The link to the review is posted on the Berkeley site, under the News page. Review is a rave. WDW
Absolute Sound Review Of Alpha USB on-line
- Thread starter WDW
- Start date
Thanks for spotting that. Here is the copy of the review amongst other products in the new guide: http://media.avguide.com/Digital_Source_Components_Buyers_Guide.pdf. Once you open it, click on the alpha USB link.
The review is of course very positive although I was puzzled by him saying he found a difference between a PC and Mac source. The point of an asynchronous and isolated bridge is to eliminate such differences. I will have to ponder what remains to explain the differences Robert Harley perceived.
The review is of course very positive although I was puzzled by him saying he found a difference between a PC and Mac source. The point of an asynchronous and isolated bridge is to eliminate such differences. I will have to ponder what remains to explain the differences Robert Harley perceived.
Not surprising that the BADA was improved with their own low jitter USB-SPDIF converter.
I heard a $70K DAC today that was improved with a $1K low jitter USB-SPDIF converter.
Amir, have you heard the Alpha USB? I would be interested in a PC vs. Mac source comparison from you.
I heard a $70K DAC today that was improved with a $1K low jitter USB-SPDIF converter.
Amir, have you heard the Alpha USB? I would be interested in a PC vs. Mac source comparison from you.
Not yet. We put an order in immediately but still waiting to get ours. As soon as I get it, I will put it through its paces including comparing it to lower cost products.
And yes, it is so counterintuitive that even expensive DACs improve with a low jitter source.
And yes, it is so counterintuitive that even expensive DACs improve with a low jitter source.
Asynch USB mode and PC/Mac noise floors
Here's some ideas...
From what I've heard subjectively and measured quantitatively, laptops and their hideous switchers are incredible noise dispensers, unfortunately, as demonstrated by looking at what happens at the output of even high-end DACs with an APx515. The noise levels vary widely within a given brand also. My Toshiba Qosmio is well-behaved; an Acer notebook, otherwise a delightful computer, is pretty dreadful. It's about 15 dB noisier than the AES output of the APx, for example. Yucko.
BTW while doing this work, it was fun to see the results of turning dithering on and off in the S/PDIF test signals. It's right out of a textbook: The spectral spurs go down, but the noise floor rises... just like in all the papers explained the technique. It really works!
Anyway, my understanding is what asynch mode accomplishes is allowing one to use internal clocks in the USB-S/PDIF transport to generate the clocks in the S/PDIF, thus feeding the downstream DAC much lower jitter inputs. With (mal)adaptive mode, the framing clocks of the USB are used, with generally poor results in comparison. Apart from that, galvanic isolation is critical for keeping the laptop from contaminating everything.
We expect to have some of the Auraliti mini-servers in shortly, and from the reports of other listeners the lack of 90% of the stuff inside a typical PC/Mac reduces the noise floor considerably.
Also, Robert has in the past used a custom PC server built by Goodwin's, which may be very quiet noise-wise. Haven't had a chance to read the review yet, but this may still be what he uses.
Here's some ideas...
From what I've heard subjectively and measured quantitatively, laptops and their hideous switchers are incredible noise dispensers, unfortunately, as demonstrated by looking at what happens at the output of even high-end DACs with an APx515. The noise levels vary widely within a given brand also. My Toshiba Qosmio is well-behaved; an Acer notebook, otherwise a delightful computer, is pretty dreadful. It's about 15 dB noisier than the AES output of the APx, for example. Yucko.
BTW while doing this work, it was fun to see the results of turning dithering on and off in the S/PDIF test signals. It's right out of a textbook: The spectral spurs go down, but the noise floor rises... just like in all the papers explained the technique. It really works!
Anyway, my understanding is what asynch mode accomplishes is allowing one to use internal clocks in the USB-S/PDIF transport to generate the clocks in the S/PDIF, thus feeding the downstream DAC much lower jitter inputs. With (mal)adaptive mode, the framing clocks of the USB are used, with generally poor results in comparison. Apart from that, galvanic isolation is critical for keeping the laptop from contaminating everything.
We expect to have some of the Auraliti mini-servers in shortly, and from the reports of other listeners the lack of 90% of the stuff inside a typical PC/Mac reduces the noise floor considerably.
