This is going to be a big comparison. My first article ended up being more of a welcome to the new Hi-Fi Cool. In that first edition, I did start the comparison process but my links and comments got buried in the linked blog. As part of this installment, I’ll be moving those comparisons and my comments with even more detail.
I’m trying to capture the smallest details between my changes. I’m doing this between simple power cable changes, ethernet cable changes, etc.
PLEASE NOTE: When listening to the comparison files, we recommend you use a good headphone rig which includes headphones, DAC and headphone amplifier.
Comparison files:
Winner – 35MB ZIP:
https://hificool.com/audio-samples/winner_original_96k-korg-dragon-pc-Lush_usb-JCAT.zip
All files – 406 MB ZIP:
https://hificool.com/audio-samples/0001_audio_samples.zip
Tested components (see comparison files above):
Audioquest Forest USB Cable
Lush ^3 USB Cable
Curious USB Cable
Generic Stock Power Cable
Audioquest Dragon Source Power Cable
JCAT USB Card XE (powered with battery pack)
Motherboard’s USB
Furutech Alpha PS-950-18 Powerflux
In system but stayed the same throughout the tests:
Korg MR-2000S – 1-BIT STUDIO RECORDER (Audioquest Dragon Source Power Cable)
Xeon Based Custom Media Server
(HDPLEX PSU with Dell switched mode PSU w. generic power cable)
Audio GD DI-20HE DDC (Audioquest Hurricane Source Power cable
DCS Scarlatti DAC (Audioquest Dragon Source Power Cable)
Esoteric G-05 Master Clock Generator (Shunyata Sigma V2 NR Power Cable & Cybershaft Clock Cables)
Tellurium Black Diamond hf digital SPDIF (Audio GD DDC to Media Server)
Our take:
Goal: Get as close as we can to the same sound of the original files. Attempt to keep everything as neutral throughout the chain to hear what the artist, producer and engineer intended. For the time being, use these results to select the best testing rig possible for future tests where changes to the system happen further up-stream.
Round 1:
Note:
It’s all about eliminating the weakest links. Even though I’m using an Audio GD DDC to get rid of distortion and re-clock the digital signal, having a clean signal from the media server or the source feeding the DDC is critical as noise can always take a free ride affecting the DAC at the point where it converts the digital signal to analog. Everything matters. The motherboard’s onboard USB will be compared to a JCAT USB Card XE powered by a 5v battery pack. In the future, we also plan on testing a linear power supply with different USB cards. Tests have already been performed with and without an Esoteric G-05 master clock to the Scarlatti DAC and the DDC. The clock’s removal is the equivalent of removing a thin wool blanket, everything gets better. In regards to the digital signal, most reviewers believe timing is everything so using a good clock is critical. Short clock path versus long is a topic we will avoid for the time being.
Forest-Generic
Audioquest Forest USB cable with Korg ADC on generic stock power cable:
One word, muted, loss in resolution. Less expensive cable just can’t compete with higher priced competitors. This is a great cable in a lesser system where differences could not be heard.
Forest-Dragon
Audioquest Forest USB cable with Korg ADC on Audioquest Dragon Source power cable:
There is the resolution. Were the issues with the stock power cable all along. Tone sounds true. Like a real violin. Cymbal hits seem a little too emphasized lacking a little touch of realism compared to the original file. Hear just a little more resolution in the piano key strokes then I hear in the original file. Maybe this is happening because the silver utilized in the Forest is not PSS silver (perfect surface silver).
Lush-Generic
Lush ^3 USB cable with Korg ADC on generic stock power cable:
Now we can hear the uptick in resolution. Can hear Jack Broadbent breathe. Doesn’t sound as holographic as the original file. Guitar sounds a little flat, 2D. Trumpet just a little muted sounding.
Lush-Dragon
Lush ^3 USB Cable USB cable with Korg ADC on Audioquest Dragon Source power cable:
Hear more texture in violin strings then with the Forest USB cable and Dragon power cable. Violin just sounds sweeter sounding like the original file. Horn seems a little more refined and real. Snare drum sounds so natural with a perfect tone. Really picks up the touch of the piano like the original file. This was my favorite combination before I switched from the motherboard’s USB to the JCAT USB XE card.
