12/29/2023 0 Comments Define lossless![]() So if you're looking for the right bit depth to use which just covers exactly what microphones can record you shouldn't limit yourself to 16 or 24, but any number in between (or below) might also be the magic number. But what microphones etc are able to resolve has no direct relation to bytes. You know perfectly well that 16 and 24 are used cause they happen to be a multiply of 1 byte (1 byte = 8 bit). Now tell me: Why stopping at 16/48? Why not downconverting to 15/48 or 14/48 or 13/48? Or maybe a very good studio master of a 2008 movie does contain enough real information for 17/48? I don't see why 16 would be the one and only "right" magical number. While there would be a mathematical loss in that conversion process, it is critical to note if that loss really encompasses actual audio content or simply Hanky, A conversion from 24/96 to 16/48 is entirely external to this process. ![]() "Lossless" refers to basic data compression within an encoding process such that the decompressed mathematical result IS identical to the original (that was fed into the encoder). "Lossy" refers to data compression/reduction within an encoding process such that the decompressed mathematical result is not identical, but sounds nearly (possibly indistinguishably) the same as the original. That is why it is better to just stick with the classic definitions of what is "lossy" and what is "lossless". Naturally, such a definition really doesn't do any good if it applies to everything under the sun. Hence, "everything" would be "lossy" by that same definition. While some may consider that as fair justification to describe it as "lossy", that would also ignore that various degrees of noise have been left behind at every stage of the process in the making of that soundtrack. Hence, that is the reason it is possible to mix/master the soundtrack in a 24/96 and export it to a 16/48 distribution level format, w/o having lost a single thing pertaining the actual audio content, except a whole lot of noise data. If you were able to analyze the spectral and amplitudinal scale of the material directly, you would find that it doesn't even exceed the specification for 16/48. Once you consider the limitations of microphones, analog electronics, and background noise, you realize that any kind of recorded material you could possibly come up with is not even approaching the vast specification that is encompassed by 24/96. ![]() The low-level detail that is present is entirely reliant on the material, itself, rather than the medium it is stored in. If it exists in a 24/96 format, does that automatically mean it utilizes the full performance envelope of the 24/96 format? It's easy to assume yes, but it's really much the same as the cassette tape scenario. Now consider modern audio content that would end up in a movie soundtrack. The original cassette recording will have been represented to its fullest extent, as far as actual audio content. So if you convert it to 16/48, have you really lost anything? Technically, you lost a whole lot of noise data, but that's about it. Do you think the resulting recording utilizes the full performance envelope of 24/96? It will still sound like an old cassette tape. ![]() The problem with this line of thought is what if you consider an old cassette tape is captured to 24/96.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |