New Stellar DAC from PS Audio

has 8 ESS pro dac chips, 4 per channel, and a new pre-amp coming too

edit: 1 DAC chip with 8 channels, 4 each for R/L (in video, Paul said it had 8 dacs)

1 Like

I’m looking forward to the pair coming out. There won’t be any equipment moves till I’ve got the room sorted out. But even then I’ll be stuck on what I’d do - new front end for the M1200s or a complete redo. I’ve received lots of great suggestions but nothing yet is resonating.

I wonder if the new DAC will be considered a step up from the DSD Jr. If it is, I “might could be” (as they say) persuaded to consider trading my Jr and my SGCD toward the new pair. We’ll see. Price would be just as important as performance.

1 Like

Its an interesting move by them. Many have been hoping they build an integrated line at the BHK level. Now they go in opposite direction and make the Stellar line separates. They did need a DAC in the lineup cheaper than 8000.

What Paul is speaking of is the fact that the ESS pro chip is an 8 channel chip. What they are doing is the same as what Wyred 4 Sound does in their DACS. The use 4 of the channels for the right and 4 for the left. He is not saying they are using 8 chips. Just to clarify this.


Paul said it has 8 DACs

1 Like

Hopefully, Paul takes this opportunity to clarify or reinforce his statement. Personally no horse in this race.

I thought it was interesting that Paul pointed out that both the new Stellar DAC and the Stellar Preamp were designs from Darren Myers before he left PS Audio.

It did sound a bit odd. My Lumin T3 does the same, presumably other Lumins as well, two ES9028Pro chips in dual mono configuration. The low noise level comes mainly from the power supply design. Having recently bought the Lumin, no skin in the game either. Anyone with Stellar amplifiers will need a pre-amp.

Paul mentioned Darren left with a lot of designs well progressed, one of which was a BHK level phono stage.

1 Like

Reread what I said. It is a single ESS pro chip that is designed for 7.1 channel surround applications. 4 channels paralleled for the right and left channels.



Yup, it’s a new class of product we’re working on—between Stellar and PerfectWave. Still scratching for a series name.

8 Dacs—4 per channel in parallale.


A new Perfectwave DAC? TSS? Sign me up for the betatest!

Interstellar OctoDAC seems the obvious name, but you might get sued.

Surely more DACs mean more noise, not less? That’s about $400 worth of DAC chips alone.

1 Like

No, quite the opposite. When they are in parallel, they lower noise. Think of it like paralleling transistors: one transistor has X amount of noise while another in parallel, cuts in half the noise. These paralleled devices are part of the secret to how we make ultra low noise MC phono stages.

Love Interstellar!


It was a great movie and soundtrack. Hans Zimmer and Christopher Nolan. What could possibly go wrong? It was recorded at Air Studios and Temple Church, a beautiful place in Central London, almost 900 years old, which featured in the Da Vinci Code movie. It’s in our legal district and there are regular concerts there. Christopher Nolan fell in love with their organ, which is prominent in the soundtrack.

As for 4x2 DACs, it’s been done before.

Apparently multiple DACs has long been argued over. A search came across a discussion on ASR and a gentleman from Colorado gave a few ±’s off the top of his head:


  1. Randomizes threshold errors (especially helps with steady-state signals) to improve linearity (this includes aperture – jitter – errors);
  2. Depending on architecture can improve SNR since signal adds coherently and noise does not so you gain a little with each extra DAC;
  3. May increase drive level from parallel outputs (again architecture-dependent);
  4. May reduce timing/settling requirements (may reduce sampling rate for each DAC, providing more time to settle); and,
  5. May provide higher sampling rate (flip side of point (4) – use multiple DACs to provide higher effective sampling rate).


  1. Can be very difficult to align levels and time (clocks) to integrate the DACs – any errors will introduce spurs at combinations of sampling and signal frequencies;
  2. Combining process can add errors (switches add distortion and glitches, passive combiners can cause gain errors, distributing a low-jitter clock to parallel DACs, can be a challenge, etc.);
  3. May not gain SNR if signals do not combine (some architectures just rotate among DACs);
  4. Increased SWaP (size, weight, and power) and cost for DACs and ancillary circuits; and,
  5. Output capacitance may be higher from combined outputs, reducing output bandwidth.

None of that means anything to me, but it sounds like it’s nice in theory but really needs good implementation and can go wrong.

So it seems very interesting that Darren took up the challenge, because it seems like it needed someone as good as Darren clearly was/is to make a go of it.

this new dac will cost less than the DSD MkII

is the planned ESS pro 8 dacs-in-one cheaper than the programmable FPGA?

I’ve been a happy Stellar stack (GCD, M700) camper, very intrigued by these products, particularly if the new pre has 2 sets of balanced inputs and an HT bypass.


  1. Will the new ‘Stellar+’* DAC solve the problem of no-volume-control-synching among 3 PSA DACs?
  2. Will it process what I call raw-DSD out of a Memory Player or just regular DSD that virtually all modern DACs can process?

Volume matching might require 3 identical ‘Stellar+’ DACs?

As you can see, I’m STILL trying to improve my SACD-playing system for 6-channel service.

  • There’s your new series name!