Abstract: Images of an actual rendering environment are acquired through image sensors operating in conjunction with a media consumption system. The acquired images of the actual rendering environment are used to predict audio characteristics of objects present in the actual rendering environment. Spatial audio rendered, to a user in the actual rendering environment, by audio speakers operating in conjunction with the media consumption system is adjusted or modified based at least in part on the audio characteristics of the objects present in the actual rendering environment.

Abstract: In some embodiments, a pitch filter for filtering a preliminary audio signal generated from an audio bitstream is disclosed. The pitch filter has an operating mode selected from one of either: (i) an active mode where the preliminary audio signal is filtered using filtering information to obtain a filtered audio signal, and (ii) an inactive mode where the pitch filter is disabled. The preliminary audio signal is generated in an audio encoder or audio decoder having a coding mode selected from at least two distinct coding modes, and the pitch filter is capable of being selectively operated in either the active mode or the inactive mode while operating in the coding mode based on control information.

Abstract: Novel methods and systems for compensating for ambient light around displays are disclosed. A shift in the PQ curve applied to an image can compensate for sub-optimal ambient light conditions for a display, with the PQ shift being either an addition to a compensation value in PQ space followed by a subtraction of the compensation value in linear space, or an addition to the compensation value in linear space followed by a subtraction of the compensation value in PQ space. Further adjustments to the PQ curve can also be made to provide an improved image quality with respect to image luminance.

Abstract: One or more media contents are received. A viewer's light adaptive states are predicted as a function of time as if the viewer is watching display mapped images derived from the one or more media contents. The viewer's light adaptive states are used to detect an excessive change in luminance in a specific media content portion of the one or more media contents. The excessive change in luminance in the specific media content portion of the one or more media contents is caused to be reduced while the viewer is watching one or more corresponding display mapped images derived from the specific media content portion of the one or more media contents.

Abstract: At a first time point, a first light capturing device at a first spatial location in a three-dimensional (3D) space captures first light rays from light sources located at designated spatial locations on a viewer device in the 3D space. At the first time point, a second light capturing device at a second spatial location in the 3D space captures second light rays from the light sources located at the designated spatial locations on the viewer device in the 3D space. Based on the first light rays captured by the first light capturing device and the second light rays captured by the second light capturing device, at least one of a spatial position and a spatial direction, at the first time point, of the viewer device is determined.

Abstract: The present disclosure provides methods, devices and computer program products for encoding and decoding of a vector of parameters in an audio coding system. The disclosure further relates to a method and apparatus for reconstructing an audio object in an audio decoding system. According to the disclosure, a modulo differential approach for coding and encoding a vector of a non-periodic quantity may improve the coding efficiency and provide encoders and decoders with less memory requirements. Moreover, an efficient method for encoding and decoding a sparse matrix is provided.

Abstract: A novel projection system includes a light source, a phase modulator, an amplitude modulator, and a controller having temporal lightfield simulation capabilities. The phase modulator spatially modulates a lightfield from the light source to generate an intermediate image on the amplitude modulator. The amplitude modulator spatially modulates the intermediate image to form a final image. The controller models the phase state of the phase modulator during transitions between phase modulator frames and generates lightfield simulations of the intermediate image during the transition. The controller utilizes the lightfield simulations to generate and provide sets of amplitude drive values to the amplitude modulator at a faster rate than that at which the phase modulator is capable of switching.

Abstract: An audio processing system (100) accepts an audio bitstream having one of a plurality of predefined audio frame rates. The system comprises a front-end component (110), which receives a variable number of quantized spectral components, corresponding to one audio frame in any of the predefined audio frame rates, and performs an inverse quantization according to predetermined, frequency-dependent quantization levels. The front-end component may be agnostic of the audio frame rate. The audio processing system further comprises a frequency-domain processing stage (120) and a sample rate converter (130), which provide a reconstructed audio signal sampled at a target sampling frequency independent of the audio frame rate. By its frame-rate adaptability, the system can be configured to operate frame-synchronously in parallel with a video processing system that accepts plural video frame rates.

Abstract: Example embodiments disclosed herein relate to a transducer assembly and associated signal processing. A transducer assembly includes two voice coils in a telescopic arrangement and having unequal sizes, and two suspension systems connected to the two voice coils, respectively. The two voice coils extend in opposites directions from their suspension systems. Dimensions of respective wires of the two voice coils are determined based on respective magnetic flux densities in magnetic gaps for receiving the two voice coils. As a result, a residual vibration caused by the unequal-sized voice coils can be further reduced.

