened in an audio encoder, whereas claims 1 and 14 defined an audio decoder.
(b) In its third paragraph, claim 1 defined that the spectrum processor applied a spectral shaping to the spectral coefficients in dependence of linear-prediction-domain (LPD) parameters. This definition implied a scaling of the spectral coefficients, which corresponded to spectral- (or "frequency-") domain noise-shaping (FDNS). The same definition also implied that the LPD parameters had to be transformed into scaling values in the frequency-domain. Hence, an additional explicit definition of an "LPC to Frequ. Dom." block and a "Scaling/FDNS" block was not necessary in claims 1 and 14. The spectral shaping carried out by the spectrum processor corresponded to the weighted LPC synthesis filtering that was carried out in WD4 of USAC in the time-domain.
(c) The algebraic-code-excited-linear-prediction (ACELP) mode, as disclosed in P1 (see Figure 2) was not essential for the audio decoder. It was apparent, to the skilled person reading P1, that alternative kinds of decoding could be used instead, with the same advantages. Hence, it was not necessary to include the definition of an ACELP block, or of an entire ACELP decoding mode, in claims 1 and 14.
(d) Also, the frequency-domain (FD) mode disclosed in P1 was not essential and did not have to be defined by claims 1 and 14. Again, it was apparent, to the skilled person, that other decoding modes could be used, and that the same advantages would apply.
Hence, the entire subject-matter of claims 1 and 14 enjoyed the priority of P1.
4. The Board finds the proprietor's arguments under points (a) and (b) persuasive. It is not necessary for claims 1 and 14 to define features that the skilled person would have recognised on the basis of its common general knowledge as not essential for the invention. This applies, in particular, to details of the synthesis signal that are not even disclosed by P1, but only vaguely referred to by reference to document D1. This also applies to known details of the techniques used in USAC. WP4 of USAC uses LPC parameters in the time-domain. Their use in the frequency domain is described, in P1, under the heading "Frequency Domain Noise Shaping (FDNS) - Application of LPC in Frequency Domain". This heading alone already expresses the correspondence between FDNS and the definitions of the spectral shaping used in claims 1 and 14. It is not necessary that claims 1 and 14 stick to the exact denotations used in Figure 2 of P1 in order to enjoy the priority, as long as the claimed subject-matter at least implicitly derives from its teaching (G 2/98, OJ EPO 2001, 413).
5. However, the appellant's arguments under points (c) and (d) are not persuasive. As explained under the above point 1., P1 proposes an improvement to WD4 of USAC. The proposed modifications and their effects are inextricably linked to the three particular decoding modes used therein (FD-mode, TCX-LPD mode, ACELP-mode), and to the transitions between respectively coded frames. Nowhere does P1 direct the skilled person to other possible decoding modes.
6. In contrast to P1, claims 1 and 14 are not restricted to the three particular decoding modes used according to the USAC standard. The claims broadly define a "multi-mode" decoding without defining (or implying) a FD decoding mode or an ACELP decoding mode. The modes of claims 1 and 14 do not even necessarily refer to decoding modes, but could also refer to, for example, suspended and active modes or high- and low-quality modes of the decoder. The skilled person would not have derived any modes different from FD and ACELP decoding modes from P1.
7. The independent claims are, therefore, to be understood as generic "OR"-claims that can be divided into two parts (see G 1/15, OJ EPO, 2017, A82, and the analysis thereof set out in the Case Law Book, 9th edition, sections II.D.5.3.2 and 5.3.3, which establishes this approach to partial priorities):
(a) The first part of each independent claim corresponds to the invention disclosed directly and unambiguously in P1 and enjoys its priority.
(b) The second part is the remaining part of the respective generic "OR"-claim that does not enjoy this priority.
8. This means that claims 1 and 14 enjoy a partial priority only (Article 87(1) EPC), covering only the subject-matter that comprises the three particular decoding modes used in USAC according to WD4.
