super bit mapping

SBM or NO SBM, that CD is great !!!

Miles Davis - Kind of Blue. It is reference material.

Try and hear the LP if you get the chance, it will bang the CD by a mile !!! [This is a comment for SBFX]

Regards,

BHAGWAN69

P.S. I would like to know the technical details of SMB [Super Bit Maping] though; Since we are at it, DSD too could be explained in detail [this is for CD & not for SACD].

Hold on guys, before you take off, DSD is for SACD, I am aware of that, but there are a lot of CD recordings that come on DSD these days, I want to understand what that means....... Please..........Thanks !!

As far as I understand it, Super Bit Mapping or SBM is Sony's proprietary method for using 20 bits during re-maastering. It's Sony's answer to HDCD & to JVC's K2 re-mastering process both of which also use 20 bits & are also proprietary methods. I'm not sure who the original owner of HDCD is but today Microsoft owns it.

Anyway, during re-mastering what SBM (or JVC's K2) does is that it digitizes the original analog master tapes into a stream of 20 bit digital music (rather than traditional 16 bits that they used to). Then they process that 20 bit digital stream the way they want to during the re-mastering process (whatever enhancements/tweaks they want to do to the intruments, vocals, sound-staging, reverb, etc, etc). The key reason for using 20 bits is that during the process of re-mastering the digital music goes thru a DSP (digital signal processor) which is a bunch algorithms implemented in hardware/electronics. We all know that electronics adds noise. So, as the music signal is being enhanced/tweaked, noise is being added (the engineer has no choice in this matter as he HAS to process the signal to get his enhancements/tweaks) to it.
Noise is always the smallest amplitude component vs. the music signal. Hence it always affects the LSB (least significant bits). So, as the music signal gets more & more enhanced/tweaked during mastering, the accumulated noise begins to affect the bottom of the 20 bits i.e. if you write a 20 bit digital word, which is a 1 or 0 pattern, the bits on the extreme right, which are the LSBs, are the ones that are affected by noise.
By the time the entire remastering process is complete what started out as a 20 bit stream of digital music now degenerates to a 19-bit or an 18-bit or even a 17-bit stream of digital music.
Note that, with today's technology, the word length of the digital music is 99.9% greater than 16 bits post remastering!
So, when they take the remastered music & make a redbook CD of it, which is 16 bits, the consumer is assured that he/she is getting 16 true bits of music.
If the remastering process had begun w/ a 16 bit digital stream & then DSP'd that data, the consumer would have a remastered redbook CD that would be 13-15 bits at best. That would mean a much reduced dynamic range (DR) of the music & it would make the consumer say "CDs are not delivering their promise of high fidelity music!"

For SACDs, DSD format, as I understand, is recording at a much higher rate of 2.8224MHz. This is 64 times the redbook CD rate of 44.1KHz.
The premise of DSD is that if you up/oversample an analog waveform at a high enough rate, you can make it *appear* as tho it's analog!
The act of up/overampling makes an analog waveform digital BUT if you take sooooo many samples of that analog waveform, you can fool the system into thinking that the waveform is analog (because there are so many consecutive samples of the analog waveform that it is assumed that nearly every nuance of the analog waveform has been caught by the multitude of samples).
That is why SACD enthusiasts claim SACDs sounding akin to LPs. (this is an eternal debate on several forums & nobody seems to be winning as there are equal #s of people pro & against the format).

For data that is resampled to DSD (like dCS gear), the CD is read off at 44.1KHz & up/oversampled 64X to 2.8224MHz using the company's proprietary algorithm. Essentially, this algorithm is a complex & highly tweaked digital filter whose sonics are considered to be acceptable to the end user. Needless to say, each company's algorithm sounds a bit different from the other as there is no "correct" sound. So, you are likely to prefer one manuf over the other.

Very well written bombay walla..
thanks for the info.

Wow! Thanks bombaywalla, that was informative.