Abstract: Described are methods of processing an audio signal for packet loss concealment. The audio signal comprises a sequence of frames, each frame containing representations of a plurality of audio channels and reconstruction parameters for upmixing the plurality of audio channels to a predetermined channel format. One method includes: receiving the audio signal; and generating a reconstructed audio signal in the predefined channel format based on the received audio signal. Generating the reconstructed audio signal comprises: determining whether at least one frame of the audio signal has been lost; and if a number of consecutively lost frames exceeds a first threshold, fading the reconstructed audio signal to a predefined spatial configuration. Also described is a method of encoding an audio signal. Yet further described are apparatus for carrying out the methods, as well as corresponding programs and computer-readable storage media.

Abstract: Described is a method of performing automatic audio enhancement on an input audio signal including at least one speech-articulation noise event. The method comprises: segmenting the input audio signal into a number of audio frames; obtaining at least one feature parameter from the audio frames; and determining, based at least in part on the obtained feature parameter, a respective type of the speech-articulation noise event and a respective time-frequency range associated with the speech-articulation noise event within the input audio signal.

Abstract: Multiple virtual source locations may be defined for a volume within which audio objects can move. A set-up process for rendering audio data may involve receiving reproduction speaker location data and pre-computing gain values for each of the virtual sources according to the reproduction speaker location data and each virtual source location. The gain values may be stored and used during “run time,” during which audio reproduction data are rendered for the speakers of the reproduction environment. During run time, for each audio object, contributions from virtual source locations within an area or volume defined by the audio object position data and the audio object size data may be computed. A set of gain values for each output channel of the reproduction environment may be computed based, at least in part, on the computed contributions. Each output channel may correspond to at least one reproduction speaker of the reproduction environment.

Abstract: Audio objects are associated with positional metadata. A received downmix signal comprises downmix channels that are linear combinations of one or more audio objects and are associated with respective positional locators. In a first aspect, the downmix signal, the positional metadata and frequency-dependent object gains are received. An audio object is reconstructed by applying the object gain to an upmix of the downmix signal in accordance with coefficients based on the positional metadata and the positional locators. In a second aspect, audio objects have been encoded together with at least one bed channel positioned at a positional locator of a corresponding downmix channel. The decoding system receives the downmix signal and the positional metadata of the audio objects. A bed channel is reconstructed by suppressing the content representing audio objects from the corresponding downmix channel on the basis of the positional locator of the corresponding downmix channel.

Abstract: A method of noise reduction includes using a neural network to control a Wiener filter. The gains estimated by the neural network are combined with the gains produced by the Wiener filter. In this manner, the noise reduction system provides improved results as compared to using only a neural network.

Abstract: The present disclosure relates to a method of decoding audio scene content from a bitstream by a decoder that includes an audio renderer with one or more rendering tools.

Abstract: Creative intent input describing emotion expectations and narrative information relating to media content is received. Expected physiologically observable states relating to the media content are generated based on the creative intent input. An audiovisual content signal with the media content and media metadata comprising the physiologically observable states is provided to a playback apparatus. The audiovisual content signal causes the playback device to use physiological monitoring signals to determine, with respect to a viewer, assessed physiologically observable states relating to the media content and generate, based on the expected physiologically observable states and the assessed physiologically observable states, modified media content to be rendered to the viewer.

Abstract: Methods and systems for designing binaural room impulse responses (BRIRs) for use in headphone virtualizers, and methods and systems for generating a binaural signal in response to a set of channels of a multi-channel audio signal, including by applying a BRIR to each channel of the set, thereby generating filtered signals, and combining the filtered signals to generate the binaural signal, where each BRIR has been designed in accordance with an embodiment of the design method. Other aspects are audio processing units configured to perform any embodiment of the inventive method. In accordance with some embodiments, BRIR design is formulated as a numerical optimization problem based on a simulation model (which generates candidate BRIRs) and at least one objective function (which evaluates each candidate BRIR), and includes identification of a best one of the candidate BRIRs as indicated by performance metrics determined for the candidate BRIRs by each objective function.

