Tailspn,
what about modern recordings rather than remastering original tapes that are then done to DSD?
Hi Orb,
I'm talking about modern recordings. Actually, I'm talking only about acoustical event recording, mostly classical music and jazz, recorded in an event space like a concert hall. What Linn does in its production of recordings is applicable only to Linn. If you read again my post above, I detail the options available to any label/production company. How they implement those options is their decision, reflecting their priorities.
Non studio recording today, and almost as far back as electrical recording, is divided between the actual session, and post processing and mastering. The session recording can be either simply tracked with one mic per channel, and mixed/balanced later in the editing and mastering studio, or mixed/balanced in an analog desk at the session, and simply archived to the recording device. The former is the safest, and most common procedure, and the latter the best sounding, but risky.
In both instances, the session master must be edited. This is simply the assembly of "takes" by the producer/editor, with conductor input, to achieve the best musical interpretation. At this same time, for projects tracked with one mic per channel at full gain, the various mic/channels are mixed and balanced to present the orchestra/band/music piece to the producer's and conductor's liking. This form of production is what Linn is referencing needing to be converted to PCM. That is not the case however with editing on either a Sonoma or SADiE DSD workstation. There is no conversion to PCM, and no down sampling with attendant decimation. Editing (splicing) and level changes (mixing/balancing) are done at 64fs DSD sampling rate, but in an 8 bit parallel word format.
If the project was mixed and balanced in analog at the session, then editing is just the assembly of takes in DSD. Again, if the editing was preformed on a Sonoma, the edit region is the only period where DSD Wide is implemented. If done on a Pyramix, the same is true, but the edit region only is DXD. The remaining time is unaltered native DSD.
If the label determines it also wants sweetening of some form, EQ/reverb/et al, then the entire project must be converted and processed in DXD, just like level balancing and mixing, or as Bruce states, converted to analog, externally processed, and converted back to DSD. There are as many different processes in making a recording as there are labels and production houses.
As to recordings, as I stated in my post above, Telarc are all Sonoma processed, Blue Coast tracks to tape, and edits in tape, then using a Sonoma and ADC, archives to DSD. Channel Classics mixes and balances (stereo) in analog at the session, and edits (splices) with a Pyramix in DXD (affecting the edit regions only). Those are three that I know of, and there are certainly more.
Tom