Dear all, in chapter 2 of http://jmvalin.ca/papers/aes135_opus_celt.pdf it is mentioned, that encoder and decoder has an algorithmic delay of 6.5 ms (lookahead and resampling delay). In Table 54 of https://tools.ietf.org/html/rfc6716 different delays for the resampler of the SILK encoder are given. However, I measure the same round trip delay for opus encoder->decoder regardless which maxplaybackrate I use. The round trip delay corresponds to the 6.5 ms. Therefore my question: How is this 6.5 ms algorithmic delay distributed between encoder and decoder? May it be, that the reference implementation uses 5.75 ms at encoder side and 0.75 ms at decoder side (using equal resampling delay for encoder and decoder)? With best regards, Martin -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20150424/fd534587/attachment.htm