Public availability of documents D3, D4 and D6
9. D4 and D3 are "input contribution documents" that were discussed in the 90th. and 91st meeting of the Moving Picture Coding Experts Group (MPEG) in October 2009 and January 2010, respectively. MPEG formally is the Working Group 11 (WG11) of Subcommittee 29 (SC29) of the Joint Technical Committee 1 (JTC 1) under the auspice of the International Organisation for Standardization (ISO) and the International Electrotechnical Commission (IEC). D6 is an "approved working document" from the 92nd meeting of MPEG in April 2010. D6 was not intended for publication (E13, point 6.3: "TBP: No"). E1 - E14 and D5 were cited to shed light on the public availability of D3, D4 and D6.
10. The Examining Division found that D3, D4 and D6 were not subject to confidentiality and were, therefore, publicly available at the date of filing of the application, and hence prior art in the sense of Article 54(2) EPC.
11. This is contested by the appellant, who points to E11, a document from the year 2010, describing the Moving Pictures Coding Expert Group (MPEG), operating "in the framework of the Joint ISO/IEC Technical Committee (JTC1)" (page 1, first paragraph). Further according to E11 (page 1, third paragraph), ISO standard committees were
carefully regulated by "Directives" issued by ISO/IEC and "Procedures for the Technical Work" issued by JTC1.
The input and output documents produced at the MPEG meetings were "restricted to MPEG members" (page 1, sixth paragraph and page 3, third last paragraph). E14 disclosed similar information.
12. E4 showed the directives of ISO/IEC JTC1 that were referred to in E11 in a version from the year 2007. Annex H, point 7.1 (page 135) of E4 disclosed that
while some information is publicly available, other information must be kept private to defined recipients.
An identical formulation was found in E7, point 13.1, first paragraph, and E6, point 12.1, first paragraph.
13. According to the appellant, the combined cited passages of E11 and E4 proved not only that access to the input and non-published output documents was restricted, but also that the information contained in those documents was to be kept private. Hence, the documents and their content was confidential and restricted to MPEG members only. This interpretation was supported by E3, page 36:
Input documents are all considered private to MPEG and may only be distributed outside of MPEG by the author or permission from the author in separate communication;
and E9, lines 341 - 342:
The recordings and any transcripts are the property of the meeting secretary, who is expected to respect the confidentiality of ISO meetings.
14. It followed that D3, D4 and D6 were only made available to the restricted group of MPEG members under a confidentiality agreement. The documents were, therefore, not part of the prior art.
15. These arguments are not persuasive.
16. The Board agrees with the appellant in so far as the above cited documents E3 and E11 disclose that input documents (such as D3 and D4), and working documents not intended for publication (such as D6), were "restricted to MPEG members". The restricted documents were stored on a password-protected server (E3, page 37, confirmed by E10). This does not say more than that the access to these documents was restricted, but it does not unambiguously imply an obligation to keep their content confidential. Such an obligation does also not result from the sworn statement E10.
17. In the absence of an unambiguous confidentiality agreement, it must be evaluated, whether it was realistic to expect all members of MPEG to keep the content of the restricted documents private.
18. The group of people having access to the restricted documents includes the attendees of the MPEG meeting. According to E11, second paragraph (see also E14, second paragraph), MPEG meetings were typically attended, in 2010, by
some 350 experts from some 200 companies and organisations from about 20 countries...
19. According to ISO/IEC JTC 1 regulations (E7 and E8), the password-restricted documents are electronically available "within the ISO/IEC system" (E7, item 13.1), which is an even larger number of persons than that of the meeting participants.
20. The participants of the MPEG meetings are delegates or experts that are nominated by members of the JTC1 or by representatives of "other organizations in liaison" (E9, item 6 and E4, items 126.96.36.199, together with 188.8.131.52 and 2.6.1). The nominating bodies can be companies, organisations or national institutions. ISO/IEC does not seem to impose restrictions on how these bodies perform their nominations. Also, there is no limit on the number of participants to the meetings (E9, item 6).