Curious-Generic
Curious USB cable with Korg ADC on generic stock power cable:
I used to own this cable when I had a Chord Hugo 2. Even more resolution than the Lush but just a little hard, less sense of touch. Trumpet definitely sounds more alive and not muted.
Curious-Dragon
Curious USB cable with Korg ADC on Audioquest Dragon Source power cable:
Horn just has too much focus, not as 3D or as refined as with the Lush ^3 USB Cable with the Dragon source power cable. Maybe just a little too sharp. Piano touch is just a little too over emphasized.
Forest-Dragon-JCAT
Plucks on upright bass just don’t sound as real as with the Lush or Curious USB cables
Lush-Dragon-JCAT (WINNER)
Using JCAT card with 5V battery power back, I hear even more resolution. Everything seems to come from a more black background. It seems to get rid of the hash. I hear more of the performance space and more tonal complexity when piano keys are hit. When a cymbal is hit, it shimmers with vitality, like the real thing. In violin strings, I hear just as much texture as with the Furutech power cable but there is just so much more frequency extension in the highs with the Dragon source power cable.
Curiuos-Dragon-JCAT
Just like before, cymbal hits are just too strident, tizzy. This might be even more emphasized now when using the JCAT card. This just shows you as you move up, things that didn’t matter as much very much matter. Plucks on upright bass not as visceral as with the Lush-Dragon-JCAT combination, more like Forest in this regard. Just more 2D sounding compared to Lush
Lush-Furutech-JCAT
Plucks of upright bass just don’t go as deep as with the Dragon power cable. Also, cymbal strikes don’t quite shimmer like the real thing. Just a little muted
What next?
I chose the above files because they could be found on Tidal. In round 2 below, let’s stream and see how close we get to flat raw files on my media server’s local drive.
Round 2
Note: This is where I get into changing the fiber converter on the clean side and other changes on the dirty side. There will always be naysayers but I want there to be proof you can hear using a quality headphone rig.
Past Baseline for Round 1:
What I tested in round 1 should not have been affected by the changes I’m making in round 2 as the files recorded were what I call a flat raw files or music that I have stored locally on my media server. Since I wasn’t streaming, the sound quality should not have been affected by the upstream components. As I talked about before, noise can take a free ride along the cables but the use of a fiber converter should take care of a lot of that noise but not all. If that is the case, we would not be able to hear the difference between changes on the dirty side of the fiber kit but we can. Proof is not in the pudding, it is in the sound file. I believe timing errors and other issues are creating jitter. As they say in the computer industry, garbage in, garbage out. I can go visit hundreds of websites that explain it better than I ever could but let’s just listen to the recordings and try to make our own judgment. I’m moving quick with quick comments. For now, please make comments in the YouTube comment section.
Since we will now be streaming from Tidal, all the upstream components are going to be utilized. Let’s go through all of those (in order of flow):
Digital Signal flow:
AT&T Cable Modem
TP Link Router (built in switched mode power supply and no wi-fi so maybe a little less distortion)
TP Link Switch (built in switched mode power supply)
Audioquest Cinnamon Ethernet Cable
Fiber Kit
End of Dirty side
Cheap Media Converter (powered by Small Green Computers SMG 5v power supply & stock fuse)
100ft of OD-5mm Industrial TPU OS2 LC to LC Outdoor Armored Fiber Optic Cable, Duplex Single Mode Fiber Patch Cable, 9/125um Uniboot LC Fiber Direct Bury(purchased off Amazon – Ytianhome)
Start of Clean side
SMG OpticalModule (powered by SMG 7v power supply with upgraded Quantum Science Lab Yellow Fuse)
Audioquest Vodka Ethernet
AQVOX SE Switch
Shunyata Venom Ethernet Cable
Media Server (motherboard’s built-in ethernet port)
Power flow:
Dirty side power conditioning on a Audioquest Niagara 1000 Power conditioner (wall – Audioquest Hurricane HC power cable – conditioner)
Clean side power conditioning on a Audience Adept Response aR1p Power Conditioner (wall – Balanced Power Technologies – BPT – L9-C Power Cord – strip)
Furutech e-TP86G 4-Outlet Power Distributor with FPX- G AC Outlets – has 2 outlets (4 banks of 2)
Note: AQVOX SE switch on nice AV style surge/conditioner strip)
File korg-dragon pc-Lush usb SMG-Vodka-AQVOX-Venom show how bad streaming is compared to the original file. Not much texture, edgy, noisy background, etc.
Change 1
First thing we did was get rid of the intermediate AQVOX SE switch from the chain and most likely the noise it was picking up from the cheap AV conditioner/surge strip that had many attached switched mode power supplies including a desktop computer with its typical noisy ATX power supply. These switched mode power supplies are known to be very detrimental to the sound as they spew all kinds of noise downstream. Most of my audiophile friends try to eliminate hooking up switch mode power supplies to their audiophile-grade power conditioners or at least isolate them on to their own power strip/conditioner.
So we started off with just the SMG OpticalModule with the Audioquest Venom ethernet cable going straight into the media server. No extra switch.
Change 2
Moved from the Audioquest Venom ethernet cable between the SMG OpticalModule and the media server to the Audioquest Vodka Ethernet Cable.
Change 3
Move from using the SMG OpticalModule to using a SOTM sNH-10G High-End Audio Switch Hub. Started off using the Shunyata Venom Ethernet Cable between the SOTM switch and the media server.
Change 4
Moved from Audioquest Venom ethernet between the SOTM switch and the media server to the Audioquest Vodka Ethernet Cable.
Change 5 (WINNER)
Went from the cheap switch mode power supply on the SOTM switch to a SOTM SPS500 Signature Edition power supply that has OCC silver internal wiring and EMI absorption inside. Started off using the Shunyata Venom Ethernet Cable between the SOTM switch and the media server.
Change 6
Moved from Audioquest Venom ethernet between the SOTM switch and the media server to the Audioquest Vodka ethernet cable.
Change 7 (RUNNER UP)
Moved from Audioquest Vodka ethernet between the SOTM switch and the media server to the Final Touch Audio Ethernet Cable (FTA).
Verdict:
The difference between having the extra AQVOX SE switch in the circuit is really not even worth talking about. Also, using the cheap switch mode power supply the SOTM switch came with was not a great experience. It just can’t compare to using the SOTM SPS500 power supply. There is a hard edge to everything when using the stock switched mode power supply. So going forward, I’m only going to discuss using the SOTM switch with the SPS 500 power supply.
What did I hear when evaluating the recorded files using Round 1’s winning Lush/Dragon combination:
SOTM switch vs SMG OpticalModule: Jack Broadbent’s voice on the SOTM switch is just so much more intimate, like you can hear the space or atmosphere around the singer. I’m hearing all the complexity of his voice. The hard edge of the SMG OpticalModule seems gone.
With the SOTM switch, Chet Baker’s horn is right in front of me.
Venom ethernet, same as vodka but just more organic, more natural, real. Really get a sense of the soft touch of the piano keys. Cymbals strike’s ring true with decay. Also wins on tone, violin and guitar just seem more natural. More like you in a 3D space around the performer or instrument. What made me switch my winner from FTA to venom was tone. Violins, guitars and really everything sounds more like the real instruments that were recorded, more 3D space making everything sound realistic.
FTA Ethernet: Chet Baker sound sample at the end of the file, cymbal seems to have more texture, where Vodka seems just a little blurred and the Venom seems a little soft. Cymbals just seem more real at the end of the sound sample. Lush USB with the FTA just seems to have a wonderful tone but Venom just sounds more realistic than the others. Can hear more of the atmosphere around the singer like with the Vodka and Venom on the SOTM switch unlike the SMG OpticalModule. Violin just sounds the tiniest bit sweeter and just a little more extension in the flat file. Maybe swapping out the TPLink switch with its built in switch mode power supply with the AQVOX switch with its very custom tailored switch mode PSU on the fiber kit’s dirty side will make all the difference. See round 3. Also need to try swapping the 5v linear PSU better with the better 7v linear psu with it’s upgraded fuse on the dirty side. One person I read about used two SOTM switches, one on the dirty side and the other on the clean. Is that overkill? Only one way to find out.
