Laptop to Berkeley

[earflappin]

jkeny - I sent you a private message regarding your modded HiFace.

 

[jkeny]

Just replied, David!

 

[earflappin]

Absolutely! I'll let you know when I have a modded HiFace in the house.

 

[Bob Stern]

DavidR said the HiFace is not Async. However, my interpretation of the HiFace FAQ is that it is Async:

"28)Does hiFace works in asynchronous mode?

"Sure, the transfer of data over the USB bus happens in a totally asynchronous fashion, with respect to their retransmission to the DAC, in short packets which use all the available bus bandwidth. Actually, hiFace is master in the transmission from PC to interface. As it uses its low jitter oscillators to time the S/PDIF transmission, line jitter is totally uncorrelated to data transmission from PC to interface."

 

Also, this comment says it has a transformer for galvanic isolation of the S/PDIF output:

http://www.computeraudiophile.com/content/Just-received-M2Tech-Hiface-USB-interface#comment-37100

 

[Part-Time Audiophile]

David -- did you order that modded HiFace?

 

[jkeny]

Wow, the look of this forum has changed since my last visit

 

Yes Danny & I are talking about his exact needs. The units should now be in stock in M2Tech & I will be ordering his one shortly. So look out for his review soon.

 

PS. I know Empirical Audio is launching a 192KHz USB DAC based on the HiFace. I think I've heard that Wavelength are possibly working on something.

 

[Matias]

Now I also have the ESI Juli@ and the HiFace RCA, because when I tested both, in a very high end system by the way, the HiFace was clearly a lot more detailed then the Juli@'s digital out.

That is why I bought myself a HiFace even having a Juli@ available here.

Test system was Thiel 3.7, BAT preamp, Gamut power, Accuphase player/DAC and lots of top cables:

http://i132.photobucket.com/albums/q39/mreccius/DSC_9239.jpg

 

[earflappin]

Scot, yes, I ordered JKeny's modded HiFace today, including his modded USB cable which allows the use of an external 5V power source.

 

I received a stock HiFace BNC today and am VERY impressed. It outperforms my Lynx AES16 handily reminding me of the improvements I heard when we A/B'ed your Legato. The HiFace makes the Lynx sound congested and muddy. With the HiFace there is greater resolution, extension, clarity and less glare.

 

While aural memory is fleeting, I think the HiFace is giving me better sound quality than when I had the Antelope DA in the signal path providing noise isolation for the AES16. I now have to question the prevailing wisdom that the Berkeley sounds its best via its AES input. Perhaps this was true before the advent of async USB-to-SPDIF converters where one was forced to use the SPDIF output of CD players, but the game has changed.

 

To think the modded HiFace will up the performance several levels is pretty amazing to think about. Can't wait to try it. One can only imagine what could be achieved if you could take the I2S output from the HiFace directly into the Berkeley DSP and eliminate the SPDIF-SPDIF conversion step.

 

I tried the HiFace in a couple of different configurations - (1) active extender USB cable with HiFace directly coupled to the Berkeley's BNC input, (2) directly plugged into my Zalman mobo's USB port with a 15' BNC/BNC cable and (3) same as #2, but using a USB port on the front of the Zalman than connects to the mobo with a cable. #3 sounded the best. I think this is due to: (a) a stock USB cable degrades sound quality more than a BNC/BNC cable and (b) the mobo USB port carries more noise than the remote USB port. I also A/B'ed plugging my Zalman into the wall versus my PurePower 2000. There was a slight improvement with the Zalman plugged into the PP2000. Lastly, I have found proper grounding is important - I use a star ground configuration.

 

What this has demonstrated to me is what John Kenny has stated and proven with his mods - power quality to the clock and output circuits of an AES or SPDIF converter is very important to realizing the best sound quality (this is likely a key reason why the Diverter and Legato sound so good), and secondly, properly implemented async USB is a lower jitter approach for getting the bits out of the computer.

 

 

 

[Bob Stern]

It's interesting that the HiFace performed better with a direct USB connection instead of the Active USB Extender. I would have expected the latter to be superior because it isolates the HiFace from the computer's power supply noise. This suggests that the HiFace has excellent power supply filtering to attenuate such noise. (Either that, or the Active USB Extender has a noisier power supply than the computer.)

 

I'm confused by the comment about the importance of high quality power to the S/PDIF converter. Didn't the best configuration #3 simply derive power from the computer's USB port?

 

[jkeny]

Dan,

It's great that you have a stock Hiface & you will have a my modded one to compare against. The first listen will not leave you in the position of some upgrades - wondering if the improvement is there or not? It's a huge jump in SQ. It's the sound of lower jitter As you say a cleaner supply to the clock & clock handling circuitry allows the clock's low jitter to be realised - if it has a low jitter clock So this is where the big improvement in sound comes from.

 

Improvements to the other power supply areas will also improve the sound but it's might be a case of diminishing returns here depending on the quality of downstream equipment.

