[audioengr]

For those that don't subscribe to TAS, the review of the Peachtree iDAC in the July/August issue is very interesting.

http://www.avguide.com/review/peachtree-audio-idac-tas-214

Steven Stone concludes that the digital source is actually more important than the DAC itself.  When the iDAC and the W4S DAC's were driven with his Off-Ramp 4, not only did they each sound a lot better, they sounded almost identical.  No surprise to me

This is all the more reason to invest in a really good digital source, with the best clock you can get.  Anyone know where to get one of those?

Steve N.

[serengetiplains]

Congratulations, Steve.  I think you're quite right about the importance of clocking.  It is, after all, the essential aspect of digital signal transfer fidelity.  What else does digital fidelity concern?

[Tyson]

Hi Steve,
At some point I was thinking about getting a miniDSP 4x8 since I actively tri-amp my speakers.  Do you think the off-ramp would be a good fit with something like that?

[audioengr]

Congratulations, Steve.  I think you're quite right about the importance of clocking.  It is, after all, the essential aspect of digital signal transfer fidelity.  What else does digital fidelity concern?

I put analog distortion from compression as #2.  Eliminating the preamp improves this.

Other principal digital things that affect fidelity are:

1) digital filtering  - use it sparingly or not at all
2) D/A conversion technique - affects resolution and S/N ratio - use modern D/A chips, but avoid switched-capacitor techniques
3) digital cable - make sure it's 1.5m long and low-loss

Steve N.

[audioengr]

Hi Steve,
At some point I was thinking about getting a miniDSP 4x8 since I actively tri-amp my speakers.  Do you think the off-ramp would be a good fit with something like that?

The Off-Ramp provides a low-jitter digital source for any system.  What is a miniDSP?

If it's like a DEQX, then this will work well.  A number of my customers use the Off-Ramp to drive DEQX.

Steve N.

[Tyson]

Details here - http://www.minidsp.com/onlinestore/detail/7-minidsp-kits/flypage/73-minidsp-2x8-kit?sef=hcfp

[werd]

Thats good news Steve. How about usb cables into the off ramp? Is it insensitve to change on the computer side of it ? If so you could possible have the first audio streamer thats also the starting gate in an audio chain.  All you want to hear is differences in file resolution and thats it imo.

[audioengr]

Thats good news Steve. How about usb cables into the off ramp? Is it insensitve to change on the computer side of it ?

I have not done this test since I updated the voltage regulators.  I previously had to use a Polestar with the Off-Ramp 3, but I have found that a 16-foot Belkin is fine with the Off-Ramp 4.  It probably does make a small difference however.  I had one report from a customer to this effect, but he did not have the latest voltage regs.

Steve N.

[targa02]

I put analog distortion from compression as #2.  Eliminating the preamp improves this.

Other principal digital things that affect fidelity are:

1) digital filtering  - use it sparingly or not at all
2) D/A conversion technique - affects resolution and S/N ratio - use modern D/A chips, but avoid switched-capacitor techniques
3) digital cable - make sure it's 1.5m long and low-loss

Steve N.

Hi Steve,  With regards to the digital cable length - Should it be less than or equal to 1.5m long or is there something special about 1.5m in length?  Also, how does one determine if a cable is "low-loss"?

Thanks for your response!

[audioengr]

Hi Steve,  With regards to the digital cable length - Should it be less than or equal to 1.5m long or is there something special about 1.5m in length?  Also, how does one determine if a cable is "low-loss"?

Thanks for your response!

1.5m is the minimum length.  I would try to keep it shorter than 3 meters.

Low-loss means low dielectric constant in the insulation, such as Teflon, expanded Teflon, or air.  Also, to mininize skin-effect, conductors should be 26 gauge or less.

Steve N.

[audioengr]

Just for grins, I measured a new (no break-in) Off-Ramp 4 with Turboclock and USB Hynes reg and BNC output for P-P jitter.

It comes in around 170 psec for overall worst-case jitter on the S/PDIF output now.  This is a reduction of about half due to the USB Hynes reg.

Nice.  I think the Off-Ramp 5 should measure even a bit lower due to the additional Hynes regs that are designed into that one.

Steve N.

[DaveBSC]

Just for grins, I measured a new (no break-in) Off-Ramp 4 with Turboclock and USB Hynes reg and BNC output for P-P jitter.

It comes in around 170 psec for overall worst-case jitter on the S/PDIF output now.  This is a reduction of about half due to the USB Hynes reg.

Nice.  I think the Off-Ramp 5 should measure even a bit lower due to the additional Hynes regs that are designed into that one.

Steve N.

Steve, any idea how this compares to some of the competition? And what is meant by "worst case"? 170 psec seems kind of high, IIRC the Northstar USB DAC measured somewhere around 170ps.