21. When considering, how this large and diverse group of people treated the information discussed in the MPEG meetings, it is also important to shed light on the general philosophy behind the process of standardization. E6 is a document that reflects the ISO/IEC JTC1 policy (see the foreword on page 4) and contains the following statements:
In order to make information on ISO/IEC JTC 1 and its activities available to the widest range of audiences... (E6, point 1),
... the ISO/IEC standardisation process encourages the widest possible dissemination of working documents needed for preparing standards [...] to ensure that all interested parties have the opportunity to contribute to the development of a standard (E6, point 12.1; see also E7, point 13.1).
22. In view of the above observations, the Board concludes that it was not realistic to expect that the content of D3, D4 and D6 was kept confidential, considering the inclusive and transparent process of standardisation, the large and diverse number of experts, companies and organisations involved, and the lack of a binding confidentiality agreement. If a strict and binding confidentiality of the content of the restricted documents would have been intended, a - possibly signed - agreement or regulation for those persons with access to the restricted documents would have been expected. However, none of the cited documents, which, in parts, go to considerable lengths to lay down different aspects of the JTC1 working group, discloses such an agreement or regulation.
23. The Board, therefore, considers that the content of documents D3, D4 and D6 was not subject to a confidentiality agreement, and that their content was publicly available before the filing (but after the priority date) of the present application. Hence, these documents form part of the prior art in the sense of Article 54(2) EPC for the claims of the Main Request.
24. Attention is drawn to the decision T 2239/15 from the same Board (catchword and items 8 - 34). This decision addresses in detail the question of public availability of MPEG input documents. Based on even broader evidence than in the present case, the Board came to the same conclusion that an obligation of confidentiality was not in place regarding the content of such input documents.
Main Request - inventive step
25. D3, D4 and D6 form part of the prior art within the meaning of Article 54(2) EPC for the part of the subject-matter of claims 1 and 14 that does not enjoy the priority of P1. Namely, for the part that does not include FD and ACELP decoding modes. It is this part only that will be referred to in the following.
26. D3, D4 and D6, whose author(s) are affiliated with the appellant, disclose the exact same subject-matter as P1 and only differ from the latter in the amount of detail. Hence, claim 1 differs from those documents in that the decoder comprises modes (which are not necessarily decoding modes; see above point 6.) that are not FD and ACELP decoding modes.
27. There is no apparent technical effect that would apply to all modes within the broad scope of claim 1. In the absence of a technical effect, however, no inventive step can be recognized.
28. The appellant argued that the technical effect of using FDNS and aliasing-cancellation was not restricted to the three particular decoding modes used in USAC. The skilled person would have understood that also other decoding modes could be used with the same benefit.
29. This argument is not pertinent, because the technical effect relevant for assessing inventive step must be the effect resulting from the distinguishing features. It is not FDNS and aliasing-cancellation that distinguish claim 1 from D3, D4 or D6. Rather, claim 1 differs from these documents in that the additional modes employed by the multi-mode decoder are not FD and ACELP modes but some undefined other modes. Due to the undefined and varied possible nature of these modes, there is no technical effect that would apply to all of them.
30. Hence, the part of the subject-matter of claim 1 that does not enjoy the priority of P1 does not involve an inventive step over each of D3, D4 and D6 (Articles 52(1) and 56 EPC). The same holds for the subject-matter of claim 14.
31. If, for the sake of partly following the appellant, only those modes were considered that the skilled person would have considered as alternatives for the FD and ACELP modes used in USAC (see point 28.), then the difference might have the technical effect of finding alternative decoders that also profited from the same advantages when using FDNS and aliasing-cancellation. However, if the alternative decoding modes would have been considered by the skilled person in view of claim 1, then the skilled person would have also considered them as obvious alternatives to the FD and ACELP decoding modes when starting from D3, D4 or D6. Hence, even if only considering such kinds of decoding modes, this part of the subject-matter of claims 1 and 14 would not involve an inventive step over each of D3, D4 and D6.