Abstract: Some implementations involve analyzing audio packets received during a time interval that corresponds with a conversation analysis segment to determine network jitter dynamics data and conversational interactivity data. The network jitter dynamics data may provide an indication of jitter in a network that relays the audio data packets. The conversational interactivity data may provide an indication of interactivity between participants of a conversation represented by the audio data. A jitter buffer size may be controlled according to the network jitter dynamics data and the conversational interactivity data. The time interval may include a plurality of talkspurts.

Abstract: Given input HDR and SDR images representing the same scene, a prediction model to predict the HDR image from a compressed representation of the input SDR image is generated as follows: a) generate noise data based at least on the characteristics of the HDR image b) generate a noisy SDR image by adding the noise data to the SDR image c) generate an augmented HDR data set and an augmented SDR data set by using the input HDR and SDR images and the noisy SDR image d) generate a prediction model to predict the augmented HDR data set based on the augmented SDR data set and e) solve the prediction model according to a minimization-error criterion to generate a set of prediction parameters to be transmitted to a decoder together with a compressed representation of the input SDR image to reconstruct an approximation of the input HDR image.

Abstract: Images of a user’s head are acquired at a plurality of different orientational angles through image sensors operating in conjunction with a media consumption system. The acquired images of the user’s head are used to select or predict a specific personalized head related transfer function for the user. Spatial audio rendered by audio speakers operating in conjunction with the media consumption system is adjusted or modified based at least in part on the specific personalized HRTF selected for the user.

Abstract: In an embodiment, a spatio-level filter (SLF) is created by obtaining a first set of samples from a plurality of target source level and spatial distributions in frequency subbands in a frequency domain, obtaining a second set of samples from a plurality of background level and spatial distributions in frequency subbands in a frequency domain, adding the first and second sets of samples to create a combined set of samples, detecting level and spatial parameters for each sample in the combined set of samples for each subband, within subbands, weighting the detected level and spatial parameters by their respective level and spatial distributions for the target source and backgrounds; storing the weighted level, spatial parameters and signal-to-noise ratio (SNR) within subbands for each sample in the combined set of samples in a table; and re-indexing the table by the weighted level and spatial parameters and subband.

Abstract: In an embodiment, a method comprises: transforming one or more frames of a two-channel time domain audio signal into a time-frequency domain representation including a plurality of time-frequency tiles, wherein the frequency domain of the time-frequency domain representation includes a plurality of frequency bins grouped into subbands. For each time-frequency tile, the method comprises: calculating spatial parameters and a level for the time-frequency tile; modifying the spatial parameters using shift and squeeze parameters; obtaining a softmask value for each frequency bin using the modified spatial parameters, the level and subband information; and applying the softmask values to the time-frequency tile to generate a modified time-frequency tile of an estimated audio source.

Abstract: The present document relates to audio coding systems which make use of a harmonic transposition method for high frequency reconstruction (HFR), and to digital effect processors, e.g. so-called exciters, where generation of harmonic distortion adds brightness to the processed signal. In particular, a system configured to generate a high frequency component of a signal from a low frequency component of the signal is described. The system may comprise an analysis filter bank (501) configured to provide a set of analysis subband signals from the low frequency component of the signal; wherein the set of analysis subband signals comprises at least two analysis subband signals; wherein the analysis filter bank (501) has a frequency resolution of ?f.

Abstract: An audio bitstream is decoded into audio objects and audio metadata for the audio objects. The audio objects include a specific audio object. The audio metadata specifies frame-level gains that include a first gain and a second gain respectively for a first audio frame and a second audio frame. It is determined, based on the first and second gains, whether sub-frame gains are to be generated for the specific audio object. If so, a ramp length is determined for a ramp used to generate the sub-frame gains for the specific audio object. The ramp of the ramp length is used to generate the sub-frame gains for the specific audio object. A sound field represented by the audio objects with the sub-frame gains is rendered by audio speakers.

Abstract: Described is a method of training a neural-network-based system for determining an indication of an audio quality of an audio input. The method includes obtaining, as input, at least one training set comprising audio samples. The audio samples include audio samples of a first type and audio samples of a second type, wherein each of the first type of audio samples is labelled with information indicative of a respective predetermined audio quality metric, and wherein each of the second type of audio samples is labelled with information indicative of a respective audio quality metric relative to that of a reference audio sample. The method further includes: inputting the training set to the neural-network-based system; and iteratively training the system to predict the respective label information of the audio samples in the training set.