Round 3:
Now we are going to make changes to components on the dirty side of the fiber kit with one last change to the clean side where we will add the Esoteric G05 Master Clock to the SOTM switch. Please note that the clock is 50 ohm and the SOTM switch is 75 ohm so I need to order an adapter and one extra 50 ohm clock cable from Japan. Have heard that using the wrong ohm cable can introduce jitter but I’m not hearing odd right now. Please note, changes on the dirty side are much harder to detect. I made some huge changes a month ago and the biggest was swapping out the wireless Google mesh network for some hard wired TP Link gear with Small Green Computer’s fiber kit. The second best improvement was adding the Audioquest Niagara 1000 with Hurricane HC between the wall and the Niagara to power everything on the dirty side.
Change 1
Swapping out the cheap fiber converter with the SMG OpticalModule powered by the SMG 7v linear power supply with the Quantum Science Lab Yellow Fuse
Change 2
Swap the TPLink switch with its built-in switch mode power supply for the AQVOX SE switch.
Change 3
Swap the Van den Hul The Mainsstream Hybrid Power Cord powering the SMG 7v linear power supply with a Furutech Alpha PS-950-18 Powerflux power cable
Change 4 (WINNER)
Clock the SOTM switch 75 ohm input with the 50 ohm Esoteric G05 Master Clock. Will get another 50 to 75 ohm adapter soon.
Verdict:
Like I said, we are now getting closer to the performance of the raw flat file. Also many of these changes are getting harder to detect especially on Sennheiser HD800S headphones where everything before this point is sounding wonderful.
Switching now to the Dan Clark E3 planar closed back headphones, I would say that the cymbals sound just a little more real when the super cheap fiber converter was removed on the dirty side and replaced with the SMG OpticalModule and 7v power supply with upgraded fuse.
The biggest change was reclocking the SOTM switch. Was almost like the music got louder. Chet Baker’s horn sounds more sonically pure and pristine, like it was directly connected to the synapses of my brain, just beautiful. As I was writing this review. I let Tidal play its random music from my music picks of the day. I have never heard such visceral drums, sonic beauty out of those Sennheiser HD800S headphones. I’ve never heard Sennheiser’s top, very expensive HE1 headphone system but if you told me that is what I was listening to, I would believe you. What I was hoping kind of happened. When using the FTA ethernet cable, I ended up missing the more organic, natural presentation of Jack Broadbent’s voice, like what I was getting out of the Shunyata Venom ethernet cable. I might have heard more fine details out of the FTA but would trade that for Venom’s beautiful natural tone and sense of 3D beauty. I’m now thinking Shunyata Omega could be my end game ethernet cable between the SoTM switch and my media server. I believe it would give me that increase in articulation I heard in the FTA over the Venom v1 but keep things organic letting me keep a more emotional connection with the music. With the SoTM switch, it’s almost like my brain can relax and ease more into the music. With the clock on the SOTM, it is like I’m inside Jack’s throat. I hear more of everything, like the background got even darker, less hash. When I was analyzing the violin previously with the SOTM switch powered by the linear power supply and FTA ethernet, I noticed that the violin didn’t seem to have the exact same high frequency extension as the flat original file. Well now with the SMG Optical module in place of the super cheap fiber converter on the dirty side, I’m now hearing that same amount of extension. Also, one other huge thing I’m now hearing is more texture in the violin strings with the clock on the SOTM switch just like I can hear in the raw flat file. It almost makes up for the loss in resolution I heard on the Venom compared to the FTA. I really think the sweetness came back when I got rid of the cheap fiber converter on the dirty side and got everything sounding more natural and organic when I clocked the SOTM switch.
What is next:
Well I couldn’t resist and I purchased a JCAT Optimo DUO Power Supply so I’ll be comparing that to the power brick that usually charges cell phones but in my case, was powering my JCAT XE USB card. I also purchased a JCAT XE NET card since the Optimo can power the USB and NET cards. That means another comparison of the motherboard’s network input versus the NET card.
I need to test the Niagara 1200 to see how it compares to the Audience/Furutech combination of conditioning. I also need to try putting the balanced power back on the DCS Scarlatti since I remember it making everything more dynamic.
Maybe put the Audioquest Dragon Source power cable on the power supply that powers the SOTM switch. This could make everything much more dynamic. My friend and I both heard large gains using very good power cables on the clean side’s OpticalModule fiber converter. It would make sense the same would be true on the power supply powering the SoTM switch.
