eclassical Christmas sale

[tailspn]

Bob,

Although a DSD diehard, I admit to not getting technical enough yet with it. My understanding is that the Sonoma workstation will mix and master within DSD, no PCM off-shoring required....

 

Mixing is the process of both combining multiple sources and/or level changing. The algorithms are the same, regardless of whether it's a 1-bit stream, or a multiple parallel bit (word) stream. However, to operate on a value in a 1-bit data stream (not possible since any one sample does not constitute a magnitude level), or to efficiently use the existing computational hardware, and software, 1-bit data streams (DSD) are converted to parallel word based data streams, either DSD-Wide (as used in the Sonoma or SADiE DAW), or PCM. This is explained well here under the heading DSD:

 

Direct Stream Digital - Wikipedia, the free encyclopedia

 

The is a sound quality issue is in the conversion from DSD to PCM for processing/post production. Making a 8-bit word for DSD-Wide has no negative effect in the conversion. It's simply a packaging scheme. Making a parallel PCM word(s) from a DSD data stream does have sonic consequences, and they increase as the sampling rate ratio between the DSD and resulting PCM increases. The conversion requires two steps:

 

1, Decimation filtering to filter high frequencies that would exceed the Nyquist frequency of the new, lower PCM sampling rate

 

2, Computing a value within the PCM range of data values that most approximates the DSD value.

 

It can be argued (and it is incessantly) whether these conversions affect the quality of the sound. Many, including myself believe it does, and is particularly noticeable in the corruption of spaciousness information.

 

To Ted's points, The Sonoma and SADiE Workstations both use the Sony (Oxford actually) developed E-Chip, which operates in both the 1-bit and multi-bit domain, but always at the original DSD sampling rate. Therefore no decimation and no conversion approximation. Telarc and Kimber's recordings are edited and mixed on Sonoma. Sony (Sonoma) made a mixer, more like an oven that uses multiple E-Chips, and I believe Telarc used that for its multi-mic mix downs. Genex also made a 48 channel DSD Mixer, using FPGA's. The recent Pentatone Music For a Time of War with the Oregon Symphony was produced with it. Ray Kimber doesn't do multi-mic mixing, since his mic count is always less than eight.

 

Channel Classics, Polyhymnia (Pentatone releases, among others), and many other quality recording production companies use Pyramix for DSD recording, and DXD (PCM) for mixing and editing. Editing in DXD however, is a special case. Only the edit interval ( much less than a second) is actually in DXD (352KHz PCM). The surrounding DSD content is not converted. Recently, with the release of Pyramix 8, they are now performing edits and crossfades entirely in DSD.

 

Channel Classics is more akin to Blue Coast, but without the original recording to tape stage. Jared's recordings are mixed in analog at the session on a custom purpose built analog mixer. He uses best quality Van den Hul microphone cabling, and super transparent battery powered mic preamps to achieve his sound. Jared also used Pyramix for editing, but at that time, the levels have already been established in the analog mixing stage. He, like Cookie uses DSD primarily for archiving and delivery.