Abstract: The present disclosure provides methods, devices and computer program products for encoding and decoding a stereo audio signal based on an input signal. According to the disclosure, a hybrid approach of using both parametric stereo coding and a discrete representation of the stereo audio signal is used which may improve the quality of the encoded and decoded audio for certain bitrates.

Abstract: One or more perforation hole pattern methods are applied (402) to generate a spatial distribution of perforation holes forming a semi-random pattern for an image display screen. The image display screen is perforated (404) with the spatial distribution of perforation holes forming the semi-random pattern. Image rendering light is emitted (406) with a light projector toward the image display screen that is installed in an image rendering environment. At least a portion of the image rendering light emitted from the light projector is reflected (408) by the image display screen, toward a viewer.

Abstract: Described herein is a method for improving dialogue intelligibility during playback of audio data on a playback device, wherein the audio data comprise dialogue audio data, and at least one of music and effects audio data, the method including the steps of: determining a volume mixing ratio based on a volume value for playback; mixing the dialogue audio data and the at least one of music and effects audio data based on said volume mixing ratio; and outputting the mixed audio data for playback. Described are further a respective playback device and a respective computer program product.

Abstract: The present document relates an audio encoding and decoding system (referred to as an audio codec system). In particular, the present document relates to a audio codec system which is particularly well suited for voice encoding/decoding. A transform-based speech encoder is configured to encode a speech signal into a bitstream is described. A speech decoder configured to decode audio signals from a bitstream is further described.

Abstract: Methods and audio processing units for generating an object based audio program including conditional rendering metadata corresponding to at least one object channel of the program, where the conditional rendering metadata is indicative of at least one rendering constraint, based on playback speaker array configuration, which applies to each corresponding object channel, and methods for rendering audio content determined by such a program, including by rendering content of at least one audio channel of the program in a manner compliant with each applicable rendering constraint in response to at least some of the conditional rendering metadata. Rendering of a selected mix of content of the program may provide an immersive experience.

Abstract: The present document relates to audio source coding systems which make use of a harmonic transposition method for high frequency reconstruction (HFR), as well as to digital effect processors, e.g. exciters, where generation of harmonic distortion add brightness to the processed signal, and to time stretchers where a signal duration is prolonged with maintained spectral content. A system and method configured to generate a time stretched and/or frequency transposed signal from an input signal is described. The system comprises an analysis filterbank configured to provide an analysis subband signal from the input signal; wherein the analysis subband signal comprises a plurality of complex valued analysis samples, each having a phase and a magnitude. Furthermore, the system comprises a subband processing unit configured to determine a synthesis subband signal from the analysis subband signal using a subband transposition factor Q and a subband stretch factor S.

Abstract: Embodiments are disclosed for non-intrusive transducer health detection in an audio system. In an embodiment, a method performed by the audio system comprises outputting one or more encoded inaudible acoustic signals into an acoustic transmission medium using a first transducer. The one or more encoded inaudible acoustic signals are received from the acoustic transmission medium using a second transducer of the audio system. The received one or more encoded inaudible acoustic signals are used to identify failure or degradation of the first or second transducer.

Abstract: A method for distributing High Dynamic Range (HDR) content to playback devices for displaying images where the HDR content is encoded to an HDR bitstream and the HDR bitstream is subsequently decoded by a playback device. The HDR bitstream contains auxiliary metadata packets that are based upon the processing capability of the playback device.

Abstract: Described herein is a method of determining parameters for a generative neural network for processing an audio signal, wherein the generative neural network includes an encoder stage mapping to a coded feature space and a decoder stage, each stage including a plurality of convolutional layers with one or more weight coefficients, the method comprising a plurality of cycles with sequential processes of: pruning the weight coefficients of either or both stages based on pruning control information, the pruning control information determining the number of weight coefficients that are pruned for respective convolutional layers; training the pruned generative neural network based on a set of training data; determining a loss for the trained and pruned generative neural network based on a loss function; and determining updated pruning control information based on the determined loss and a target loss. Further described are corresponding apparatus, programs, and computer-readable storage media.

