Skip to content

P800 Tests - pre-amble and concatenation per category, repeat sequence before encoding

The processing-plan updates the steps on pre-amble, concatentation as follows:

Bildschirmfoto_2023-05-24_um_11.19.48

So, basically it consists of the following steps:

  • Processing is done per category --> will be covered in different issue
  • per-category item consists of
    • 10s pre-amble [-4;+4]
    • concatenation of pre-amble + all samples per category
    • this results in | pre-amble | item01 | item02 | item03 | item04 | item05 | item06 | item07 |; the background-noise is added on this sequence
  • This sequences is repeated before coding (| sequence a | sequence b |), resulting in the "Pre-processed input", which is then fed to the encoder The result of this repetition is | pre-amble | item01 | item02 | item03 | item04 | item05 | item06 | item07 | pre-amble | item01 | item02 | item03 | item04 | item05 | item06 | item07 |
  • after coding/decoding/rendering, the first iteration of the sequence is discarded, i.e. the second iteration of the sequence is used for the listening tests.