(2017-01-26, 12:31)noggin Wrote: So is 5.1 E-AC3 the same in compression terms as 5.1 AC3 (i.e. they haven't introduced a new, more efficient scheme) If both Type 0 and Type 2 streams can be re-wrapped from E-AC3 to AC3 with no transcoding that suggests the core compression scheme is the same?
You are a tenacious person
I thought I could get away with the one sentence reply. I didn’t want to get into the details, that’s why I said “re-wrapped” and “changes in the audio block”. There is actually more to it. @
Soli is correct in saying that Dolby may be stretching the truth when the say there is no decoding/reencoding in the conversion process. That is true if none of the enhancements of E-AC-3 is used (these enhancements are what contribute to the greater compression efficiency of E-AC-3). In such cases you can just "re-wrap" the bit stream. If E-AC-3 features like vector quantization (VQ), gain adaptive quantization (GAQ), channel coupling/phase preservation, spectral extension etc. are used in the bit stream, there is indeed a partial decode/encode for requantization during the conversion process. Unlike the open source encoders/decoders,
there is NO decoding to PCM during the conversion. The partial decode/encode remains in the frequency domain. Encoding parameter data such as
snroffset used for bit allocation, exponent/coupling exponent strategy, audio blocks in the frame used for calculations etc. are already carried in the E-AC-3 bit stream. The converter can utilize those data for converting the E-AC-3 bit stream to AC-3. In short, the conversion process is less computational than a full decode/encode.
Hope you are satisfied with the answer
I may not be able to give more details.