I agree, though power noise on things like computers is complicated, because even if your feeding it a fixed DC voltage, it’s going require multiple voltages internally, and it likely is using a switching regulator internally to get them. And they pump out a lot of RF, which can have impact interconnects and other components.
I haven’t personally tried the Hypsos, the only person who’s opinion I would trust who has one likes his, but his system is VERY different to yours.
I’ve given up trying to understand how any of this actually happens, most of it I can justify, and I just go with my ears. But yes I’d rank digital interconnects as being one of the biggest changes in cables. up there with power cables, and neither seem like they should have any significant effect.
The Dave is also VERY sensitive to the quality of it’s input.
FWIW the MScaler does a pretty good job of cleaning up the incoming signal all on it’s own.
But if you are looking for a better USB source, I’ve found the Sonore Rendu’s to be really good at the price point, I have an Optical Rendu feeding my main DAC ATM, (my DAC is better through USB than even with a really good AES source), and I’m seriously considering picking up their signature Rendu if I sell my NS1.
100%. I always trust my ears more than anything else, and there is almost always a pretty clear preference for me - often unexpected (as here).
I don’t think there’s a better way for me to do this and still have a Roon server; at least I can’t think of one. I’m going USB out of the NUC-based Roon server into a DDC, and BNC out of the DDC. The M-Scaler has always preferred BNC to USB from any source I’ve tried.
This is what I did when I still had a Dave.
I have my Roon server on one side of my office connected to a switch in my network, I ran a 25ft optical cable to an SFP media converter, to a 12 inch ethernet cable to an EMM labs NS1, via AES into the Dave.
Honestly the optical run was a minor win, but it gives no path for the server to pollute the streamer/DDC, I know one of the big streamer manufacturer suggests WiFi for the same reason (might be lumin), the NS1 was night and day better than any of the lower priced streamers I tried (again this is all relative and differences get overstated).
I didn’t have an MScaler at the time, and I suspect that would reduce the overall impact of the streamer.
But what your doing is fine, and certainly not why your experiencing the issue with the Hypsos.
I’ve always been curious about optical, but never tried because of it’s rather severe bandwidth limitation. I have a decent amount of content at 24/192+ that I listen to, and don’t want to be forced to down-sample.
Re-reading your notes here, I also thought it relevant to mention that I noticed a downgrade in sound quality when I had my Roon server connected via wire to my router rather than wi-fi only. I discovered this by pure accident and wasn’t even A/B-ing. I just decided to move my server to my router and hard-wire and assumed that it wouldn’t matter, as I was using a wi-fi Roon bridge (ifi Zen Stream) at the time anyway. I thought nothing of the change until I started listening and something sounded off. I drove myself nuts trying to figure out what it was and what had changed, and I thought “no - couldn’t be the server - if anything this should be better.” But, sure enough, once I unplugged it from the router, the sound quality was back.
So, that’s all to say that I’m intentionally avoiding having a hard wired connection between my Roon server and router, as it seems to do better connected to the M-Scaler via USB → DDC → BNC → M-Scaler than removing the connection altogether. I’m guessing that having the power plant before the server also helps.
I do want to try a high-end streamer at some point (like the Emm Labs NS1 or Auralic Aries G2.1), but so far, the Singxer SU-6 DDC has had the most favorable results. I’ll likely be trying the Holo Red soon, as it supposedly has incredible jitter performance and seems like a really good power supply.
You can a/b wifi on/off and hear a degradation with it off and connected to lan. Have you also tried just disabling the lan adapter to see if that is enough?
Yep, during my time with iFi, the Ferrum Hypsos kept getting brought up as a better replacement to even our best in-house power supplies like the iPower Elite. It couldn’t be a coincidence with how often they were mentioned in our threads.
This isn’t rare in my experience. On several occasions I’ve noted serious video degradation when connecting by wire and wi-fi at the same time. IMO it follows from network hardware that either can’t keep up or that gets confused when processing traffic from two sources.
Fwiw, I have an Aries g1 and a G2 (not 2.1) on the way. It’s probably the best implemented streamer for wifi. Auralic recommends it over wired for their products as they pay special attention to it. I do recommend doing the one to one PSU mod for it otherwise, I personally wouldn’t recommend it sound-wise with its stock configuration.
Oh, and also I don’t recommend using roon on the auralic streamers, it sounds obviously worse. So that’s something to keep in mind too.
Yeah. I haven’t read a single bad review of the Hypsos. Everyone who had reviewed one liked it, and everyone who compared it to other products preferred it. I’m pretty bummed. The only explanation I can think of is that most folks probably haven’t gone to quite the lengths I have to work on noise floor prior to trying the Hypsos, and that I have somehow tweaked my way to below the noise floor of the unit that would otherwise be an improvement.