Auxiliary Request 0 - consideration
32. In its provisional opinion, the Board found that claims 1 and 14 each enjoyed a partial priority. This assessment differed from the finding of the Examining Division, and the reasons given by the Board in this respect were new to the appellant. The Board, therefore, acknowledges exceptional circumstances in the sense of Article 13(2) RPBA 2020.
33. Auxiliary Request 0 is an immediate reaction to the Board's provisional finding that could not have been filed earlier. Since the claims of this request, prima facie, overcome the inventive-step objection resulting from the partial priority of the claims of the Main Request (Article 13(1) RPBA), the Board does consider Auxiliary Request 0.
Auxiliary Request 0 - priority right
34. Claims 1 and 2 of Auxiliary Request 0 are based on claims 1 and 14 of the Main Request, and, additionally, define the three decoding modes used by USAC: a FD mode, a TCX-LPD mode and an ACELP mode.
35. Thereby, the subject-matter of claims 1 and 2 is restricted to subject-matter that enjoys priority of P1 (see the above section on priority of the Main Request).
36. It follows, that claims 1 and 2 fully enjoy the priority of P1 (Article 87(1) EPC).
Auxiliary Request 0 - added subject-matter
37. The features added to claims 1 and 2 with respect to claims 1 and 14 of the Main Request have a basis on page 4, lines 5 - 9 (three modes), on page 4, lines 31 - 35 (TCX-LPD and FD modes), and on page 5, lines 25 - 29 (TCX-LPD and ACELP modes). A further basis is found on page 69, line 30 - page 70, line 34, as well as in original claims 3 and 4.
38. In their communication dated 8 August 2016, the Examining Division found that the feature of deriving an aliasing-cancellation synthesis signal was not originally disclosed together with the purpose "for cancelling aliasing artifacts". The Board does not subscribe to this objection. First, the skilled person understands that an aliasing-cancellation synthesis signal is there precisely to cancel aliasing artifacts. Second, the last feature of original claim 14 even provides a literal basis for this feature, considering that the application implies that claim 14 defines the same method that is to be carried out by the decoder of claim 1.
39. Hence, the claims of Auxiliary Request 0 have a basis in the application as filed (Article 123(2) EPC).
40. The appellant requests "to directly issue an Official Communication under Rule 71(3) EPC or to order the Examining Division to issue such a Communication, without further examination", meaning, without referring the case back for further prosecution.
41. The Examining Division based its decision on the finding that the priority of the claims was not valid. It followed that the effective filing date of the claims was later than the dates at which D3, D4 and D6 were made available to the public. As a consequence, D3, D4 and D6 were part of the state of the art according to Article 54(2) EPC, and the assessment of patentability was made exclusively on the basis of these documents.
42. The Examining Division did not assess patentability on the basis of other documents. It did not comment on the respective assessment in the Written Opinion of the International Search Authority (WOISA).
43. The Board asserted that the claims of Auxiliary Request 0 enjoyed the priority of P1 (see point 36.). Hence, D3, D4 and D6 are not part of the prior art in the sense of Article 54(2) or (3) EPC. Whether the remaining two prior art documents D1 and D2 on file are sufficient to fully examine patentability is a question, which the Board cannot answer without performing its own, extensive, investigation.
44. The Board finds itself, therefore, in a situation in which it would have to consider aspects which would lie entirely outside the scope of the appealed decision. Considering the circumstances of the present case, issuing a decision without referring the case back for further prosecution would go beyond the primary object of the appeal proceedings as defined in Article 12(2) RPBA 2020. The Board considers this as a special reason that prevents it from deciding on the case on its own.
45. Hence, the case is remitted to the Examining Division for further prosecution (Article 11 RPBA 2020 and Article 111(1) EPC).
For these reasons it is decided that:
1. The decision under appeal is set aside.
2. The case is remitted to the Examining Division for further prosecution.