[audioengr]

Steve, any idea how this compares to some of the competition? And what is meant by "worst case"? 170 psec seems kind of high, IIRC the Northstar USB DAC measured somewhere around 170ps.

Take a look at JAs measurements in the Nov/Dec issue of Stereophile.  He reviews the Off-Ramp.

Its important to draw a distinction between RMS and P-P jitter

Steve N.

[DaveBSC]

What I'm wondering is how Audiophilleo is managing to get 15ps p-p jitter for their devices. Are they not measuring their S/Pdif output?

[audioengr]

What I'm wondering is how Audiophilleo is managing to get 15ps p-p jitter for their devices. Are they not measuring their S/Pdif output?

I dont believe it.  Even on the bench, just measuring the clock alone driven from the very best power supply, you will not get 15psec.

It is best to make the decision based on sound quality anyway, not measurements.  Even JA realizes this now.


Steve N.

[DaveBSC]

What do you make of this measurement, comparing the Audiophilleo1 to a specialized, ultra low jitter clock, and a Grimm Audio CC1 master clock? The AP1 is right in league with these devices, suggesting that Audiophilleo is not making up their jitter measurements. I'm not making claims about sound quality, indeed from what I've read from owners it seems like the OR4 is among the very best sounding converters out there. It's just that 170ps p-p jitter on the output (or 300+ without the USB regulator) just doesn't seem that impressive of a result.



Stereophile recently measured Simaudio's Moon Evo CD player for example at 69ps p-p on its output. Granted it costs $8K, but that includes dealer markup, and its also a CD player which adds more than a little to the cost of its internal digital circuitry.

[jtwrace]

This is a pretty interesting thread about jitter.

http://www.computeraudiophile.com/content/Timing-errorjitter-reduction

[audioengr]

The plot looks good, however the different signal levels is concerning.  Is this really apples to apples?

Are we comparing clocks to clocks or S/PDIF data to clocks?

BTW,  a customer in Australia who has Audiophillio just got his Off-Ramp 4 with Turboclock, no USB Hynes reg.  He says it beats the Audiophilleo.  He is sending it back for the USB Hynes upgrade too.

Steve N.

[DaveBSC]

I think this test was measuring the clock of the AP1 vs. the other two, rather than the usual S/Pdif output. AP's specification of 15ps p-p using their WaveCrest I think was done with S/Pdif. I've heard from a few people that using bus power seems to reduce the performance of the AP quite a bit, despite the "regenerative" internal power supply. I'm amazed that the $2900(!) Sonicweld Diverter HR continues to use bus power. They obviously go to herculean lengths to try to clean it up but, why bother with all that when you can just use either batteries or a nice linear supply? I don't get it.

Do you know of any comparisons of the OR vs. the Berkeley Alpha USB? These are the two I'm most interested in. Even if the Diverter solves all of the USB power issues, the price is so high that it's silly. If you're going to spend nearly $3,000 you might as well just skip the converter altogether and buy a USB DAC.

[audioengr]

I think this test was measuring the clock of the AP1 vs. the other two, rather than the usual S/Pdif output.

That's what I suspected.  My clocks are just as good if not better.

I also have a NEL clock that I can put on my Turboclock which measures phase noise at -128dB at 10Hz.  When comparing this to my current clock from Crystek, the audible difference is small.  The Crystek is -155dB at 10kHz.

AP's specification of 15ps p-p using their WaveCrest I think was done with S/Pdif.

This may be at 10kHz.  At 10Hz, the plot puts the RMS jitter at about 500psec, and this is just the clock alone.

I've heard from a few people that using bus power seems to reduce the performance of the AP quite a bit, despite the "regenerative" internal power supply. I'm amazed that the $2900(!) Sonicweld Diverter HR continues to use bus power. They obviously go to herculean lengths to try to clean it up but, why bother with all that when you can just use either batteries or a nice linear supply? I don't get it.

Me either.  Whats the point?  Make it simpler and then charge $3K??

BTW, I discovered that my new Hynes design actually outperforms LifePO batteries.  I put both in the OR4 and compared them.

Do you know of any comparisons of the OR vs. the Berkeley Alpha USB? These are the two I'm most interested in. Even if the Diverter solves all of the USB power issues, the price is so high that it's silly. If you're going to spend nearly $3,000 you might as well just skip the converter altogether and buy a USB DAC.

Agreed.  I dont know anyone that has one yet, only the review in July/Aug TAS.  Pflaumers background is not really digital design according to the review.  Harley compares it driven by a Mac compared to a Lynx card in a PC, which is not interesting or particularly useful IMO.  It's really too bad that Steven Stone did not do this review. He would have put it head-to-head with my Off-Ramp 4.

Steve N.