Table of Contents

 1     Scope          
 2     References 
 3     Definitions 
        3.1     Terms defined elsewhere           
        3.2     Terms defined in this Recommendation 
 4     Abbreviations and acronyms
 5     Conventions             
 6     Introduction to ITU-T P.863 
        6.1     History of objective speech quality        
        6.2     Basics of a subjective test          
        6.3     Benefit of using Recommendation ITU-T P.863  
        6.4     Relationship with other Recommendations          
        6.5     Moving from ITU-T P.862 to ITU-T P.863            
        6.6     Challenges with ITU-T P.863    
        6.7     MOS misconceptions   
 7     Operational modes  
        7.1     Why two operational modes?    
        7.2     When should the narrowband mode be used?      
        7.3     When should the superwideband mode be used?
        7.4     Why no wideband (up to 7 kHz) mode? 
        7.5     Are narrowband signals scored equally in narrowband and superwideband modes?
        7.6     Can I map a narrowband score to a superwideband score?              
        7.7     Is it recommended to mix bandwidths in subjective testing?           
 8     Influence of reference speech on scores          
        8.1     What characteristics should a reference signal contain?   
        8.2     Are there constraints on the recording environment?        
        8.3     Can we use artificial speech signals?      
        8.4     What filter specification should be used?             
        8.5     Does reference material influence scores?           
        8.6     How much reference material should I use?         
        8.7     What is the impact of silence padding?  
        8.8     How do I validate a reference signal?     
        8.9     How do I use a 10 minute reference file?              
 9     How to inject test signals       
        9.1     Over which interfaces can signals be injected into a network?       
        9.2     Should the test signal be filtered?            
        9.3     What level should be used?       
10     Influence of recording process on scores        
       10.1     What influence does file speech level have on score?     
       10.2     Should we pre-align signal levels?         
       10.3     Should we apply a filter to the recording?           
       10.4     Does ITU-T P.863 handle silence padding?         
       10.5     Can I assess a narrowband recording in superwideband mode?    
11     Behaviour of ITU-T P.863   
       11.1     Global level changes  
       11.2     Local level changes    
       11.3     Bandwidth limitations
       11.4     Stretching and compressing speech       
       11.5     Step delay changes      
       11.6     Long-term gaps in speech         
       11.7     Short-term gaps in speech         
       11.8     Cellular handovers      
12     Comparison of ITU-T P.862.1/ITU-T P.862.2 and ITU-T P.863
       12.1     Scale differences         
       12.2     Degradation handling differences          
       12.3     Standard codecs          
13     Procedure for comparing subjective test results to ITU-T P.863 results 
14     Validation scope     
       14.1     Validated       
       14.2     Not yet validated         
       14.3     Outside scope              
Appendix I – ITU-T P.863 Reference validation checks    
Appendix II – Typical scores expected from ITU-T P.863 for a given codec    
Appendix III – Applications for further investigation    
Bibliography