Bob - it’s a very big deal in pro audio. The best available vintage hardware (mic preamps, compressors, EQs, tape machines, etc.) are found and their circuitry modeled and turned into plugins. Recently it has gotten so good that studio engineers that have owned the original gear for 30 years are saying some of the emulations are as good as the real thing. Universal Audio is doing some of the best work in this area, while still producing current versions of their classic hardware units.
Badbeef, wild! Wonder what’s in the code…
Yes, when comparing sound quality to the Bridge I am always assuming people will use the I2S connections and that they have PS Audio DACs. That might be a shortsighted view but anything else assumes I might have a crystal ball into what folks would connect and that’s where the music stops for me. Bridge III can only be connected via its internal I2S connection within a PS DAC so naturally…
Octave the server will not have an Ethernet output, only an Ethernet input.
I can’t yet comment on the hardware innards because they are a moving target right now and far from completion.
Bob - re: code, I’m the wrong guy to answer that. It would be interesting if and when this sort of thing can be implemented as a standalone chip, rather than requiring a computer and software platform on which it runs. I suppose they’re on that, since I’ve thought of it…
Interesting thought for Paul and Ted. Dang near anything you record sounds better run through the Studer A800 Tape plugin. (I bet it could beat MQA ; ). It could be the future of “tone controls”. Obviously, it could be implemented on a computer-based music system as is, as long as you use the required software and hardware. The UAD stuff requires a UAD interface, which has a totally different intent from an audiophile rig, but hey, people run Amarra with the Sonic Studio studio interfaces…
@tedsmith - maybe you can allocate a chunk of a future FPGA to some sort of processing along these lines - best of old and new tech in one : )
Paul McGowan saidYou’ll be fine with Bridge III if you’re sticking with your NAS. It’ll be a step down from a performance standpoint because of the hardware and isolation benefits of the separate chassis. At this point I see no reason it will have any fewer capabilities for DSD rates, etc.
OK, for those of us who want the best SQ from our NAS could you please produce a (small) box containing Bridge III and the necessary power supply, isolation, etc with minimum connectivity - ethernet in, I2S HDMI out?
Before switching to the PS Audio DS DAC & Bridge I used a stand-alone renderer (Sonore Rendu) so I can appreciate the potential SQ gains to be had from isolating this function from the DAC.
Bob said Badbeef, wild! Wonder what's in the code..You do not want to know. Seriously.
For example, the Studer A800 was a multi-channel tape deck, introduced in 1978. Micro-processor controlled! Highly respected and very popular. Many classic highly regarded pop albums were recorded on a Studer (Prince, Jeff Buckly, Metallica, Tom Petty, Stevie Wonder, Tom Petty, etc.).
Universal Audio’s plug-in, which replicates the sound of the A800, adds the sound of tape compression and tape saturation, replicate the typical calibration choices people apply to the Studer (with the concomitant differing noise floors, headroom characteristics, and frequency shifts), the low frequency boost associated with tape, various distortions, etc. Studer even endorses the plug-in as sounding like their tape deck.
Working together, these distortions create the studio sound many love: the warm, cohesive presence many audiophiles adore and which vinyl takes even further.
As Badbeef states, “Dang near anything you record sounds better run through the Studer A800 Tape plugin.” 
But we need to remember the resulting sound is intentionally distorted from the input.
davidl saidOK, for those of us who want the best SQ from our NAS could you please produce a (small) box containing Bridge III and the necessary power supply, isolation, etc with minimum connectivity - ethernet in, I2S HDMI out
Interesting problem. I like the fact that the Bridge is internal (I have so darn many boxes of various sizes in my audio rack, along with their power supplies ). How much of a SQ difference are we talking about here? If it’s small I would prefer to leave the Bridge as an internal component. I always assumed that having the Bridge inside the DAC didn’t affect SQ appreciably or PSA would not have put it there, given the efforts both Ted and the team in Boulder have put into getting the best sound from the DS. But maybe I’m wrong.
