LHY UIP USB audio isolator
Apr 25, 2024 at 9:10 AM Post #31 of 45
I’ll second that, Phoenix has really grown on me, or opened up, burned in… now I’m not sure I’d ever give up the innous usb. Most noticeable in the low end but really everything is better. Course I think it’s 10 times the cost but you can get what you pay for
 
Apr 25, 2024 at 3:54 PM Post #33 of 45
Talking about FMC, just installed mine with a LHY LPS thanks to the guys over in the R26 thread Mickey and Jake included. Very good results and an even darker background. Question when I get my OCK 2 do I connect it to both the UIP and R26 or just to the UIP?
If/once you have an external clock connect it to both. Reason is to give the DAC the benefit of better timing too (but trust your ears as to whether it is in fact better as there’s many potential points of signal degradation with external clocks, cable quality - shielding and terminations, power quality, impedance matching etc etc), not because you want to ‘synchronise’ the internal oscillators of the two devices.

Relevance of oscillator precision in an asynchronous context
There may I guess be some benefit to having a single master clock synchronising both devices, but I don’t think it likely this is the main reason / mechanism of action for any audible benefits here as USB is asynchronous. Meaning the recipient DAC will be using its own clock (internal or external ref) to convert this asynchronous packetised stream to a synchronous / precisely timed internal stream before converting to analogue (versus the DI20HE which is doing the digital conversion and outputting a synchronous stream over i2S etc). Which begs the question what role does a more precise external (or internal in the case of the Phoenix) reference oscillator play in improving a USB output stream. As best I understand from reading on this topic (and re switches) try to find a plausible proffered theory as an explanation for what I’ve been hearing - is that the more precise timing along with cleaner power - helps generate a much cleaner electrical eye pattern* in the output USB stream (which at an electrical level in the conductors in a USB cable is an analogue wave) that makes the job of the recipient device much easier. Less jitter at the USB data interface level (not to be confused with jitter at the DAC/DDC synchronous conversion step) less buffering, error correction required, meaning less activity/thinking by its tiny circuit/brain = less micro voltage variation impacting the ground plane of the recipient circuit and the downstream loss of precision (downstream jitter) associated with that. Sounds an incredibly subtle electrical effect I know, but it’s the most plausible explanation/theory I’ve read for this.

Really good discussion of eye patterns here in an ethernet context.

@naynay FYI.. sorta on point to your OG question, I did warn you it was complex. 😉
 
Apr 25, 2024 at 4:13 PM Post #34 of 45
I’ll second that, Phoenix has really grown on me, or opened up, burned in… now I’m not sure I’d ever give up the innous usb. Most noticeable in the low end but really everything is better. Course I think it’s 10 times the cost but you can get what you pay for
Cheers for the feedback. This one was recently acquired second hand so assume well burned in.
 
Apr 25, 2024 at 5:06 PM Post #35 of 45
Last edited:
Apr 25, 2024 at 5:07 PM Post #36 of 45
As best I understand from reading on this topic (and re switches) try to find a plausible proffered theory as an explanation for what I’ve been hearing - is that the more precise timing along with cleaner power - helps generate a much cleaner electrical eye pattern* in the output USB stream (which at an electrical level in the conductors in a USB cable is an analogue wave) that makes the job of the recipient device much easier.
Not enough. It must be synchronised with GPS for the best effect. .LOL.
 
Apr 25, 2024 at 5:10 PM Post #37 of 45
But goddamnit, and I have to call it as I hear it inconvenient or not, otherwise what’s the point/who am I kidding.. the Phoenix takes it with a more warm and full sound with this remarkable bass weight.. incl more PraT and tempo. Not a gulf but it’s in the lead for now - as you might expect/hope for the price.

I hate you... :stuck_out_tongue_closed_eyes:
 
Apr 25, 2024 at 8:50 PM Post #38 of 45
Very sorry @m-i-c-k-e-y .

Well needless to say I couldn’t resist the temptation to - in Ghostbusters speak - cross the streams and use the LHY UIP to feed the Phoenix, with the UIP both internally and externally clocked.

You don’t really need me to tell you what happened.. just plot a mental trend line through my scatterplot of observations to date involving the UIP and you’ll have a pretty good idea. Thankfully not kaboom! 😅
 
Apr 25, 2024 at 10:10 PM Post #39 of 45
Here’s some impressions adding the LHYUIP (+OCK2) in front of the Phoenix:
  • cavernous soundstage depth with extended decay, amazing layering and instrument placement.. so much space between soundstage elements
  • Very busy multilayered music sections are a little more coherent, easier to follow the threads/laterals
  • Electronic music bass notes’ leading edges hit like a hammer, but still nicely textured, a little less of the Phoenix’s default generous bass quantity and warmth
  • Remarkable resolution, speed and dynamic agility
  • The extra precision means more delicacy and fine texture in vocals, if a touch less warmth and body, but still good relatively speaking
Removing the OCK-2 and running off the UIP (INT) the sound smooths over a fair bit, a touch warmer, bass more generous and flowing with less absolute command and speed, less dynamic weight and scale. Soundstage less expansive, closer to Phoenix alone. Vocals smoother/fuller so more forgiving with less than perfect recordings - the one area I’m not sure about with the EXT clocked UIP as definitely a touch leaner than the super lush baseline of the Phoenix.

