It absolutely can be both, and effectively, too:

Hi-Rez Highway

New high resolution SACD releases, players and technology.

In Reply to: HDCD +Direct SBM posted by azure on May 31, 2005 at 05:17:51:

Nothing says that the SBM (essentially division and quantization from 2.8mhz DSD down to any sort of PCM) has to be done at 44.1/16. The SBM stage could easily be done at 1/64 with 20 bit quantization (44.1/20 PCM) and then passed through the HDCD encoder which further reduces it to 16bit+HDCD encoding cues.
SBM is nothing but quantization and dither. It's billed as (and I do believe it ) a "smart" way of doing dither (read: noise shaping), but that's all it is. If you "SBM" (or dither) down to 44.1/16, you can get around 18-19 bits of 'perceived' resolution, but your data words are still 16 bit.
However there's nothing saying that SBM's target -has- to be 16 bit. You can dither from 24 to 20, from 20 to 16, from 24 to 18: it's all a matter of how you handle that noise floor.
I'd even guess that this may provide a more effective use of HDCD encoding, since it's dealing with a 20 bit input.
HDCD on the other hand is not just dither: it actually does use 'cues' encoded into the LSB (of note: the bit that is 'dithered' generally, but doing it post-SBM would work) to change filter slopes, emphasis, and extension.
Both methods are effectively *two different and disparate ways* of squeezing a bit more "perceived" bit depth out of a 16 bit word. Dither helps, and HDCD's cued routines help a little bit further. They aren't the same thing, and aren't mutually exclusive.
I'm sure that MANY if not MOST HDCD's out there include a dithering phase before it hits the HDCD encoder. Whether you call that SBM, POW-R, UV22 or whatever, it's still just noise shaping, and it's a practice that has been shown to increase the perceived resolution of PCM recordings. HDCD encoding operates outside of that domain, and is generally not harmed by or precluded by dither.
And, since I know this question will come up: Yes, the source was multi-track (16 i believe?) 44.1/16. That does NOT mean the mixdown is limited to an effective 16 bits. That is 16 bits PER TRACK; and post fader, you can EASILY get 24 bits -or more- of effective bit depth and dynamic range. Think of -each instrument- having its own 16 bit space, versus the entire recording.
So yes, SBM+HDCD would certainly be possible, and could absolutely end up sounding better than redbook, even from a multitrack 44.1/16 source.

This post is made possible by the generous support of people like you and our sponsors:

VH Audio

Follow Ups Full Thread

Topic - HDCD +Direct SBM - azure 05:17:51 05/31/05 (3)

It absolutely can be both, and effectively, too: - Michi 22:07:14 05/31/05 (1)

Re: It absolutely can be both, and effectively, too: - azure 03:34:47 06/01/05 (0)

Re: HDCD +Direct SBM - Soundboy 09:37:31 05/31/05 (0)

Follow Ups

It absolutely can be both, and effectively, too: - Michi 22:07:14 05/31/05 (1)

Re: It absolutely can be both, and effectively, too: - azure 03:34:47 06/01/05 (0)

You can not post to an archived thread.