It isn’t trivial to compare sound quality with an internal vs external bridge. Clearly the bridge has a processor on it and that in itself can generate significant noise (both conducted and radiated.) But DS’s also have a processor on the display card and the display itself, which already generate these kind of noises.
An internal implementation will have to isolate the bridge both on the power supply side and possibly in a box (no box on the DS or the DS Jr.) On the other hand, for example: it can use a very clean data path to the FPGA, it can share power connections cutting down ground loop opportunities, it can share clocking (keeping beating between the FPGA and bridge clocking down: the FPGA on the Jr generates all clocks for the bridge module.)
An external implementation can be better isolated electrically and better shielded for RF. But it will require another power supply which can make more ground loops and it will need transceivers on both ends of all of the connections between it and the DS, these will add jitter and electrical noise on each signal even if everything else was perfect.
Thank you for sharing this info, Ted. Each implementation obviously has its pros and cons, but it does seem that a carefully engineered internal Bridge is not necessarily more noisy than an external one would be.
Paul
Have you given any thought to integrating some voice control into the server, for example using Amazon Alexia or Google Assistant?
I am not sure if this makes sense or how much effort it entails.
So, for example you good say " Ok Alexa create a playlist of Van Morrison songs from the seventies", or whatever?
magicknow
It’s a great idea and I like it though that’s not something trivial without having to add an Echo device itself (all the microphone technology they have is critical).
thanks please keep this on the list of possibilities.
Ted Smith saidWe have an interesting situation here on the relative SQ from PS Audio devices:It isn’t trivial to compare sound quality with an internal vs external bridge. Clearly the bridge has a processor on it and that in itself can generate significant noise (both conducted and radiated.) But DS’s also have a processor on the display card and the display itself, which already generate these kind of noises.
Paul was clear when first auditioning the DMP that to his ears it gave better SQ through the DS DAC than both the PWT (via I2S) and Bridge II fed by his old server via ethernet(?). At that time it was not clear why this should be so.
Judging from Paul’s comment in this thread he now thinks the SQ from the new Octave server (via I2S) is better than that from the bridge (II or prototype III?). So I guess PS Audio may have identified what made the DMP give better SQ and implemented some aspects of this in the Octave server.
If this is so then am I being too simplistic to expect that stripping the Octave server down to ethernet input > processing > I2S output (i.e. omitting all the storage and processing associated with the server functionality) should give SQ as least as good as that from the Octave server?
I would take his comments to indicate that PS Audio has learned how to make quieter devices over time. I’d like to think I had something to do with that (some of my suggestions went into the PWD Mk II) but their engineers are no slouches and any good engineer makes improvements with each additional product.
My point about internal vs external was only that such simple binary choices are rarely actually simple and it’s the design of the unit as a whole that matters.
magicknow saidHave you given any thought to integrating some voice control into the server,
This seems like something that would take a lot of programming resources. Before doing so, I suggest that PSA try to get some sense of how many people would use this and therefore whether it is worth devoting resources to. (I personally never use such devices.) There are plenty of other claims on the programmers’ time, I’m sure (how’s that DMP upgrade coming?).
It would take an insane amount of resources for limited functionality. Servers actually do very little, voice controlled or not. It is not as if a music server is going to give you the weather and read your email aloud.
It would be last on my wish list.
It wouldn’t be on my wishlist. If you can make playlists, choose one then just put it on random.
Plus there has to be a licensing fee, that would add to the price of every unit, for a feature, most audiophiles would never use.
Below a Hello Kitty face plate?
I just realized I can talk to my Android-powered TV. Probably had it a year. It’s what I get for not reading manuals. So soon, our devices will be battling for supremacy.
TV: “WHICH ONE OF US do you want to play the music, Mark?”.
STEREO: “You may make pretty pictures and all, but you sound awful!”
ME: “BOTH OF YOU shut up.”