Back in July, Lonely Raven wrote on the USB cable thread that he was happy with a Pangea silver. By chance, I had to send my system SSD back under warranty and cloned another to fill in. When I went to play the DS (JPlay, foobar), I discovered it wasn’t showing in Windows Sound. The driver showed as installed but the dac as not connected. After unplugging and replugging my highly regarded $400+ TotalDac1 USB cable a few times and getting nowhere, I switched to the $30 silver plated copper Pangea and voila, everything worked. Then switching back, the TotalDac worked. Go figure. But now I had a sound comparison. Going back and forth on a few good PCM CDs and FLAC files, I found telling the difference between the two took a lot of careful listening. There is a difference, but at this point, with more listening to do, I can’t quite tell if it’s more a difference in listener preference, as opposed to one in sound quality (be easier with a helper).
Paul McGowan said My favorite still remains Paul Pang's work. You can get it here for 200 pounds - not sure what that is in dollars, probably around $300?Paul, do you prefer that over a network connection with the Bridge?
Am mainly using my Directstream via the Bridge, but I did try it with Kimber B-bus CU and the cheaper Pangea USB cable, both seemed to work without any problems. I wasn’t really listening for sound quality that way so don’t really have any comparisons to make, but if anyone wants I can try again.
It really depends on a number of factors but in general, I prefer the sound of the network connection to the sound of the USB - but only by a little and it really depends on the USB cable. Lots of variables. For my own setup I use both: USB and network. I mostly use USB because I have a Mac Mini tied into the DS through USB that allows me to use iTunes as the interface on my iPad mini. Being lazy and a sucker for Apple interfaces, the ease of use of the iTunes interface outweighs the small difference in sound quality between the two. So mostly, that’s what I do.
Your mileage may vary.
aliaswolf saidPaul McGowan said My favorite still remains Paul Pang's work. You can get it here for 200 pounds - not sure what that is in dollars, probably around $300?It took me three months and 300€ to follow that recommendation. Finally, I got it.
The jCat reduces the gap between the I2S inputs (PWT and bridge) remarkably and allows me now to enjoy the convenience of jRiver/jRemote for gapless playback until the bridge II will come up.
Hi
That PPA link points to my site but I don’t think you ordered from me (if you did please pm me your order number), an order from Germany (or anywhere really) is usually between 10 and 20 days depending on how busy Pang is and whether I have stock and customs (orders received by me are sent the day I get them by Royal Mail International Tracked and Signed and a tracking number is provided).
I’ve never sent a JCAT cable or card to Germany so I’m not sure what your first paragraph relates to. I can’t control shipping but 3 months would be excessive by any measure and I monitor orders and would have been chasing it up well before then, I’ve never received a request from Germany to look into a late delivery either so I’m confused by your post and the link to my site.
Has anyone else tried the IFi IUSBpower and Gemini cable combo (or separate)?. I could not get my set to work on my Directstream. (My Audioquest Carbon works fine). Neither the IUSBpower, Gemini USB or combination works for me.
It used to to work fine on my MKII.
(I contacted Ifi support - not much help there).
Yes IFI power and Gemini cable work fine for me I first used it from an IMac to the DSD, now I use it from an Antipodes DS Reference to the DSD no problem. I am going to order a JCat so I can compare the two. The Antipodes is fantastic using the Ipeng as my interface.
I just received the ifi power supply and the conditioner. I had to turn the Direct Stream off and then on at the back, then the Direct Stream recognized the usb.
At first playing Hi Rez files I thought it sounded pretty good. But, when I switched to some regular files, I got a sound that was distorted and speeded up. (A glitch?)
So, I turned the Direct Stream off at the back and then back on again and everything is playing ok.
I’m not really sure that the sound quality is better, it perhaps sounds a little worse. I guess I’m going to have to let it play for awhile be for I make final judgment.
If that glitch happens again I will definitely be concerned.
I agree with your assessment the Antipodes DS Reference has a powered and non powered hub, I removed the IFI power to the DSD I think it sounds better without it in the system. I use it with my IFI IDSD for headphones where it sounds fine. I do think their 99.99 USB cable sounds great, my only frame of reference is a cardas USB cable I have been using.
Has anyone directly compared either the JCat or the JCat Reference to either Wireworld Platinum 7 or Audioquest Diamond? I’ve heard the latter two and I am curios about the JCat. Unfortunately the JCat would be a blind purchase on recommendation.
Any thoughts would be greatly appreciated.
Other folks have mentioned Pangea’s pure silver USB cable, and I have to agree. For $99 1m length, Pangea Audio’s 24-gauge pure silver USB cable is going to be one of the best values in terms of price/performance. I use this to connect my Auralic Aries to my DirecStream and found a huge improvement in sound quality over the stock USB cables I used before. I also noticed an improvement in the sound over a full week of continuous use. The improvement in sound quality from my USB cable upgrade was so great, you really need to consider a must-have vs. nice-to-have if you are investing the $$$ for a DirectStream.
On a related note, in another thread I shared positive results I got upgrading my ethernet cabling from Cat 5e to Cat 7. While the exact “why” an upgraded USB cable or ethernet cable improves the sound makes for lively discussion, there is no denying that improving the quality of digital cabling (be it USB or ethernet) has a very audible, and in my experience, significant positive impact on sound quality. The ears don’t lie!
I can confirm that JCAT Reference USB Cable 1.0m works fine between the DS and Aries.
Even without need to use the second (power) thread of the cable (as DS USB interface does not require 5V power to operate).
+1
I did get my JCAT within a week to Germany. Ordered direct at JCAT.
lestes saidComments below are from AudioStream reviews (Stereophile's sister site for computer audio).Has anyone directly compared either the JCat or the JCat Reference to either Wireworld Platinum 7 or Audioquest Diamond? I’ve heard the latter two and I am curios about the JCat. Unfortunately the JCat would be a blind purchase on recommendation.
Any thoughts would be greatly appreciated.
PS Audio DirectStream DAC review (Michael Lavorgna):
I will also mention that I greatly preferred the DirectStream DAC when paired with the AudioQuest Diamond USB cable. I initially used the Light Harmonic Lightspeed USB cable and found the overall performance to be a tad flat. This perceived flatness was not a factor with the Diamond USB cable and I would recommend this particular pairing for DirectStream buyers.
JCAT USB cable review (Steven Plaskin):
One of my favorite USB cables is the Light Harmonic LightSpeed. The JCAT has a richer sounding midrange and more prominent bass. The LightSpeed has more detail at the high end and has a bit more midrange presence than the JCAT. The LightSpeed has the largest soundstage of any USB cable I have heard.
Another USB cable I like is the Audioquest Diamond. The JCAT is a warmer sounding cable with a more prominent low end. The JCAT has a slightly larger soundstage.
JCAT Reference USB cable review (Steven Plaskin):
Other USB cables used in this review were the Synergistic Research Galileo LE, JCAT USB cable, and the Audioquest Diamond USB cable. …
For those that own the JCAT USB cable that I previously reviewed, the new cable builds on the strengths of the JCAT USB, but goes far beyond it in sonic performance. …
The JCAT Reference is the best balanced of any USB cable I have ever heard. There is absolutely no accentuation of highs, midrange, or lows with this cable. The bass is extremely articulate and tight sounding. But it can deliver plenty of punch when called upon to do so. The midrange has the essential quality of the JCAT USB cable without the extra warmth found in this cable.
“On a related note, in another thread I shared positive results I got upgrading my ethernet cabling from Cat 5e to Cat 7. While the exact “why” an upgraded USB cable or ethernet cable improves the sound makes for lively discussion, there is no denying that improving the quality of digital cabling (be it USB or ethernet) has a very audible, and in my experience, significant positive impact on sound quality. The ears don’t lie!” - jtrimm
Yes, I’ve had the same experience with both ethernet and USB. I have recently learned a great deal about cabling for analog signals from a fellow who used his ears, and 30+ years of engineering electronics cabling, to design a new line of audio cables. I’ll ask him about why digital cable would make a difference. One thing for sure, what appears to be a simple bit of plastic and copper has a whole lot going on under the hood the instant that you send a signal down it. There is no unknown “magic,” but the interaction of known variables is quite complex and in the audio sphere you can’t fully predict the outcome. You have to use your ears as the primary design tool.
Cinnamaker said Comments below are from AudioStream reviews (Stereophile's sister site for computer audio).Thank you for the reply[Bunch of quotes deleted]
Since my initial question I have bought, and now use the JCAT Reference. While I no longer have the Wireworld 7, I do own the Audioquest Diamond. I’d say the JCAT is more inline with the Wireworld 7, but a bit more detail and timber…however memory can be a fleeting thing. Regardless I am quite happy and no longer looking.
The Diamond is now overkill in my office computer system. I think the cable is more than I spend on the speakers, DAC and JPlay but who’s tracking it really?
Here is Galen Gareis’ response to my question about why digital cabling might cause a difference in the sound quality. Those of you familiar with networking will see a lot of review here but the discussion is really about the cabling. It was mostly new information for me (understatement).
"Will,
Ethernet is most driven by what is called the ACR ratio. This is a means to measure the “signal” level OVER the noise floor in the cable. The noise floor is everything but the signal (cross talk, shot noise ETC). So the larger the signal over the noise, the less the NIC card has to say, “what’s that again?” and request an error correction packet re-transmission after it detects an error.
Remember, serial digital audio is NOT Ethernet.
I need to dig into what’s really being used to build the right cable for audio. There also seems to be mixed signals (pun intended) on devices as to format choices. One cable cannot address the differing schemes.
Some of the issues for “sound” quality are what happens when the CRC detects an error, and the ARQ “correction” steps start and this impacts the audio quality. If the corrections steps are never started, the overall channel can sound much better. Audio wants to avoid any “rebuilding” steps if possible to remove the distortions caused by the interpolations and wait periods that can time out entirely and lose information.
Older audio circuits would use the previous information cache of so many mille-seconds until readable new data arrived, mitigating data apparent loss with no “correction” steps. Audio can’t wait like a download to print a picture that needs all of the data correct.
I’m not a hardware layer code expert at all, but you don’t want the signal to fall below the noise floor for audio, or anywhere near it. Modern NIC cards can see the data several dB BELOW the noise, believe it or not, and assemble a readable signal. This is frighteningly amazing as the signal amplitude is near nothing by that point. But, this is data that can look more asynchronous in arrival times where audio wants a consistent synchronous clock if possible.
This error DETECTION scheme is called CRC - Cyclic Redundancy Check;
CRC-16 (99.998% effective for errors greater than 16 bits in length)
CRC-32 (99.99999998%for errors greater than 32 bits long) are in common use today.
If an error is DETECTED through CRC checks, is tries to CORRECT it using CSMA/CD - Carrier Sense Multiple Access w/ Collision Detect. When a frame is so goofy it needs a retransmission, it use a process called ARQ – Automatic Repeat Request.
You can turn on or off error DETECTION, but it’s not advisable to do so except where lengths and signal strengths are high (never detect an error needing retransmission). The default setting for Ethernet is to use error detection without error correction. Bad frames are simply discarded and a new one is requested to be sent.
If error correction is turned “on” it uses a stop and wait ARQ methodology mentioned above. It is just that, it pings the sender for a new packet and WAITS for it to arrive. If it gets one, and it is good, it sends an ACK acknowledgement packet to the sender. If no good it sends a NAK negative acknowledgement packet to the sender and this goes on UNTIL it finally gets a good one or times out (never going to happen).
CAT7 cable is simply a very high ACR design in that each pair is shielded with foil, and the four foil shielded pairs are then braid shielded as well. This forms a VERY high ACR (signal level is well above the noise) digital line. Larger 22 AWG copper, and low loss foamed dielectrics are used to improve the signal “level” at a given distance as well as pushing the noise floor down to -85 dB to -100 dB with shielding. This is a brute force method that does not utilize electromagnetic finesse through lay length optimization required for UTP and FTP cables to achieve worst pair NEXT control.
Ethernet uses a PAM – Pulse Code Modulation voltage level “digital” scheme that allows multiple voltage levels and uses all four pairs in full-duplex mode (sends data BOTH ways all the time). This is how a Ethernet can send more digital data, it has multiple voltage levels and uses all four pairs concurrently as you go UP in speed versions. 10baseT only uses two pairs for instance with minimal voltage levels where 1000baseT on up uses all four with higher voltage level PAM encoding.
Getting to practical data rate limits; all the data isn’t really “data”;
nominal data rate (e.g., 10Mbps for 10BaseT)
error rate, since this determines the frame retransmission rate
efficiency of the data link protocol which depends on the percentage of the transmission devoted to overhead
efficiency of the media access control protocol meaning how effective is the protocol at making use of the nominal data rate.
The SIZE of the frame (small (82% efficient) to large (98% efficient)) effect the efficiency as well, and the stream is inconsistent. And now we have SHARED Ethernet where we see practical efficiencies go to ~ 50% or so based on what is called MAC – Media Access Control protocol methodology. SHARED is limited by random network traffic where SWITCHED Ethernet doesn’t have this issue and this is why switched is best for speed efficiency. For example a sharedlink with at 97% data link protocol efficiency, 10BaseT can carry 0.97 x 0.5 x 10mbps = 4.85Mbps.
So what does all this mean? For Ethernet the easy answer is you want NO retransmissions (high ACR on the cable) and a switched link that does not share itself with others and use larger packets.
USB is uncompressed serial data stream that USES Ethernet cables…but NOT Ethernet protocol (no error detection or correction), so we are left with the fundamental signal to noise characteristics of our cable (ACR ratio) to manage an uncompressed digital stream where errors are NOT corrected (we see / hear a glitch in the data).
There are tricks that can be used to improve short length digital lengths that cannot be used in LONG haul 100-meter Ethernet cable design.This isn’t magic either but optimization of variables based on cable length. We optimize Ethernet for 100 meters per standards. Raw digital streams have flexibility to optimize to shorter lengths to improve eye pattern performance. Remember I told you, and showed you how some variables can be too good, and detract from the overall end game process if let go unchecked. If I have “patch” type cables, and no TIA / EIA or UL rules, I can make some changes to improve the performance over a standardized 100 meter requirement length.
These tricks improve what is called an eye-pattern test, which is a graphical representation of the digital signal ability to, well, stay digital, and not get into a rounded off rise-time / fall time limited mess. So a modified digital cable can be smaller (cat 7 ISTP – individually shielded twisted pair cable is brute force technology @ 0.320” diameter), and STILL allow high error free speeds. The fact that the ACR ratio is NOT linear across the frequency spectrum (more ACR margins as you go lower in frequency) means that a means to distribute the energy more uniformly to enhance ACR linearity improves the data rate eye pattern test. This gets all the bits moving in a more uniform referenced to noise manner sharpening the rise / fall time energy consistency.
This all ignores the internal workings of the critical DAC, by the way! Two DAC’s with the exact same data make differing analog re-build interpretations.
If Belden’s analog cables are satisfactory, then Belden may THEN send digital around the horn, too. There is no sense in having a high quality digital link when the analog DAC signal going onto your analog cables just gets screwed up. Analog has to be first as this is the FINAL answer to your human ears.
So in a nut shell, USB is pure uncompressed digital and has no means to fix goofs. Ethernet can, but must be highly buffered in high quality audio so the ARQ automatic repeat request for error correction can’t impact consistent packet arrivals. You want a smooth highly efficient stream of data, and Ethernet is DESIGNED to allow a huge MESS of data to be managed with elaborate detection and correction schemes. This is great for buffered systems that don’t rely on linear timing like our ears. Our audible world is pure uncompressed data with no error correction.
An ideal uncompressed digital link would have;
High natural ACR through lay set design.
Linearity improved ACR with respect to frequency.
Smallest cable size to achieve an “open-eye” percentage above 90%.
Managed DESIGN control on impedance under flex."
Thanks Will, that was, uh, enlightening. Not sure I understood much of it though, I had no real knowledge of how ethernet and USB work, but Galen sure does! I think I will go sit in the corner and let my brain explode …
Hehe, yeah, lots of info here. But after several read-throughs what he is saying gives ample room for someone as bright as he to be able to work some “not-magic” to optimize transmission for audio in the digital realm. He has not put a lot of thought into this yet but has designed interconnects and speaker cable that are, gulp, um, really good let’s say.
I just hope that Belden brings them to the market!
Well, his write up impressed me! Not that that counts for much … I’ve read a few cable white papers over the years and Galen’s explanation seems as good (or better) as any, and far better than quite a few. He seems to really understand the science and how to make it improve the product. I think Belden needs to send out a batch of the interconnect and speaker cables to the PS Audio forum members for some evaluation, build up some buzz for the brand.