By the way - any idea as to why iFi hasn’t put out an iPower Elite that would be compatible with a Roon Nucleus (or NUC equivalents)? I feel like that would be a good seller, as the Roon forum is full of people looking for LPSs that aren’t willing to spend $500+. It would be 19v, 65W (3.42A).
This is good info, Im also a noise floor chaser, so anything that increases that is anathema to me. Even if its just an idiosyncrasy of your specific unit it is still a flag to be noted.
One of the curses of being blessed with a sensitive system is noting these “small” differences more easily and they are not always positive.
It’s easier to recognize as a hard to practically define background characteristic if you put something noisier in. Like changing interconnects or something. Ive tried to describe it before but it’s a nebulous audiophile term that is insanely relative to the listener. Some components just add a “fuzziness” to the background.
Indirect ways that a lower noise floor may present are increased dynamics, smaller dynamics changes may be more audible, as well as being able to pick out more details deeper in the music. To my ears it can have a large impact on the smaller stuff that you may hear.
It might be a very individual sensitivity as well. Some dont believe in the possibility of “greyness”, backgound is background.
I dont think there is one set thing I listen to to test noise floor, its more a perception across the board.
Ha! I was just completely unhelpful. Forgot why I dont post much here anymore!
As @perogie said, it’s a bit difficult to describe, but I’ll try. Like he said, there are small details, textures, and reverberations that you are used to hearing that start getting blurred or disappear altogether (or the opposite, as you lower noise floor). One will notice with any half-decently recorded track, but I find it the most obvious and easy to pick out with really well recorded drums. Instead of just the initial attack, as noise floor goes down, you start hearing the whole drum head and even micro-reverberations in the room or space around it. As @perogie put it, “more details deeper in the music”.
To attempt a metaphor, think of a track of music as a partially submerged sculpture. In a normal system, it’s 90% out of the water, and in no system is it 100% out of the water, but there are little details and frills in that last 10% that aren’t by any means critical to the song as you go from 90% uncovered to 95% uncovered, but you can definitely tell, and I personally find that extra dimensionality to be super appealing.
To extend that metaphor to the Hypsos, I’ll even grant that the Hypsos makes the above-water sculpture just a tick more sharp and defined, but it raises the “water level” from, say 3% to 7%, and I’m not willing to make that tradeoff.
My favorite track to use is on an album called “Polarity” by the Hoff Ensemble, which is incredibly well recorded, and even available in true 24/352 as a download. The first track, “Innocence”, starts with this incredibly delicate piano and drum intro that becomes outright magical with a super low noise floor. It sounds like I can hear through the drumhead and the little taps bounce off the walls. Another track that is likely more widely available is on an album called “The Gershwin Connection” by Dave Grusin, and the track is “Fascinating Rhythm”. It’s not quite as delicate as my first example, but changes to noise floor (and transient attack, for that matter) will be obvious enough.
I appreciate @andris documenting his mis-advenure with the Hypsos. It’s one of those occasions where something nominally good failed to happen. It’s rare, but it’s good to remember that not all changes are a positive.
BTW, that Hoff Ensemble recording is very good. It’s one of the few higher definition recordings I’ve downloaded.
One final (I think) follow-up on the Hypsos. I reached out to the Ferrum folks and told them about my observations, and asked if I had missed something I should try. They were super kind and quick in responding and did provide some suggestions. There were two options I tried toggling that I did not know about before, and I also tried adjusting the voltage (given that the Roon NUC can accept 12V - 19V). I also put in an even better power cord - AQ Thunder - to give the unit its best chance possible.
The better power cord did improve the sound, overall, but that alone did not match the stock SMPS. The modes and voltage adjustments also made very minor, but audible differences, but again, not enough to have the unit outperform the stock power supply in terms of noise floor.
With those options exhausted, I have indeed decided to move on from the Hypsos.
I want to close out by saying that there were audible benefits to the Hypsos - only with tradeoffs in my system, that I didn’t find satisfactory. Still, I read too many glowing reviews from folks I generally agree with and trust to believe this unit wouldn’t perform really well in other systems.
Thanks for the feedback on the Hypsos and the willingness to experiment. Goes to show how some power is different than others and you need to take the time to find the right fit.
Hopefully, this will help someone out who does land on the Hypsos eventually.
Hello all!
So, need some newbie advice…
Remodeling (thanks to hurricane Ian), and I am having the contractor put in a dedicated outlet for current and future plans of equipment. (at this time all I have is an iFi Gryphon and Focal Radiance)
They are using the yellow 12/2 wiring. Will go to a dedicated 20 amp breaker - with a basic electrical panel surge protector that was already there.
Q#1: is this a good enough wire to use (resistance issues, other, etc.)? And do I need to do anything else with this run from panel to outlet?
Q#2: What other not-crazy-$s surge/conditioner unit do you recommend for the equipment to plug into? (Audioquest, etc)