# US 10,504,536 B2 · Audio quality in real-time communications over a network > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,504,536 B2 - Document ID: US10504536B2 - Status: Granted - Kind: B2 - Application number: 15/827,387 - Filed: 2017-11-30 - Published: 2019-05-30 - Granted: 2019-12-10 - Earliest claimed: 2017-11-30 - Assignees: LogMeIn, Inc. - Applicants: GetGo, Inc. - Inventors: Bjorn Volcker; Matthieu Hodgkinson - Web page: https://patentgrep.com/patent/US10504536B2 ## CPC - G10L 21/0364: for improving intelligibility - G10L 21/0232: Processing in the frequency domain - G10L 25/60: for measuring the quality of voice signals - G10L 25/69: for evaluating synthetic or decoded voice signals - H04L 12/1827: {Network arrangements for conference optimisation or adaptation} - H04L 65/80: Responding to QoS - H04M 3/2236: {Quality of speech transmission monitoring} - H04W 72/542: using measured or perceived quality - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) - H04M: TELEPHONIC COMMUNICATION (circuits for controlling other apparatus via a telephone cable and not involving telephone switching apparatus G08) - H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72) ## Abstract A technique for estimating and enhancing audio quality in a real-time communication session between parties over a computer network produces real-time measurements of factors that are known to impact audio quality, assigns a separate MOS value to each of the measured factors, and combines the MOS values for the various factors to produce an overall measure of audio quality. At least one party to the real-time communication session operates a computing device that runs a software program, and the technique further includes directing the software program to render an indication of the overall audio quality, thereby enabling the party operating the computing device to take remedial action to improve the audio quality. ## Claims ### Claim 1 A method of improving audio quality in real-time communications over a computer network, the method comprising: generating, during a real-time communication session over the computer network between multiple parties operating respective computing devices, real-time measurements of multiple audio factors of the communication session, each of the audio factors being susceptible to degradation in a way that affects an overall audio quality of the communication session; assigning a respective MOS (Mean Opinion Score) value to each of the measured audio factors, each MOS value providing a measure of quality of the respective audio factor independent of other audio factors; combining the assigned MOS values to produce an overall measure of audio quality; and directing a software program running on one of the computing devices to render an indication of the overall measure of audio quality, the indication enabling the party operating said one of the computing devices to take remedial action to improve the overall audio quality. ### Claim 2 (depends on 1) The method of claim 1, wherein combining the assigned MOS values is performed by a server apparatus coupled to each of the computing devices over the computer network, and wherein the method further comprises the server apparatus directing the software program to display a message on said one of the computing devices, the message providing a recommendation of a particular remedial action for improving the overall audio quality. ### Claim 3 (depends on 1) The method of claim 1, wherein combining the assigned MOS values is performed by a server apparatus coupled to each of the computing devices over the computer network, and wherein the method further comprises the server apparatus directing the software program to change a setting to improve the overall audio quality. ### Claim 4 (depends on 3) The method of claim 3, wherein the method further comprises the server apparatus: generating a per-party audio quality measure for each party to the real-time communication session, and directing the software program to render a respective indication of the per-party audio quality measure for each party to the real-time communication session, such that the party operating the respective computing device is enabled to see, via the software program, the respective indication of the per-party audio quality measure of each other party participating in the real-time communication session. ### Claim 5 (depends on 3) The method of claim 3, wherein assigning a respective MOS value to each of the measured audio factors includes: receiving the measured audio factor as a raw value; and converting the raw value to the respective MOS value. ### Claim 6 (depends on 5) The method of claim 5, wherein converting the raw value to the respective MOS value includes, for at least one measured audio factor: performing prior off-line testing on human subjects, the human subjects assigning subjective quality scores to audio at each of multiple raw value levels of the audio factor, each of the multiple raw value levels corresponding to a respective level of audio degradation in the audio factor with all other audio factors held constant; computing a mean quality score across the human subjects for each of the multiple raw value levels; constructing a mapping that associates raw value levels with corresponding mean quality scores; and applying the mapping to convert the raw value to the respective MOS value. ### Claim 7 (depends on 5) The method of claim 5, wherein converting the raw value to the respective MOS value includes, for at least one measured audio factor: performing prior off-line characterization of multiple raw value levels of the audio factor using VQT (Voice Quality Testing), the characterization using VQT providing a mapping between the raw value levels and corresponding MOS values; and applying the mapping to convert the raw value to the respective MOS value. ### Claim 8 (depends on 5) The method of claim 5, wherein combining the assigned MOS values to produce the overall measure of audio quality includes converting each of the assigned MOS values to a