Also, Robert has in the past used a custom PC server built by Goodwin's, which may be very quiet noise-wise. Haven't had a chance to read the review yet, but this may still be what he uses.
Low jitter USB transport to the rescue...
The GTE Audio Trinity DACs provide a USB-S/PDIF transport as an accessory for computer source enthusiasts, so one can try different manufacturer's devices.
When Brand X (no real phase noise or jitter specs published) was replaced with Brand Y (which Madrona Digital just happens to sell
) that has very low phase noise, the results are quite noticeable. Brand X is probably 50 ps period jitter, who know what the phase noise it), while Brand Y is under 5 ps period jitter (which is at 44 kHz; at 176 the random jitter as determined by the PN is in the femtoseconds).
With the Trinity DACs, the S/PDIF hits a PLL. My preferred architecture is to forget the clocks on the S/PDIF; just put everything into a big buffer memory and clock it out with OXCOs inside the DAC: Guaranteed zero sensitivity to external clocks, because the DAC is completely disconnected from them. This is what MSB does on its higher-end products, and also may be what TAD does.
So, yes, as Dan says, the $495 Brand Y USB transport greatly improves the performance of the Trinity DAC. This is one of these mouse helping a lion stories. Another way of saying this is that with really, really low jitter clock inputs, the full potential of the Trinity is revealed, and it's pretty impressive. There's a soundfield that subtends about 60 degrees of arc horizontally, and goosebump-inducing realism. It will be interesting how the whole thing sounds when the Acoustic Geometry diffusers and traps go in.
It's possible, remotely, that there are some issues relating the PC versus the Mac drivers.
The GTE Audio Trinity DACs provide a USB-S/PDIF transport as an accessory for computer source enthusiasts, so one can try different manufacturer's devices.
When Brand X (no real phase noise or jitter specs published) was replaced with Brand Y (which Madrona Digital just happens to sell
) that has very low phase noise, the results are quite noticeable. Brand X is probably 50 ps period jitter, who know what the phase noise it), while Brand Y is under 5 ps period jitter (which is at 44 kHz; at 176 the random jitter as determined by the PN is in the femtoseconds).With the Trinity DACs, the S/PDIF hits a PLL. My preferred architecture is to forget the clocks on the S/PDIF; just put everything into a big buffer memory and clock it out with OXCOs inside the DAC: Guaranteed zero sensitivity to external clocks, because the DAC is completely disconnected from them. This is what MSB does on its higher-end products, and also may be what TAD does.
So, yes, as Dan says, the $495 Brand Y USB transport greatly improves the performance of the Trinity DAC. This is one of these mouse helping a lion stories.
Another way of saying this is that with really, really low jitter clock inputs, the full potential of the Trinity is revealed, and it's pretty impressive. There's a soundfield that subtends about 60 degrees of arc horizontally, and goosebump-inducing realism. It will be interesting how the whole thing sounds when the Acoustic Geometry diffusers and traps go in.It's possible, remotely, that there are some issues relating the PC versus the Mac drivers.
You have a gift for highjacking a thread...my post to the Absolute Sound was quickly replaced with the photo image of you in your SUV....are you an important person or are you selling something....this thread is now looking at you and your parrot.
I'm a simple man, please explain.
WDW
I'm a simple man, please explain.
WDW
Key design issues of USB-S/PDIF interfaces
@WDW... You must be wondering why Dan, Amir and I responded to your post about a review going up, and to each other, in the way we did. The main thing is that you'd like to learn, so here goes...
Basically, one of the most fundamental issues in making "digital" sound well is phase noise, meaning the accuracy of the clocking signals. If one uses a S/PDIF or AES interface, there are already a lot of issues. If the S/PDIF driving a DAC is in turn generated by a USB interface, there are even more issues. Phase noise analysis may also help visualize the possible origins of various types of time-domain jitter.
The Berkeley Audio Design alpha DAC is of course an iconic product, and the fellows behind it are both extraordinarily skilled and demanding, thus it took them a rather long time to finally evolve an interface that they thought was satisfactory.
