I have a large headphone collection which actively competes with my 552/500 system and am heavily involved with Stax running several 3rd party amplifiers (universally acknowledged to be significantly better than stock Stax offerings). If you are interested ask Peter Swain for my email and I would be happy to bring you up to date on Stax.
Best
Gregg
Ah… end stage audiophilia.
How is the Atom he more convenient vs your 500 system?
One box vs three (or two vs four). Fewer cables. Better control over volume (digital vs analog). Versatile device placement. Nicer remote.
Hi Bud
One single Linn DSM box with a single cable going to each active Linn speaker
or
Your ND555 into a pre and then to active ATC speakers
Best
Gregg
ND555 (2 boxes), NAP 552 (2 boxes) headphone amp (1 box). DIN/DIN IC plus System Automation cable, Two remotes.
Atom HE (1 box, 1 remote) Can be placed anywhere with WiFi.
Naim 300 DR amp (2 boxes)
Naim 250 DR amp (1 box).
I’ll keep my Harbeth 30.2’s.
I will ask Peter thanks Gregg
The Stax amp I was looking at was the T8000 by the way - I know Woo (?) make competing devices, so I’m certainly interested in hearing your experiences 
Hi Matt
Stay away from both
Best
Gregg
Bud
A good friend on this forum (@MagicBus) who has been a Naim user for many decades now runs active ATC’s and a single box Ayre preamp. He has had various sources including ND555, Linn KDSM, Chord Dave and currently dCS Rossini.
The Linn, dCS, and Chord can all be single box source solutions - and when combined with a single box like his Ayre pre - result in a simple 2 box solution.
Best
Gregg
There are reasons why I need to stay with Naim. Otherwise, I’d get a dCS Bartok. It includes a headphone amplifier.
I’ve had the whole gamut of Chord Electronics including Dave. I don’t care for them.
I’ve tried several speakers in my small apartment and the only ones that worked for me are the Harbeth’s. I don’t want to limit myself to active speakers.
Headphone listening is more important to me than speaker listening because of my small apartment.
Bud
Bartok was great for headphone listening. It gave my Linn a run for the money. So good that I traded it back in for a Rossini.
Best
Gregg
It is disappointing that the analog input on the Atom goes through A/D conversion. Since you want to stick with Naim, you’ve got the best available solution. Most would likely find the transition from the 555 difficult. You’ve indicated you’re going to not immediately change anything so perhaps a good temporary solution would be to connect the Atom to your 552 using AV bypass mode? This would only require a more common RCA → Din cable vs the RCA → XLR.
Actually, that’s what I’ve done. I’m using a DIN Out to RCA cable from Teddy P. It does sound a bit better than driven by an Ethernet cable.
Even so, I find the standalone HE very enjoyable. I’m thinking about having Chord build me an RCA to dual XLR cable so I can try using the HE with my 300 DR. That should give me an idea about what the HE can do.
If that sounds encouraging, I’ll have Chord modify the cable snd pick up a 250 DR. I can see how that works and if necessary obtain an NXD-2 with the HE acting as a preamp between the devices.
In any case, I can use the NDX-2 and 250 DR’ as replacements in my current system. I’d need to find a replacement for my NAC 552.
Your post convinced me to demo the Bartok back in March. Now I’m waiting for my Rossini & clock to be delivered. Prior to hearing the Bartok I was really just looking for a <$500 Roon endpoint with USB out to connect to my V1.
In between the Bartok & Rossini, I upgrade my SN2 → 282/300DR + full SL cables + upgraded rack. Still considering speakers.
The Bartok HPA was like the heroin of audiophilitis.
That’s what I’m afraid of… going down the dCS rabbit hole.
Some time back I thought I needed to downsize. I went all the way to a Naim Nova. That was premature and too far.
Then I built back up to a Naim 500 level system.
Because of age and health I think I do need to do it now but this time with more control and thought.
Sorry for that 
The Bartok has been a great gateway drug for dCS. Congratulations on your incoming Rossini. I have 3 Naim friends here in the States who have joined the dCS crowd this past year. Based on membership on the dCS forum there is a rather large Naim contigent.
Best
Gregg
My only problem with dCS is their instance that an external master clock is critical. It’s only a problem because the rationale requires sophisticated understanding of electrical engineering and I lack even basic insight. Initially I assumed the external clock is better design and the internal clock is turned off during replay. Of course that’s not the case. The internal clock is still used. The external clock somehow regulates the internal clock. Then I looked at the design of the Vivaldi clock and it is identical to internal Vivaldi clock design. In fact the Vivaldi dac is built on top of the Vivaldi clock. It looks like the primary source of improvement is the fact the external clock has its own power supply.
My theory is 100% based on examining pictures of the pcbs in the Vivaldi clock and dac. After several hours I was unable to find a single component that was different between the two. The upsampler also uses the same Base pcb.
Coming from Naim ecosystem, I appreciate the role of power supplies. Even so… the conspiracy theorist in me can’t prove that dCS hasn’t simply place some code in the dac like “if external_clock == true then Do (SoundBetter(external_clock_model))”. Another alternative is the code running on the external clock is better, could run on the dac board, and dCS wants to sell clocks.
At the end of the day, I’m not about to undertake a course in electrical engineering and it doesn’t matter why it sounds better if it’s a reasonable value. Just curious how MSB & dCS both produce excellent results with completely different design theories. Unfortunately the answer is likely incredibly complex. 
Just a gentle reminder to members before the thread veers way off topic into all things DCS - let’s try to keep the thread on topic (Atom HE). Thanks.
Maybe like Naim, there is quite a bit of testing and measurement taking for incoming components and component matching. The components with the lowest noise, tighter tolerance, more time spent with component matching, etc. end up in the top-level boxes, those at a slightly lower quality go into the range below and so on.
What headphones are people using with the HE?