Another reclocker comparison

0 Members and 1 Guest are viewing this topic. Read 11132 times.

audioengr

Another reclocker comparison
« on: 28 Apr 2019, 05:02 pm »
I was fortunate enough to get one of the nicest reclockers, the Titans Audio Helen (Japan) to do some mods on, so of course I made some jitter measurements.  Helen is one of the few reclockers with I2S in and out.

http://www.titansaudiolab.com/en/products/Helen/Helen_guige.html

They advertise "residual" phase jitter at 2psec.

I also measured jitter on the Synchro-Mesh again, this time with the new upgraded oscillator.  Both using S/PDIF in and out and same cabling.  BTW, the Helen is $1700 with wall-wart.  The Synchro-Mesh by comparison is $699, less than half, but I used the $800 LPS for this, so $1500, but mine has LPS, not wall-wart like the Helen.

Titans Helen jitter distribution:



Synchro-Mesh jitter distribution:



Helen has more like 60psec of jitter based on these measeurements, not 2psec.  Synchro-Mesh has ~1/6 of the jitter of the Helen at ~10psec.

Titans Helen spectra:



Synchro-Mesh spectra:



Synchro-Mesh spectra is an order of magnitude lower in amplitude across the spectrum.

Steve N.

audio.bill

  • Full Member
  • Posts: 387
Re: Another reclocker comparison
« Reply #1 on: 28 Apr 2019, 08:05 pm »
Steve - Thanks for sharing your results which look quite impressive! I'd be curious if you'd also be able to share the results of your Synchro-Mesh with a more modest power supply instead of the $800 LPS. Maybe just with a typical wall wart or something around $200 or so. TIA for taking the time to share your data!

audioengr

Re: Another reclocker comparison
« Reply #2 on: 29 Apr 2019, 12:26 am »
Already did that in this thread:

https://www.audiocircle.com/index.php?topic=157348.0

About the same jitter distribution.  Differences with the Dynamo are more in the phase error noise floor and spectra.  This leads to a more organic analog sound quality.

Steve N.

path73

  • Jr. Member
  • Posts: 34
Re: Another reclocker comparison
« Reply #3 on: 30 Apr 2019, 04:39 pm »
There are 2 differences I see between Synchro-Mesh and Helen:
1) Helen claims to be Bit-Perfect
2) Helen can reduce jitter on I2S

Steve, how is the final result at the DAC output impacted by these 2 aspects when fed by the lower jitter SM?
Did you hook it up to your Overdrive using I2S?

Thx for sharing your results and thoughts, /path

audioengr

Re: Another reclocker comparison
« Reply #4 on: 30 Apr 2019, 06:35 pm »
There are 2 differences I see between Synchro-Mesh and Helen:
1) Helen claims to be Bit-Perfect
2) Helen can reduce jitter on I2S

Steve, how is the final result at the DAC output impacted by these 2 aspects when fed by the lower jitter SM?

Usually when there are two peaks separated like the Helen has, I have noticed some hardness in the sound quality.  The slower risetime will also add to the jitter, reducing clarity.

The voltage on their S/PDIF output does not adhere to the specification of .5vP-P.  It was .8VP-P with twice the risetime of the SM.  This is indicative of a designer that does not know how to properly design the S/PDIF output stage.

Quote
Did you hook it up to your Overdrive using I2S?

I wish I could have done this, but I would need an adapter cable.  Their I2S is not the same as PSAudio and Wyred4sound.  Not sure why they deviated from that standard pinout, except that I have found other info on the web that is wrong in the same way.  I sent them an email asking why their I2S pinout is different than the standard and no response....

Steve N.

path73

  • Jr. Member
  • Posts: 34
Re: Another reclocker comparison
« Reply #5 on: 30 Apr 2019, 08:56 pm »
Regarding my point 1), would you say that in your experience the possible loss in fidelity caused by the non bit-perfectness of the SM is of less influence on sound quality than the benefit of having lower jitter (in combination with the DACs that you tested)?

I guess you are also able to evaluate this aspect when comparing to Off-Ramp and Interchange with similar levels of jitter.

In my system, nothing comes close to the SM with your reference coax cable, so I would tend to think that bit-perfectness (at this level of quality) is secondary to reducing jitter. The convenience of the SM, not having to worry too much about the jitter in sources placed upstream (apart from poorly designed SRC found in some devices), makes it an absolute killer device to me.

I also find the Dynamo to give a "more organic analog sound quality". Can you actually measure the "phase error noise floor and spectra" that you associate with this subtle but totally addictive effect?

Cheers, /path

audioengr

Re: Another reclocker comparison
« Reply #6 on: 1 May 2019, 03:34 am »
Regarding my point 1), would you say that in your experience the possible loss in fidelity caused by the non bit-perfectness of the SM is of less influence on sound quality than the benefit of having lower jitter (in combination with the DACs that you tested)?

I determined some time ago that jitter is much more important than being bit-perfect, assuming that the algorithm for upsampling is a good one.

Quote
I guess you are also able to evaluate this aspect when comparing to Off-Ramp and Interchange with similar levels of jitter.

This is how I determined this.  Also, some customers with the Synchro-Mesh tried the Interchange and found that it sounded virtually identical.  Both have very low jitter.

Quote
In my system, nothing comes close to the SM with your reference coax cable, so I would tend to think that bit-perfectness (at this level of quality) is secondary to reducing jitter. The convenience of the SM, not having to worry too much about the jitter in sources placed upstream (apart from poorly designed SRC found in some devices), makes it an absolute killer device to me.

Wonderful.  Many customers just use a Sonos and a Synchro-Mesh.  It's a great product that keeps getting better.  It does not work by itself however.  You always need a source with it.

Quote
I also find the Dynamo to give a "more organic analog sound quality". Can you actually measure the "phase error noise floor and spectra" that you associate with this subtle but totally addictive effect?

It is measurable I'm certain of that.  I don't have the gear to do that measurement however.  This probably comes closest:

https://www.audiocircle.com/index.php?topic=163027.0

Steve N.

audioengr

Re: Another reclocker comparison
« Reply #7 on: 15 Jun 2019, 10:49 pm »
Customer asked me to mod the Helen reclocker, so here is the jitter distribution and spectrum before and after the mods:

Jitter distribution before mods:




Around 75 psec between the three peaks.

Jitter distribution after mods:




About 8psec of jitter now with one peak. Scale is 10psec/division versus 20psec/division before mods.


Jitter spectrum before mods:




Jitter spectrum after mods:




The vertical scale is 1/10 of the before mods plot.  21.5psec versus 2.3psec.

path73

  • Jr. Member
  • Posts: 34
Re: Another reclocker comparison
« Reply #8 on: 19 Jun 2019, 06:57 pm »
And... how does it sound?
Did you get a chance to test the I2S output?
/path

audioengr

Re: Another reclocker comparison
« Reply #9 on: 19 Jun 2019, 07:43 pm »
And... how does it sound?

Good question.  My Synchro-Mesh even with the wall-wart beats it at half the price.  When you add the Dynamo, it gets left in the dust.  It must be the jitter spectrum because the magnitude is very similar.  Sounds a bit strident and thin.

Quote
Did you get a chance to test the I2S output?

Only on the bench.  My I2S outputs are about the same as my S/PDIF given similar cable quality, so I would not expect much difference.  The main advantage of I2S is it bypasses the internal reclocking in some DACs.

Steve N.

Steve N.