 

Your experimental configurations are interesting but difficult to draw conclusions from. The Hiface has no especially good PS filtering so the PS feeding it will have a direct SQ effect. The active USB extender may well have a worse PS than your PC or it's active circuitry may be causing interference in the USB signals.

 

The idea of my USB cable is to just run the USB signals (& ground) on their own within the cable without any potential detrimental influence of a USB power line running alongside the signal lines. It also allows you to control the quality of the 5V supply by providing an external connector.

 

Mentioning I2S - I have modded my Hiface to take out I2S & feed it into a ESS DAC & the sound is amazing. I don't use the top of the line Sabre DAC (although that is next) but their DAC with in-built analogue output stage (ES9022) - this is the same as is used in the NuForce uDac which there is a lot of talk about on head-fi. I'm actually using two DACs in balanced mode & feeding it directly to balanced headphones. So it can do 24/192 in balanced mode direct to headphones. Drives AKG701s without difficulty.

 

[earflappin]

"I'm confused by the comment about the importance of high quality power to the S/PDIF converter. Didn't the best configuration #3 simply derive power from the computer's USB port?"

 

True, Bob. What I meant was based on John Kenny's findings (and others) that the stock HiFace can be made to sound much better with a high quality power supply versus the computer's power supply. Again, I suspect this is a key reason why the SonicWeld Diverter and ART Legato have been reviewed so highly versus other USB/SPDIF converters. That being said, it is pretty amazing to me how good it sounds just pulling its power from the computer USB port.

 

 

 

[earflappin]

Hi John. It's David actually and not Danny...:-)

 

Since the active USB extender cable also pulls power from the USB port how would power quality come into play? The one configuration I did not try is to use both the extender USB cable AND the bnc/bnc digital cable. Is there any reason that direct connecting the HiFace to the DAC via a male/male coupler would degrade sound quality? Anyway, I'll give it a go. Also, does the HiFace have a burn in period?

 

As I said, I can't wait to hear your modded HiFace. I know there was one or two posters (on another forum) who challenged your comment about the modded HiFace with a good DAC approaching the sound quality of really good analog. Just based on what I'm hearing now that claim doesn't sound that far fetched.

 

I've read your posts about your i2s direct connection to the Sabre. Cool. I don't know much about I2S. Some of what I've read says you must keep the wire length super short 1-2 cm. But then I see that P.S. Audio uses a long cable length and a HDMI connector. To a layperson like me it stands to reason, however, that if you can skip conversion steps (i.e. going async USB to I2S versus async USB to SPDIF and then SPDIF to I2S) you should get better sound quality.

 

That being said, the Berkeley Audio designers believe it is best to not put USB or firewire processing circuitry into the same chassis as the DAC. This is the reason they are bringing their own async USB/AES converter to market. This would seem to say they don't believe it to be optimal to bring USB straight into the DAC and do the conversion to I2S inside. Does that make any sense to you?

 

Anyway, I would love to hear the Sabre DAC sometime with the async USB-to-I2S direct interface. Does the new model have a high quality balanced output stage with variable volume control?

 

David

 

[jkeny]

Sorry David,

I haven't looked into active USB extenders - I presume they boost the USB power so as to allow longer cable lengths. This active circuitry might add noise to the PS or cause some interference with the USB signals on the adjacent wires, it's hard to know.

 

I'll let you decide on that question of the modded Hiface sound quality & I'm sure you will report your findings here.

 

It's wisest to keep my opinion to myself about other equipment that I haven't experience of.

 

[earflappin]

Your comments about the USB extender cable are plausible to me. Do you have a recommendation yet about what to use for the external USB 5V source? I was thinking about getting a high quality bench power supply with 2 configurable output voltages. I could use one for the 5V and try the other for the 3.3V comparing it to the lithium batteries. I'm also looking into some way to trickle charge the lithium batteries so one would not have to swap them in and out.

 

Sorry, I was not asking you to comment about the Berkeley (!), but rather about your experience with I2S. I know you use a very short connection path to your Sabre DAC....have you in other projects any experience as to how long one can run I2S wires without introducing audible levels of jitter? From what I've read it needs to be super short.

 

Anyway, thanks again for taking the initiative on the HiFace mod. It's going to be fun being able to directly A/B it to the stock version and some other USB/SPDIF converters.

 

[jkeny]

Your bench supply sounds like a good general purpose tool & can be used for a number of other functions too. I've tried transformer & battery 5V supplies but haven't really had a chance to do A/B tests. I know lots of people report Paul Hynes power supplies are excellent. He does a 3 leg regulator replacement for 30 or 36Euro.

 

About I2S, I know the quality of the signal is sensitive but I haven't worked out a maximum operational length.

 

[audiozorro]

Your comment that the Berkeley Audio designers believe it is best to not put USB or firewire processing circuitry into the same chassis as the DAC is one I would expect Wavelength or Weiss to disagree with.