OK, after adding back the OCK-2 to confirm the above impressions then removing the UIP entirely and listening to the Phoenix alone impressions - vocals definitely sweeter and more palpable than even the UIP (INT), a lovely effortless flow if a less obviously layered soundstage or as much dynamic scale. Hmm confusing… darn, far from clear cut. 🤔

More listening required I guess..
 
Apr 26, 2024 at 2:16 AM Post #40 of 45
Here’s some impressions adding the LHYUIP (+OCK2) in front of the Phoenix:
Same observations here. When I was trying out different external power supplies of the streamers USB output with a dedicated 5V in, which was further connected to the PhoenixUSB and then to the DAC, I observed some palpable changes to character of the sound mainly regarding its density, flow and liveliness. It was very controversial at first as per description the reclocker shall regenerate the signal from scratch with no respect of what happens upstream. But the reality appears to be different...
Needless to say that daisy chaining of different solutions along the USB signal path leads to amazing results!
 
Last edited:
Apr 28, 2024 at 3:14 AM Post #42 of 45
Here’s some impressions adding the LHYUIP (+OCK2) in front of the Phoenix:
  • cavernous soundstage depth with extended decay, amazing layering and instrument placement.. so much space between soundstage elements
  • Very busy multilayered music sections are a little more coherent, easier to follow the threads/laterals
  • Electronic music bass notes’ leading edges hit like a hammer, but still nicely textured, a little less of the Phoenix’s default generous bass quantity and warmth
  • Remarkable resolution, speed and dynamic agility
  • The extra precision means more delicacy and fine texture in vocals, if a touch less warmth and body, but still good relatively speaking
Removing the OCK-2 and running off the UIP (INT) the sound smooths over a fair bit, a touch warmer, bass more generous and flowing with less absolute command and speed, less dynamic weight and scale. Soundstage less expansive, closer to Phoenix alone. Vocals smoother/fuller so more forgiving with less than perfect recordings - the one area I’m not sure about with the EXT clocked UIP as definitely a touch leaner than the super lush baseline of the Phoenix.

OK, after adding back the OCK-2 to confirm the above impressions then removing the UIP entirely and listening to the Phoenix alone impressions - vocals definitely sweeter and more palpable than even the UIP (INT), a lovely effortless flow if a less obviously layered soundstage or as much dynamic scale. Hmm confusing… darn, far from clear cut. 🤔

More listening required I guess..
Someone with a lot more knowledge than me has pointed out that the internal clock in the UIP is 10 MHz and in the Phoenix it is 24 MHz clock.
I assume the higher clock rate the better the signal?
Thanks.
 
Apr 28, 2024 at 5:13 AM Post #43 of 45
Someone with a lot more knowledge than me has pointed out that the internal clock in the UIP is 10 MHz and in the Phoenix it is 24 MHz clock.
I assume the higher clock rate the better the signal?
Thanks.
That correct re their oscillator frequencies, I understand the LHY UIP and a similar SoTM USB regenerator use an internal clock sythesiser to convert the 10Mhz reference to the 24Mhz frequency needed for USB. Higher frequency isn’t automatically better but all things equal (incl assuming oscillator quality equal) it’s better to have an internal one operating at the native 24Mhz frequency and not require synthesis which introduces another variable, another potential source of jitter on top of the extra connections and cables needed for an external clock. Same logic holds for DACs and switches, if all things equal. Here of course it’s not equal, v likely the Phoenix has a higher precision oscillator AND it is operating at the native frequency. While the OCK-2 OCXO > UIP internal, even with all the potential sources of jitter of going external.

So many variables, another really important one, the quality of the power supply feeding the USB isolator and it’s oscillator (Phoenix has a separate rail). Ultimate judge of how all the design elements come together has to be one’s ears as I’ve done.

Anyway I gotta add even comparing the two of these is pretty bonkers.. (I mean when would we do this with DACs for eg) the Phoenix is almost 10x the cost of the UIP, the fact that the LHY is not totally embarrassed by the comparison is quite the compliment indeed.

Back from a weekend away, will focus on the LHY UIP again in coming days.
 
Last edited:
May 1, 2024 at 5:49 AM Post #44 of 45
picc.jpeg
 

Users who are viewing this thread

Back
Top