Abstract: Some methods involve receiving an input audio signal that includes N input audio channels, the input audio signal representing a first soundfield format having a first soundfield format resolution, N being an integer ?2. A first decorrelation process may be applied to two or more of the input audio channels to produce a first set of decorrelated channels, the first decorrelation process maintaining an inter-channel correlation of the set of input audio channels. A first modulation process may be applied to the first set of decorrelated channels to produce a first set of decorrelated and modulated output channels. The first set of decorrelated and modulated output channels may be combined with two or more undecorrelated output channels to produce an output audio signal that includes O output audio channels representing a second and relatively higher-resolution soundfield format than the first soundfield format, O being an integer ?3.

Abstract: Embodiments are disclosed for projection systems with rotatable anamorphic lenses. In an embodiment, an optical projection system comprises: a light source; an optical integrator configured to receive light from the light source and to distribute a uniform pattern of light; a relay lens system including two or more rotatable anamorphic lenses, the anamorphic lenses oriented about an optical axis to transform the uniform pattern of light into an image having a specified aspect ratio; at least one spatial light modulator configured to receive the image and direct a spatially modulated image along an optical path; and at least one projection lens configured to receive the spatially modulated image from the optical path and to project the spatially modulated image onto an image plane with the specified aspect ratio.

Abstract: A content-creation tool includes a processor and a memory. The processor is configured to receive a first video clip and a second video clip, a respective first and second metadata-item thereof being set to a respective first and second metadata-value. The memory stores video-editing software that includes a timeline interface and instructions that, when executed by the processor, control the processor to: add the first video clip to the timeline interface as a first timeline-track that retains the first metadata-value; add the second video clip to the timeline interface as a second timeline-track that retains the second metadata-value; and generate a frame sequence that includes a plurality of video frames. Each video frame is a frame of, or a frame derived from, one of (i) the first timeline-track, (ii) the second timeline-track, and (iii) a composited time-line-track composited from at least one of the first and second timeline-tracks.

Abstract: A method comprises: obtaining softmask values for frequency bins of time-frequency tiles representing an audio signal; reducing, or expanding and limiting, the softmask values; and applying the reduced, or expanded and limited, softmask values to the frequency bins to create a time-frequency representation of an estimated target source. An alternative method comprises, for each time-frequency tile: obtaining softmask values; applying the softmask values to the frequency bins to create a time-frequency domain representation of an estimated target source; obtaining a panning parameter and a source concentration estimates for the target source; determining, using the panning parameter estimate and the softmask values, a magnitude for the time-frequency representation of the estimated target source; determining, using the panning parameter estimate and the source phase concentration estimate, a phase for the time-frequency representation of the estimated target source; and combining the magnitude and the phase.

Abstract: A system and method of editing video content includes receiving input video data; converting the input video data to a predetermined format; generating a plurality of initial metadata values for a frame of the converted video data, the plurality of initial metadata values including a first metadata value corresponding to a first fixed value not calculated from a content including the frame, a second metadata value corresponding to an average luminance value of the frame, and a third metadata value corresponding to a second fixed value not calculated from the content, wherein the first meta-data value, the second metadata value, and the third metadata value include information used by a decoder to render a decoded image on a display.

Abstract: A computer-implemented method of audio processing, the method comprising: receiving audio object data and audio description data, wherein the audio object data includes a first plurality of audio objects; calculating a long-term loudness of the audio object data and a long- term loudness of the audio description data; calculating a plurality of short-term loudnesses of the audio object data and a plurality of short-term loudnesses of the audio description data; reading a first plurality of mixing parameters that correspond to the audio object data; generating a second plurality of mixing parameters based on the first plurality of mixing parameters, the long-term loudness of the audio object data, the long-term loudness of the audio description data, the plurality of short-term loudnesses of the audio object data, and the plurality of short-term loudnesses of the audio description data; generating a gain adjustment visualization corresponding to the second plurality of mixing parameters, the audio object data

Abstract: In some embodiments, a method for performing at least one of enhancement, decoding, or rendering of a multichannel audio signal in response to compression feedback or feedback from a smart amplifier. For example, the compression feedback may be indicative of amount of compression applied to each of multiple frequency bands, of the audio signal or an enhanced audio signal generated in response thereto. The enhancement (e.g., bass enhancement) may include dynamic routing of audio content of the input audio signal between channels of an enhanced audio signal generated in response thereto. The enhancement and compression may be performed on a per speaker class basis. Other aspects are systems (e.g., programmed processors) and devices (e.g., devices having physically-limited bass reproduction capabilities, such as, for example, a notebook or laptop computer, tablet, soundbar, mobile phone, or other device with small speakers) configured to perform any embodiment of the method.