 

Do you know if Berkeley Audio ever tried to implement USB into a variant of their Alpha DAC? It would seem to me that the Alpha DAC just predates 24/192 async USB so their new, yet to be released, USB/SPDIF device conveniently fits the need.

 

[earflappin]

audiozorro, excellent point. I agree with you.

 

No, I don't know if they ran tests during their development process for the Alpha DAC involving USB or Firewire. My take is this. A key design goal for them was 24/192 support. When they designed the Alpha DAC, 24/192 USB was not available and going Firewire would have involved a long, costly development program and Mac only transport support. AES, S/PDIF and toslink were well accepted standards and interfaces they had strong experience with at Pacific Microsonics.

 

The technology has clearly evolved since the Alpha DAC release and now they have a large installed base. So offering a standalone async USB2-to-AES3 converter makes sense - they keep their customers happy and create a new revenue stream. I have to believe they are in parallel looking at developing a second gen DAC which more fully takes advantage of the async USB interface. While I'm clearly no digital engineer I would have to believe one is leaving some performance on the table going async USB-to-AES-AES-to-I2S-to-DSP. Maybe not. Perhaps they have done the testing and have truly found the outboard conversion approach is best for sound quality. What I do know is these guys are very smart and quiet about what they are actually doing....:-). I would be as well.

 

What I can also say after more listening, at least in my system which I have carefully constructed to be highly resolving and neutral, is the sound quality improvement moving from the Lynx AES16 to the stock HiFace is significant (and that gap is going to widen with the modded HiFace).

 

Perhaps in other systems the gap would be regarded as less significant. I am really eager to hear from other Berkeley owners on this point, particularly those who use other AES transports which have less jitter and noise than the Lynx. Specifically, I would love to reprise the ULN-8 vs Berkeley Alpha DAC shoot out using the modded HiFace interface. It could well be that the biggest reason for the observed sonic differences had to do with the interface and not so much the DAC itself.

 

Lastly, does anyone know why AES would be preferred over S/PDIF BNC where you are dealing with long cable runs? From what I've read, AES cannot be true 110 ohms and since this is a transmission medium this will lead to sonic degradation (I mean just look at an XLR connector, let alone a Lynx DB26 connector !). Second, I believe the AES specification is 3-7V (many use 5V I think) and at that level and with speeds up to 12.288MHz the drivers have to be very powerful which means they will typically overshoot and cause reception errors - right?

 

Bottom line, I think these async USB/SPDIF converters are going to make a lot of "legacy" DAC owners very happy and M2Tech have set a new standard in terms of price/performance that is great for the audiophile community.

 

 

 

[earflappin]

John, I see that Paul Hynes offers an SR1-3.3V 1.5amp and an SR1-5V 1.25amp supply (not cheap at 180 English Pounds each !), but I take it these take 120V/60Hz input power and could be used for the modded HiFace? Do you know of any off the shelf trickle charger for the lithium batteries or would one have to design that as a DIY project?

 

Thanks for helping to educate a newbie on power supply technology. As an aside, I would love to have someone look at the Berkeley power supply system to see if there might be some incremental performance to be had by upgraded to PH's shunt regulators.

 

 

 

[jkeny]

Yes David, but you may not need the 1.5Amp versions - I'll have to do some checking on the current requirements needed for the 5V supply to the HiFace but you won't need anything like 1.25Amp.

 

I've just done the calcs from the datasheets of the chips & it looks like 95mA max so it looks like the 30pound 3 leg regulators will do the job Z7805 or Z1705 would do but you would need to check with Paul. See here for specs http://www.paulhynesdesign.com/page6.html

 

Yes, I think a LiFePO$ trickle charger will have to be DIY project.

 

[manisandher]

"... not cheap at 180 English Pounds each !"

 

I believe Paul lives in Scotland. You may want to call them British pounds!

 

Mani.

 

[jkeny]

I've set up a basic web-site for this modified HiFace & will be adding the details of the modifications for those that want to do this themselves:

Home (Hiface Mods)

 

[earflappin]

John, love the way you have boxed it up into an easy to use and professional looking device. Well done. Hope you sell a ton of them.

 

[jkeny]

Thanks Daniel

Boxed unit contains 2 batteries, modded Hiface & connectors specified below. It's about the same dimensions as an iPhone but thicker

 

Front shows USB-in connector

Back shows BNC-out, On/Off switch, socket for charger

 

Pics:

 

 

 

 

 

 

 

[Part-Time Audiophile]

how does it SOUND?

 

Anyone have one of these to play with yet? I thought David (earflappin) was going to get one of these little boxes of sweetness?

 

[earflappin]

Scot, I had one on order, but cancelled it because I have sold my Berkeley Alpha DAC and Lynx AES16 and moved on to a MH LIO-8 which uses an async firewire interface. I believe Dan H. is getting one to use with his Berkeley Alpha DAC and I'm sure he will post his impressions once he gets his.