corresponding factor probability, each factor probability indicating a probability that the respective MOS value indicates bad audio. ### Claim 9 (depends on 8) The method of claim 8, wherein the method further comprises identifying a reference quantile, the reference quantile representing a particular percentile opinion score that subjective human listeners have assigned to un-degraded reference audio, and wherein converting each of the assigned MOS values to a respective factor probability includes, for each of the assigned MOS values: constructing a distribution of quality values whose mean equals the assigned MOS value; computing a cumulative probability value as a summation of the constructed distribution of quality values up to the reference quantile; and providing the cumulative probability value as the factor probability. ### Claim 10 (depends on 9) The method of claim 9, wherein combining the assigned MOS values to produce the overall measure of audio quality further includes combining the factor probabilities to generate an overall probability of bad audio. ### Claim 11 (depends on 10) The method of claim 10, wherein combining the factor probabilities includes computing the overall probability of bad audio as a product of the factor probabilities divided by a sum of (i) the product of the factor probabilities and (ii) a product of complements of the factor probabilities. ### Claim 12 (depends on 10) The method of claim 10, wherein combining the assigned MOS values to produce the overall measure of audio quality further includes converting the overall probability of bad audio to an overall MOS value. ### Claim 13 (depends on 12) The method of claim 12, wherein converting the overall probability of bad audio to the overall MOS value includes: constructing an overall distribution whose CDF (cumulative distribution function) maps the overall probability of bad audio to the reference quantile; and providing the overall MOS value as a mean of the overall distribution. ### Claim 14 (depends on 5) The method of claim 5, wherein at least one of the multiple audio factors includes multiple sub-factors having different causes but contributing to a common perceptual effect. ### Claim 15 (depends on 14) The method of claim 14, wherein the multiple sub-factors for one of the multiple audio factors include (i) packet loss in transmitting audio between two of the computing devices and (ii) clipping of audio signals. ### Claim 16 A computerized apparatus, comprising control circuitry that includes a set of processing units coupled to memory, the control circuitry constructed and arranged to: generate, during a real-time communication session over a computer network between multiple parties operating respective computing devices, real-time measurements of multiple audio factors of the communication session, each of the audio factors being susceptible to degradation in a way that affects an overall audio quality of the communication session; assign a respective MOS (Mean Opinion Score) value to each of the measured audio factors, each MOS value providing a measure of quality of the respective audio factor independent of other audio factors; combine the assigned MOS values to produce an overall measure of audio quality; and direct a software program running on one of the computing devices to render an indication of the overall measure of audio quality, the indication enabling the party operating said one of the computing devices to take remedial action to improve the overall audio quality. ### Claim 17 A computer program product including a set of non-transitory, computer-readable media having instructions which, when executed by control circuitry of a computerized apparatus, cause the control circuitry to perform a method for improving audio quality in real-time communications over a computer network, the method comprising: generating, during a real-time communication session over the computer network between multiple parties operating respective computing devices, real-time measurements of multiple audio factors of the communication session, each of the audio factors being susceptible to degradation in a way that affects an overall audio quality of the communication session; assigning a respective MOS (Mean Opinion Score) value to each of the measured audio factors, each MOS value providing a measure of quality of the respective audio factor independent of other audio factors; combining the assigned MOS values to produce an overall measure of audio quality; and directing a software program running on one of the computing devices to render an indication of the overall measure of audio quality, the indication enabling the party operating said one of the computing devices to take remedial action to improve the overall audio quality. ### Claim 18 (depends on 17) The computer program product of claim 17, wherein combining the assigned MOS values to produce the overall measure of audio quality includes converting each of the assigned MOS values to a corresponding factor probability, each factor probability indicating a probability that the respective MOS value indicates bad audio. ### Claim 19 (depends on 18) The computer program product of claim 18, wherein the method further comprises identifying a reference quantile, the reference quantile representing a particular percentile opinion score that subjective human listeners have assigned to un-degraded reference audio, and wherein converting each of the assigned MOS values to a respective factor probability includes, for each of the assigned MOS values: constructing a distribution of quality values whose mean equals the assigned MOS value; computing a cumulative probability value as a summation of the constructed distribution of quality values up to the reference quantile; and providing the cumulative probability value as the factor probability. ### Claim 20 (depends on 18) The computer program product of claim 18, wherein at least one of the multiple audio factors includes multiple sub-factors having different causes but contributing to a common perceptual effect.