So, one of the key problems any USB-S/PDIF interface designer has to solve is having an S/PDIF clock with very low jitter and phase noise, thus the possibly unexpected jump into a "deep dive" on that topic. From the perspective of, well, experts in this area, the sequence makes perfect sense, although not, perhaps, to more normal folk.
Thus, a simple announcement of the new interface immediately triggers a discussion about the arcana of phase noise and $70K DACs. And Amir raised the interesting observation that RH thought PCs sounded differently than Macs, which could be either a phase noise or a galvanic isolation issue. More discussion about obscure details ensued.
Some of us on the forum have worked together over the years, know each other well in other contexts, etc., thus the assumed context and offering of reports on how other USB-S/PDIF interfaces sound (Dan), Amir's remarks about differences between platforms, and my responses to Amir's and Dan's comments.
Ergo, all the posts in this thread so far are spot-on with respect to the interesting news that you posted, namely the first review of the long-anticipated advent of what imay well be one of the best 3-4 USB interfaces available out of a couple of dozen products.
Hope this helps!
@WDW... You must be wondering why Dan, Amir and I responded to your post about a review going up, and to each other, in the way we did. The main thing is that you'd like to learn, so here goes...
Basically, one of the most fundamental issues in making "digital" sound well is phase noise, meaning the accuracy of the clocking signals. If one uses a S/PDIF or AES interface, there are already a lot of issues. If the S/PDIF driving a DAC is in turn generated by a USB interface, there are even more issues. Phase noise analysis may also help visualize the possible origins of various types of time-domain jitter.
The Berkeley Audio Design alpha DAC is of course an iconic product, and the fellows behind it are both extraordinarily skilled and demanding, thus it took them a rather long time to finally evolve an interface that they thought was satisfactory.
So, one of the key problems any USB-S/PDIF interface designer has to solve is having an S/PDIF clock with very low jitter and phase noise, thus the possibly unexpected jump into a "deep dive" on that topic. From the perspective of, well, experts in this area, the sequence makes perfect sense, although not, perhaps, to more normal folk.
Thus, a simple announcement of the new interface immediately triggers a discussion about the arcana of phase noise and $70K DACs. And Amir raised the interesting observation that RH thought PCs sounded differently than Macs, which could be either a phase noise or a galvanic isolation issue. More discussion about obscure details ensued.
Some of us on the forum have worked together over the years, know each other well in other contexts, etc., thus the assumed context and offering of reports on how other USB-S/PDIF interfaces sound (Dan), Amir's remarks about differences between platforms, and my responses to Amir's and Dan's comments.
Ergo, all the posts in this thread so far are spot-on with respect to the interesting news that you posted, namely the first review of the long-anticipated advent of what imay well be one of the best 3-4 USB interfaces available out of a couple of dozen products.
Hope this helps!
Last edited:
parrotman
Wasn't asking for help....WDW@WDW... You must be wondering why Dan, Amir and I responded to your post about a review going up, and to each other, in the way we did. The main thing is that you'd like to learn, so here goes...
Basically, one of the most fundamental issues in making "digital" sound well is phase noise, meaning the accuracy of the clocking signals. If one uses a S/PDIF or AES interface, there are already a lot of issues. If the S/PDIF driving a DAC is in turn generated by a USB interface, there are even more issues. Phase noise analysis may also help visualize the possible origins of various types of time-domain jitter.
The Berkeley Audio Design alpha DAC is of course an iconic product, and the fellows behind it are both extraordinarily skilled and demanding, thus it took them a rather long time to finally evolve an interface that they thought was satisfactory.
So, one of the key problems any USB-S/PDIF interface designer has to solve is having an S/PDIF clock with very low jitter and phase noise, thus the possibly unexpected jump into a "deep dive" on that topic. From the perspective of, well, experts in this area, the sequence makes perfect sense, although not, perhaps, to more normal folk.
Thus, a simple announcement of the new interface immediately triggers a discussion about the arcana of phase noise and $70K DACs. And Amir raised the interesting observation that RH thought PCs sounded differently than Macs, which could be either a phase noise or a galvanic isolation issue. More discussion about obscure details ensued.
Some of us on the forum have worked together over the years, know each other well in other contexts, etc., thus the assumed context and offering of reports on how other USB-S/PDIF interfaces sound (Dan), Amir's remarks about differences between platforms, and my responses to Amir's and Dan's comments.