Thoughts and ramble:
I think if we can get all this right, everything downstream will stay sweet and natural sounding. The trick is to make the preamplifier, amplifier and speakers also sound neutral or with just the right amount of coloration you desire but if the sound you are after is not there upstream, how will it ever arrive downstream to your headphones or speakers? If there is no texture or sweetness to begin with and how will it ever exist downstream? Yes, you can add warmth to get that emotional connection at the preamplifier, amplifier or speakers part of your system, but is that best? Isn’t that like creating one problem and trying to correct it somewhere else? Yes, it can be corrected but I believe there will always be some kind of degradation. Loss in natural tone, decrease in dynamics, etc. Garbage in, garbage out. I think it is a bit like what I’ve heard out of silver coated copper cables in the past, some things can sound so right but there is always some kind frequency smearing or you’re trying to correct for some deficiency downstream. Some songs sound great, others seem like the tone is off and you are always chasing your tail. Why not get it right from the start, from the source all the way up. Retailers want us chasing our tails, just get a better amplifier, who cares about room acoustics.
I’m starting to think maybe it is really smart to evaluate things in this manner. With headphones, you can be a lot more critical hearing the smallest differences. You are taking out all the factors that can mess up your critical listening like what can happen many times in a two channel system. Along with upstream components that can color your system, there are many factors that can go wrong when listening to speakers when the room really is 50% of the equation.
Future Test components:
SoTM tX-USBx10G USB Host Card/SCLK-EX Clock Board
Pink Faun USB bridge v2 with OCXO Clock
Pink Fain USB bridge V2 without clock
Pink Faun SPDIF bridge
Ansuz Acoustics PowerSwitch
2nd SoTM sNH-10G High-End Audio Switch Hub on dirty side
Audience FrontRow USB Cable
TPLink Switch, Router, WAF Endpoint
Google Wireless Mesh Network
JCAT OPTIMO 3 DUO Power Supply
JCAT Reference SATA Cable
Audioquest Niagara 1200 Power Conditioner
Terry Audio EI 30A ‘Whole Room’ Cast Balanced Power & Isolation Transformer
JCAT NET NET Card
Startech PEX1000SFP2 – Modded clock input (Sean Jacobs dc4 powered)
SFP Modules: Finisar 1318 SFP, recommend any others?
Farad Super10 Power Supply (to media server’s HDPLEX 400W Hi-Fi DC-ATX Power Supply)
Plixir Elite BDC Linear Power Supply (try on SOTM switch)
Farad Super3 Power Supply (try on SOTM switch)
Digital AES/EBU Interconnects: Shunyata Omega, Jorma, Black Cat 110 Digit, Black Cat TRØN Signature Digital
Clock Cables: Shunyata Omega, Sigma v2
Various LAN and USB Noise Isolators
Ground Noise Conditioning Products (JCAT, Melco, Shunyata and others)
Media Servers (JCAT XACT S1, Innuos, Aurender, Auralic, Grimm, Pink Faun and others)
Your Suggestions?
We thank the following for the use of their sample files:
https://bluecoastmusic.com/
https://www.hdtracks.com/
Components used for analysis depending on the test:
Media Server: Xeon Based Custom (HDPLEX PSU with switched mode PSU)
USB: Motherboard & JCAT USB Card XE (Audio GD DI-20HE DDC to DCS Scarlatti DAC)
Clock: Esoteric G-05 Master Clock Generator (Audio GD DI-20HE DDC & DCS Scarlatti DAC)
A2D/ADC: Korg MR-2000S – 1-BIT STUDIO RECORDER
Headphone/Amplifier: Enleum AMP-23R Integrated, Bakoon SATRI AMP-5514 Pure Class A 15W
Headphones: Dan Clark E3, Sennheiser HD800S & Moon-Audio Silver Dragon Cable
Speakers: Decware Trapezium Speakers v1, ZU Audio Soul Mk II Speakers
Cabling: Audioquest Thunderbird RCA Interconnects and speaker cables, Cybershaft Clock Cables,Tellurium Black Diamond hf digital SPDIF, Audioquest Forest USB Cable, Furutech Alpha PS-950-18 Powerflux, Shunyata Research Sigma V2 NR, Audioquest Hurricane HC, Hurricane Source and Dragon Source power cables, ZU Audio LIBTEC speaker cables