Major QoE Parameters Used for Fault Identification in IPTV and OTT content delivery
To maintain high quality and detect violations in the stream, it is important to track various parameters. All monitored parameters are divided into two types: relating to Quality of Service (QoS) and Quality of Experience (QoE). In this article, we will consider QoE parameters that are most important for OTT/IPTV content delivery.
The following two parameters are essential for estimating the quality of experience, because lack of audio or video affect viewer experience dramatically.
Video decodability — the parameter estimates video stream decodability. The error state occurs if the number of correctly decoded frames is under the set threshold.
Audio decodability — the parameter estimates audio stream decodability. The error state occurs if bitrate of the decoded stream is under the set threshold.
Audio Track Missing and Audio Silence. To detect Audio Silence, the Audio Loudness Analysis algorithm is applied, that is implemented on the EBU R 128–2011 standard. If errors are detected, they surely affect user experience with the service. Timely notifications on occured errors help to quickly eliminate audio issues at the head-end station.
Synchronization continuity is one of the vital parameters for qualifying a high-quality broadcasting. Stream analysis based on Timestamps Discontinuity relates to the state when continuity of the PTS/DTS timestamps is broken, backward time shifts and sudden skips (packet loss and/or stream splicing) are detected in the stream. As a result, noize, pixelization, video freezes and black screen is shown to a viewer.
Audio/Video Information Changed and Program Specific Information Changed events describe changes in audio and video headers and in a stream structure. In some cases they may affect the viewer experience.
Encryption State Changed event allows swiftly detecting the case when an encoder accidentally starts producing encrypted streams, thought a decrypted stream is expected at the input.
In addition to the QoE parameters described above there are some valuable options that help in monitoring media data quality.
SCTE-35 messages related to advertisement insertion interval are detected and registered for further analysis (according to the standard ANSI/SCTE-35). These cueing messages can be used to activate more frequent capturing of thumbnails and/or to set the inserted ad recording when the event is triggered. This option can be useful for Playout Experts is charge for preparation of daily TV-programs with ads.
EPSNR is a metric showing a statistical estimate of the digital video content distortion during encoding. It is expressed in dB and defined as a ratio of peak mean square video signal to mean square deviation of the output signal from the original one. EPSNR (Estimated Peak Signal to Noise Rate) value estimation is based on encoded video stream data, so no source video is required. EPSNR is used to estimate encoders’ performance quality. The following values may be used: 25–30 dB — low quality, 45–50 dB — high quality.
Audio Loudness Analysis measures program loudness and peak levels of audio signals according to the EBU R 128–2011 standard: Program Loudness, Momentary Loudness, Short-Term Loudness, Loudness Range, Maximum Permitted True Peak level.
There are some additional QoE parameters applicable for OTT only.
Start With non-IDR Frame error occurs if a segment starts with a non-IDR frame. It can result in playback delay and/or video freezes along with pixelization and picture artifacts.
Interlaced Video event is activated if a segment contains interlaced video. Some mobile devices do not support interlaced streams decoding or such streams cause interlaced artifacts in media content playback. That is why interlaced video cannot be used in OTT, so this parameter is important to monitor.
To track all the parameters mentioned above, we recommend using Elecard Boro. Elecard Boro is a solution for UDP, RTP, HTTP and HLS streams monitoring and measurement of QoS and QoE parameters. It allows you to quickly find any stream quality violations, receive notifications about problems, and use monitoring statistics.