Abstract: Occluded image fragments are sorted in size. The largest image fragment is used to size a quadtree node in a layout mask for a disocclusion atlas used to store the image fragments. The sorted image fragments are stored into the disocclusion atlas using the layout mask such as each image fragment is hosted with a best fit quadtree node in the disocclusion atlas. A video signal may be generated by encoding one or more reference images and the disocclusion atlas storing the image fragments. The image fragments can be used by a recipient device to fill disoccluded image data in disoccluded spatial regions in a display image synthesized from the reference images.

Abstract: An encoding device evaluates a plurality of processing and/or post-processing algorithms and/or methods to be applied to a video stream, and signals a selected method, algorithm, class or category of methods/algorithms either in an encoded bitstream or as side information related to the encoded bitstream. A decoding device or post-processor utilizes the signaled algorithm or selects an algorithm/method based on the signaled method or algorithm. The selection is based, for example, on availability of the algorithm/method at the decoder/post-processor and/or cost of implementation. The video stream may comprise, for example, downsampled multiplexed stereoscopic images and the selected algorithm may include any of upconversion and/or error correction techniques that contribute to a restoration of the downsampled images.

Abstract: An encoding device evaluates a plurality of processing and/or post-processing algorithms and/or methods to be applied to a video stream, and signals a selected method, algorithm, class or category of methods/algorithms either in an encoded bitstream or as side information related to the encoded bitstream. A decoding device or post-processor utilizes the signaled algorithm or selects an algorithm/method based on the signaled method or algorithm. The selection is based, for example, on availability of the algorithm/method at the decoder/post-processor and/or cost of implementation. The video stream may comprise, for example, downsampled multiplexed stereoscopic images and the selected algorithm may include any of upconversion and/or error correction techniques that contribute to a restoration of the downsampled images.

Abstract: Projection displays include a highlight projector and a main projector. Highlights projected by the highlight projector boost luminance in highlight areas of a base image projected by the main projector. Various highlight projectors including steerable beams, holographic projectors and spatial light modulators are described.

Abstract: The described embodiments include systems and methods for producing and adapting images, such as video images, for presentation on display devices that have various different aspects ratios, such as 4:3, 16:9, 9:16, etc. In one embodiment, a method for producing content, such as video images, can begin by selecting an original aspect ratio and determining, within at least a first scene in the content, a position of a subject in the first scene. In one embodiment, the original aspect ratio can be substantially square (e.g., 1:1). Metadata can then be created, based on the position of the subject in the first scene, to guide playback devices to asymmetrically crop the content, relative to the position, for display on display devices that have aspect ratios that are different than the original aspect ratio. Other methods and systems are also described.

Abstract: Saliency regions are identified in a global scene depicted by volumetric video. Saliency video streams that track the saliency regions are generated. Each saliency video stream tracks a respective saliency region. A saliency stream based representation of the volumetric video is generated to include the saliency video streams. The saliency stream based representation of the volumetric video is transmitted to a video streaming client.

Abstract: Methods and systems for performing at least one audio activity (e.g., conducting a phone call or playing music or other audio content) in an environment including by determining an estimated location of a user in the environment in response to sound uttered by the user (e.g., a voice command), and controlling the audio activity in response to determining the estimated user location. The environment may have zones which are indicated by a zone map and estimation of the user location may include estimating in which of the zones the user is located. The audio activity may be performed using microphones and loudspeakers which are implemented in or coupled to smart audio devices.

Abstract: A method of audio processing includes generating harmonics in a hybrid complex quadrature mirror filter domain. Generating the harmonics may include multiplication, using a feedback delay loop, and dynamic compression. The harmonics may be generated based on one or more hybrid sub-bands of the complex transform domain signal.

Abstract: Systems, methods, and computer program products are disclosed for adaptive downmixing of audio signals with improved continuity. An audio encoding system receives an input multi-channel audio signal including a primary input audio channel and L non-primary input audio channels. The system determines a set of L input gains. For each of the channels and gains, the system forms a respective scaled non-primary input audio channel. The system forms a primary output audio channel from the sum of the primary input audio channel and the scaled non-primary input audio channels. The system determines a set of L prediction gains. The system forms a prediction channel from the primary output audio channel. The system forms L non-primary output audio channels. The system forms an output multi-channel audio signal from the primary output audio channel and the L non-primary output audio channels.