Ergo, all the posts in this thread so far are spot-on with respect to the interesting news that you posted, namely the first review of the long-anticipated advent of what imay well be one of the best 3-4 USB interfaces available out of a couple of dozen products.
Hope this helps!
Come on, guys, let's keep this exchange civil please. Thread drift is allowable when a topic stimulates an interesting discussion about similar matters.
Lee
Lee
Pure Music
If I'm not mistaken, the two trials were between his Goodwin Zalwin/Korean based noise cancelled PC box and a run of the mill iMac.
To respond to your surprise of Harley's noticable preference a Mac before a PC server
, this was due to one major difference between these two platforms. The Mac was running Pure Music.
If I'm not mistaken, the two trials were between his Goodwin Zalwin/Korean based noise cancelled PC box and a run of the mill iMac.
To respond to your surprise of Harley's noticable preference a Mac before a PC server
, this was due to one major difference between these two platforms. The Mac was running Pure Music.
If an asynchronous USB device that is fully isolated doesn't remove the difference between PCs and Media Players on them, then there is some unknown science I can't explain .
Maybe he has PC set up so that it doesn't output bit-exact. That would make a difference. But if the PC is bit exact, then there should not be any difference as any timing difference would be taken care of by the USB to S/PDIF bridge.
Indeed, it would be bad news if the best hardware out there can't get rid of this difference. It means we don't really know what to do next!
.Maybe he has PC set up so that it doesn't output bit-exact. That would make a difference. But if the PC is bit exact, then there should not be any difference as any timing difference would be taken care of by the USB to S/PDIF bridge.
Indeed, it would be bad news if the best hardware out there can't get rid of this difference. It means we don't really know what to do next!
Isn't safe to say that this issue isn't hardware..WDW
The memory management functions within Pure Music are doing some musically wonderful things....My profession is technical but by no means electrical/digital but I suspect this work is all to do with lowering the noise floor...writing into memory and
spooling forward. Parrotman made a similar comment about dumping all of this into ram and clocking from there forward.
If an asynchronous USB device that is fully isolated doesn't remove the difference between PCs and Media Players on them, then there is some unknown science I can't explain
Maybe he has PC set up so that it doesn't output bit-exact. That would make a difference. But if the PC is bit exact, then there should not be any difference as any timing difference would be taken care of by the USB to S/PDIF bridge.
Indeed, it would be bad news if the best hardware out there can't get rid of this difference. It means we don't really know what to do next!
The memory management functions within Pure Music are doing some musically wonderful things....My profession is technical but by no means electrical/digital but I suspect this work is all to do with lowering the noise floor...writing into memory and
spooling forward. Parrotman made a similar comment about dumping all of this into ram and clocking from there forward.
Those features help when you don't have an asynchronous connection and the PC is generating timing. To the extent the Alpha is doing that now, it should not matter.
Amir
A nagging question...I've recently read that there will be an update to the mac mini which will, rumour has it, have the new Thunderbolt protocol.....this tells me that firewire is gone. Anyone considering a Weiss DAc should pause. But it follows that USB will follow suit...admittedly on a far longer time frame but will, most certainly, follow. So my soon to be ordered Alpha DAC may have an inappropriately short window of utility as some marketing group in Cupertino decides to move on.
Enjoy this weekend WDW
A nagging question...I've recently read that there will be an update to the mac mini which will, rumour has it, have the new Thunderbolt protocol.....this tells me that firewire is gone. Anyone considering a Weiss DAc should pause. But it follows that USB will follow suit...admittedly on a far longer time frame but will, most certainly, follow. So my soon to be ordered Alpha DAC may have an inappropriately short window of utility as some marketing group in Cupertino decides to move on.
Enjoy this weekend WDW
Oh, I would not be worried about that at all. For one thing, USB will be supported on Macs for many years to come due to breadth of peripherals in that format and the little it costs to implement vs Thunderbolt. When and if the time comes that Thunderbolt is the only connector, there will be adapters (really bridges) which would convert it to USB.
You have a great weekend too. Looks like we may have some sunshine!
You have a great weekend too. Looks like we may have some sunshine!
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |