Light Harmonic LightSpeed USB Cable

Device Type: USB Cable
Availability: Authorized Dealers
Price: 0.8m = $999, 1.6m = $1,399, 3.8m = $1,999

Equipment Used in Evaluation: MSB Technology Analog DAC with Analog Power Base, early 2011 MacBook Pro 2.3 GHz Quad Core i7, 16 GB RAM, Samsung 840 Pro SSD, Promise Pegasus Thunderbolt Drive 8TB, GRAID 8TB Thunderbolt Drives. Audirvana Plus/OSX Mtn Lion, JRiver, JPlay/Windows 8 Pro 64. Interconnect cables: Synergistic Research Tesla Apex LE, and Tesla Precision Reference LE. Speaker cables: Synergistic Research Tesla Apex LE. Power cables: Synergistic Research Element C.T.S. Digital, Tesla Hologram D, T3, and Precision AC cables. 2 Synergistic Research Tranquility Bases, USB Active SE cable and Thunderbolt Active SE cables. Other USB cables on hand were the Audioquest Diamond, Synergistic Research Active USB SE, and the Wireworld Platinum Starlight.

The LightSpeed USB cable is Light Harmonic’s latest entry to an ever growing and somewhat crowded filed of audiophile USB cables. Light Harmonic, well known for their outstanding Da Vinci DAC that I previously reviewed last January (see review), has created what they feel is a USB cable that very few other manufacturers can match in terms of bandwidth. Light Harmonic claims that the LightSpeed with its Ultra-high speed differential pair provides 20 times the bandwidth of USB 2.0 high speed. With its 10 GB per second bandwidth, it offers performance that many other high end USB cables cannot come close to. In fact, Light Harmonic has measured many high end USB cables and found that they cannot meet the USB Audio 2.0 High Speed standard of 480mbps. The LightSpeed physically splits the line with separation of the 5V power supply to reduce noise and disturbances of the digital signal.

The LightSpeed is offered in two configurations
1. The Standard split line with a shared connector at each end. 2. A 3 connector version with two USB-A connectors at the computer end; one for the signal, the other for the power line. One USB-B termination of both lines for the DAC.

I evaluated the standard split line version with a shared connector at each end.

Technical Specifications

Conductors Signal: 0.51 mm silver plated 99.9997% pure copper
Power: 0.6 mm 99.9997% pure copper
Connectors USB-A: 30µ 99.9% Gold Plated A Grade Phosphor Bronze
USB-B: 30µ 99.9% Gold Plated A Grade Phosphor Bronze
Dielectric Construction Layer 1: 0.55 mm Gas Injected Polytetrafluoroethylene
Layer 2: 0.08 mm Biaxially-oriented Polyethylene Terephthalate
Layer 3: 0.32 mm Gas Injected Tetrafluoroethylene
Layer 4: 0.08 mm Biaxially-oriented Polyethylene Terephthalate
Layer 5: 0.35 mm (+/- 0.1) Polyvinyl Chloride
Final dimensions Square-shaped, 3.5 mm (+/- 0.1) X 5.0 mm (+/- 0.1)
Rated Bandwidth 10.083 Gbit/s

DSD Playback Issue
Just how important these specification of speed and signal integrity with reduction of the 5 volt line noise was made plainly evident to me. I have been having some issues playing native DSD files with Windows 8 64bit Pro / JRiver Media Center 18 / JPlay 5.1. Intermittent dropouts have been occurring with intermittent clicks with some of my otherwise excellent sounding USB cables. I tried to increase the buffer size for JPlay 5.1, but had no success eliminating these playback issues. I decided to try the LightSpeed to see if its superior bandwidth and claims of noise reduction would solve this issue. Inserting the LightSpeed between my MacBook Pro and MSB Technology Analog DAC completely eliminated these playback defects with absolutely not one dropout or click.

I asked Gordon Rankin of Wavelength Audio if he had some insight into the problems I was having with some of my cables. Gordon is one of the most knowledgeable engineer-designers on matters of USB that I know.

Gordon replied:

“The problem is what we call turnaround and has to do with the overhang of noise from when the computer sends what's called an IN packet request to the DAC. At that point the USB cable turns around and the DAC sends back the feedback pipe which is getting lost in this case because the overhang of the computers IN packet is lingering (probably too much capacitance or impedance is wrong) and corrupts the feedback sent to the computer. Since the feedback packet is corrupted the computer ignores this and continues to send out data as normal. This causes either and over run or under run of data to the dac chip which causes pops and clicks. Windows does not seem to handle this as well as OSX for some reason.”
It would appear that the selection of the USB connectors also influence the performance of the LightSpeed.

A Soundstage to Die For
I recently replaced my Levinson 32 preamp with an Ayre KX-R preamp. The Ayre has balanced outputs which did not allow me to use my run of single ended cables to my Wilson Watch Dog 2 powered subwoofer. The XLR to RCA adapters I had were incorrect for this application. Now the Wilson sub helps to create a deep and wide soundstage with my Wilson Sasha speakers. Even though the Watch Dog comes in at 35Hz, it has a substantial effect on the perceived soundstage of my system.

Using the LightSpeed provided an enormous soundstage that just about matched what I was hearing with the subwoofer in the system. The depth and width were excellent with lateral reproduction of instruments that went well beyond the outer border of the speakers.

Playing the Weavers Reunion at Carnegie Hall 24/96 Classic records DAD, I was simply amazed at the soundstage given the fact that the sub was out of the system. The depth was some of the best that I have ever heard with this recording. But this leads me into the next characteristic of this cable.

A Low Noise Background
Listening to the same Weavers at Carnegie Hall title I was also struck by the low background noise of the cable. I heard subtle hall reverberation with this cable that was masked with some other fine cables in my possession.

The excellent Reference Recordings Doug MacLeod There’s A Time 24/176.4 allowed the guitar and voice of Doug MacLeod to emerge from a deep black background that highlighted the definition of his guitar playing. The pace and rhythm reproduced by the LightSpeed was impressive. This is really a wonderful recording. The only USB cable that I’ve heard that surpasses this background silence is the Synergistic Research Active USB SE with its Active Shield powered by the Transporter Ultra SE. But the LightSpeed would be considered superior in this quality compared to most other high end USB cables.

Dynamic Reproduction with Weight and Wallop
I don’t think I’ve ever used the term wallop to describe the punchy dynamic quality of a USB cable. Aaron Neville’s My True Story 24/192 had impact and a sense of rhythm that was possibly the best I have heard with this recording. The LightSpeed is fast with no overhang. Its dynamic qualities are first rate with the ability to draw one into the musical presentation. The rhythmic drive that this cable can deliver is something that must be experienced to fully appreciate.

Hans Zimmer and James Newton Howard’s The Dark Knight Soundtrack 16/44.1 delivered fast powerful dynamic bass lines in track 1 "Why So Serious?" Again, even without the subwoofer in the system, the bass moved the walls and rattled the windows in my room.

Detail and Definition
The LightSpeed reproduces instruments and voices with exceptional clarity and definition. The LightSpeed is not bright; it does not soften the high end nor is it euphonic sounding as some USB cables I have heard. Complex symphonic musical passages are beautifully unraveled with this cable. I listened to the Mercury Living Presence native DSD Howard Hanson / Eastman-Rochester Orchestra playing the Song of Democracy. The chorus was well separated from the up-front orchestra. The voices were easily understood. The orchestra was reproduced with excellent dynamics and sounded very real. What a superb recording from the golden age. The LightSpeed‘s exceptional clarity and resolution allowed this recording to easily surpass my early vinyl pressing; especially in the loud passages.

What About DXD Playback?
DXD files or Digital eXtreme Definition is an audio format recorded at 352.8kHz. 2L Records utilizes 32 bit floating point at 352.8kHz for their masters. This leaves headroom for their editing and balancing before converting to DSD. sells the 24/352.8 files for 2L recordings. I was interested to see how the LightSpeed would handle DXD files; certainly a more difficult task to deliver to a DAC than DSD64 files.

The Hoff Ensemble’s Quiet Winter Night was downloaded at the master sample rate of 352.8kHz/24. I have enjoyed the 24/192 version purchased from HDtracks and was interested to see if I could hear a difference between the two formats. More importantly, I want to see if the LightSpeed could handle the DXD format.

Not all DACs are comfortable playing back 192kHz files, let alone, 352.8kHz files. The MSB Technolgy Analog DAC with its modified Signature Ladder DAC modules and precision femto clock can handle these files easily.

This title was the first true DXD recording I have ever listened to. While the 24/192 version was very good, the 24/352.8 easily bettered it. The ambience seemed to dry up in the 24/192 version with a collapse of the soundstage. Now the soundstage is very good in the 192 version, but nowhere as good as the DXD version. The percussion that plays in the background-left became less defined and a little muted with the 24/192 version. The decay of the percussion was not as prominent with the 24/192 file.

I’m not suggesting that music needs to be recorded at 24/352.8, but when carefully executed as 2L Records seems to do, the result can be stunning. And the point to take home with this discussion is that the LightSpeed was able to discern the differences in these files with no dropouts or noises.

The LightSpeed Has Weight and Body
I played Mel Torme Swings Shubert Alley 24/192 (HDtracks) and was quite surprised with the result using the LightSpeed. Not only was the orchestra’s instruments extremely clear, but Mel’s voice had body and weight and sounded less 2-dimensional than with some other USB cables. I suspect that the cable’s speed and low noise just allows more information to get to the DAC . Also, timing errors are probably reduced with the LightSpeed.

A High Resolution USB Cable
The Light Harmonic LightSpeed is capable of exemplary resolution when paired with a fine DAC. Its special qualities of speed, detail, soundstage and ambience reproduction, along with its low noise capabilities, place it among the finest USB cables I have yet heard. Having had an opportunity to live with the Light Harmonic Da Vinci DAC, I can see how this USB cable would beautifully complement the low grain-analog sound quality of the Light Harmonic Da Vinci DAC.

Associated Equipment

bsm's picture

Also, it'd be great if Light Harmonic could actually provide their test data showing other cables not reaching the 2.0 standard, as well as data showing dropped bits in others vs. no dropped bits in theirs.

Archimago's picture

Yeah, wish these companies would back up their claims.

Stereo DSD128 and 24/352 isn't exactly a lot of bytes/s for a decent USB2 cable to transfer (compared to say a hard drive)...  I have not run into any pops/crackles/etc. with JRiver and the TEAC DAC at these speeds with generic USB cables.

I would submit that some of the issue may be JPlay and how it influences the CPU load. For some reason audiophiles like to also limit the buffer size to bare minimum on JPlay which of course exacerbates the problems.

Steven Plaskin's picture

I found issues with some cables using just JRiver (DoP) and OSX with Audirvana Plus.

JPlay didn't seem to make a difference. I also thought that the buffer might be an issue with JPlay, but increasing the buffer didn't solve the problem.

24bitbob's picture

Another fascinating review, thanks, but wow, $1,000 for a USB cable. 

How about this from the antipodes:  cables that also separate the power from the signal.  Not inexpensive either but closer to the 'do-able' range for more people


Wavelength's picture


If a printer misses part of a signal, it can and will ask the computer to send the signal over and over again, until it gets it right. That's not an option in streaming an audio signal to a DAC; if the DAC misses part of the signal due to noise or jitter, it's forced to "guess" at the missing bits. When that happens, you no longer have bit-perfect digital audio; besides which, if the timing of an arriving music signal package is altered, there is no way to correct it again.

While it is true that isosynchronous data sent from the computer to the DAC does not have error correction. I have set up a pretty elaborate test setup for cables and in my experience there has been obvious data errors because of this but not consistent errors. Also the DAC when it does receive a packet error does nothing to the data it receives and in most cases is only a BIT off and playback is not totally compromised.

From above... let me explain further what Asynchronous USB Class protocol is all about. This will help you understand what causes pops and clicks.

Basic USB... when the computer sends data out it's port it sends an OUT request which tells who it's going to and then it sends the data. When the computer wants a device to send data it sends an IN request with the address of the device and at that time the computer stops sending data opens the line for the device to send data to the computer. This was originally conceived as a poll and select protocol.

Now with Asynchronous USB enumerates (tells the computer what it is) two different endpoints. Each device can have 7 in and out endpoints, with endpoint 0 in/out being the control point for all devices. The two ASYNC endpoints are the input or data stream from the computer and the feedback stream which tells the computer to slow down or speed up the data stream so the dac chip does not run out of data. The 3 byte feedback pipe to the computer happens about every 32 packets of data.

So the big reason for pops and clicks is not the receiving of bad data. It's basically that the computer is not doing flow control correctly. Think about it this way... if you initiate cruise control on your car and you hit a hill and the car slows down then really there is no control. The same way with ASYNCH protocol if you need more data or less data then the device controller (at the dac chip end) has nothing to do but reset it's buffering system. This is what causes pops and clicks, not bad data.



Steven Plaskin's picture

Thanks for the excellent explanation Gordon.

Priaptor's picture


I am not sure if you received a "broken in" cable or not, but my first impressions were that "this cable sucks".  I have been fooled enough to give it some burn in time.  Lo and behold, by day 3 of constant usage, no doubt I realized my initial impressions were way too hasty.  I can say from a  resolution standpoint, this cable resolves like none I have ever used.  

Unfortunately (or fortunately) I had to leave for a vacation so I only had day 4 after burn in for my initial assessment but really like the direction the cable was going, however, I still haven't realized the soundstage of my current standard.  Just an early assessment and I fully expect the cable to continue to improve with time in all aspects.

While I have experienced "burn in" differences with other cables, as I am sure most of us have, I have never experienced such a difference with a cable as the Light Harmonics.  If the soundstaging improves as I fully expect it to and for sure hope it will, I can see this becoming my new reference-BUT, it is not there just for me.   

Steven Plaskin's picture

Yes, the cable needs a good deal of break-in. At first I thought the soundstage was a bit small and the bass lacking. It does get better with play; much better.

For DACs like your MSB Tech DAC IV Diamond plus, it will sound very good. It loves a truly neutral DAC.

Priaptor's picture

BUT this was quite a surprise as to how much character change I noticed after 3 days. 

I am looking forward to getting some more hours on it.  I really like what is was starting to do.

Jason Victor Serinus's picture

I initially tried this cable cold, and was extremely disappointed. After I read this exhaustive and extremely positive review, I contacted Bill Leebens, who is repping Light Harmonic. He informs me that the cable needs two weeks of solid break-in to fully come into its own. Alas, mine is in a box in a pod somewhere, staying warm until we sell our house and begin to settle into another. Given Steven's endorsement, I'm kicking myself for not leaving it out. I could have set my laptop on repeat and broken it in by now...

Larry Ho's picture

Jason, Glad to see your message here...

Proper break-in procedure in my mind, I know you want to make the period shorter if possible. Right? 

1. For ultra high speed differential pair 

Maybe the quickest and easiest way to break in Light Speed is to use it between USB 2.0 external Hard Drive (better larger volume and fast one)... Continously copy big video files back and forth. Random 0 and 1 and constantly high speed at 480M Hz shall break-in this part faster.  

I saw one article somewhere, one programmer even wrote a simple program for this. Copy files in both directions. Up stream and down stream... ;-)

2. Power leg part of LightSpeed

Use Da Vinci to break-in the power legs part of Light Speed is not agood idea, because it's SELF-Powered. Better use some DAC or some devices that will draw near 300mA ~ 400mA constantly...

I think by using above method, LightSpeed USB will break-in faster.

Sorry that like Da Vinci, I choose the best material and construciton method for LightSpeed and God is fair. It usually needs more time....


bsm's picture

I'm curious as to how "burn in" changes the bits.

Priaptor's picture

I was as baffled as you.  I have never had a cable with such drastic changes in the burn in process.  I was ready to bring it back from where it came after the first night.   It literally "sucked" so bad, I emailed several people with my initial impressions figuring there was no way for the cable to deliver based on my initial impressions, what the claims by the company was.

Well after 3 days, the character totally changed for the better.  I am looking forward to more sessions when I get back as this cable has me baffled.  I have never seen such dramatic changes in character in any cable I have ever used. 

Steven Plaskin's picture

Hi bsm.


While I'm hardly an expert on cable design, I have found that many USB cables I have listened to improve with use.

The 5 volt line insulation, the insulation for the signal, timing changes with insulation forming might influence the sound even if the bits are not changed.

This is an interesting link from Cardas cables:


This AudioStream discussion might also add some insight to your question:

billleebens's picture

We are investigating ways of breaking-in or burning-in the cables before shipment. 
No conclusions at this point.

Right now we have our hands full with the Kickstarter campaign for our next product, "GEEK", which is a dongle-style DAC/headphone amp that handles up to 384k in PCM, and is DSD-capable.

Lana Sander's picture

Since I can't measure the cable, I don't know. But Light Harmonic does make some really excellent sounding gear. The Da Vinci DAC was one of the best DACs I have ever lived with. 

reverendo's picture

is there any subjective comparison to the three other USB cables you mentioned in the beginning of your review. I know all three of them well and the AQ Diamond fit my system best. Am burning in a Purist Ultimate at the moment, but might be able to compare the Diamond, the Ultimate and the Lightspeed to each other.

Best regards


highstream's picture

While the Lightspeed is undoubtedly a fine cable, I agree with reverendo and find this review less than forthcoming. The author has a plethora of top quality cables on hand and doesn't mention any specific comparisons?? Was that because there wasn't enough time, or...? Each cable has its own presentation style and characteristics, just as each system is different and has its own characteristics and needs. Without some comparison, there's no way to place the Lightspeed cable sonically. "...speed, detail, soundstage and ambience reproduction, along with its low noise capabilities" doesn't really tell the reader very much.

wfojas's picture

All the cables used in the USB comparison (the Audioquest Diamond, Synergistic Research Active SE cable and the Wireworld platinum) as well as the Synergistic Research Galileo Cables seem to compare so well with this, how is this not a Greatest Bit? Just curious as to what keeps it out.