An Objective Test of USB Cables for Audio - Supra, Oyaide, Wireworld
HTML-код
- Опубликовано: 8 июл 2024
- An objective test and review of three affordable USB cables for audio: the Supra, Oyaide Neo d+ and Wireworld Chroma 8 cables. The biggest question here of course is: "Do USB cables make a difference?" Take a look at the tests and decide for yourself and/or share your thoughts in the comments. Most of these cables aren't available on Amazon it seems, but here's a link to the Oyaide Neo d+ (amzn.to/380OPB0) and AudioQuest Diamond (amzn.to/2Pkusbn)
My Curious Evolved USB cable review: • Curious Evolved USB Ca...
Don't forget about the giveaways of the Shanling UA1 and UP4. Checkout the Jotunheim 2 ( • Schiit Jotunheim 2 Hea... ) and HE400i ( • HiFiMan HE400i 2020 Re... ) videos to get in the draw. General terms and conditions are here: / 39724405
Thanks again to Melbourne ChiFi Audio for making the giveaway possible! melbournechifiaudio.com/
Stay tuned for the announcement of the winners!
00:00 Preliminary comments
02:25 Introduction
05:56 Test explanation
10:06 Audio comparisons
11:18 Demonstration of inversion approach
12:51 Demonstration of alignment of recordings
13:38 Comparison of generic cable with aftermarket (inverted)
16:57 Comparing two recordings of the same cable (Supra)
18:07 Conclusions from tests
22:46 Subjective conclusions
Like the music in this video? You can get it here: www.epidemicsound.com/referra...
SUPPORT THE CHANNEL
Share this video on your favourite forums or social media: • EarMen TR-Amp DAC/Amp ...
Merchandise is now available here: passion-for-sound.myspreadsho...
Join my Patreon: / passionforsound
One-off donations: paypal.me/lachfen
Use this link to buy ANYTHING from Amazon: amzn.to/3dtyiWQ
Or if you’re in Australia: amzn.to/2U7lqOp
Or from Apos Audio: tinyurl.com/Apos-P4S
Or from Drop: massdrop.7eer.net/d07P2
PASSION FOR SOUND GEAR LIST
(Affiliate links to Amazon and Apos Audio help to support the channel - thank you!)
Desktop Sources Playback software: Roon roonlabs.com/r/St5em2YsbUqGXc...
Primary DAC: Chord Hugo TT2 chordelectronics.co.uk/produc...
with Chord M-Scaler chordelectronics.co.uk/produc...
Secondary DAC: Schiit Bifrost 2 www.schiit.com/products/bifro...
USB cables: Curious Evolved curiouscables.com/
and AudioQuest Coffee amzn.to/3059qiA
Active Speakers: Bang & Olufsen BeoLab 3 Speakers
Amps
Bottlehead Mainline DIY kit tube amplifier bottlehead.com/product/mainli...
Burson Soloist 3XP www.bursonaudio.com/products/...
Schiit Asgard 3: www.schiit.com/products/asgard
Headphones (that I regularly use)
Meze Audio Empyrean tinyurl.com/y6gfyquz
Focal Utopia: amzn.to/2YmO7Je
HiFiMan Sundara tinyurl.com/AposSundara
HiFiMan Deva: tinyurl.com/HFMDeva
Ollo S4X: www.groovecentral.com.au/stor...
Beyerdynamic T5 (3rd gen): amzn.to/3priR7N
Portable
Phone: Samsung Galaxy S21 Ultra amzn.to/3sY1EoJ
DAC for phone: Audioquest Dragonfly Cobalt amzn.to/36AzpAB
Digital audio player: Shanling Q1 tinyurl.com/ShanQ1
Earphones
64Audio U12t www.64audio.com/products/u12t
Audiofly AF1120 Mk2: amzn.to/2ZoTuHZ
Shozy Form 1.4 amzn.to/3cIL0Az
Tin HiFi T2 Plus amzn.to/3fqCbxu OR tinyurl.com/TinT2P
Bluetooth Headphones / Earphones
Nuraphone amzn.to/2N7G2Tu
Nuraloop amzn.to/2HKVZzi
Reviews filmed using:
Sony A7R III: amzn.to/3mnR0nS
Sony FE 24mm F/1.4: amzn.to/3kuHVIf
Blue Encore 300 mic: amzn.to/3oreKae
Scarlett 2i2 Audio Interface: amzn.to/36hRFkn
Neewer Bi-color 660 LED lights: amzn.to/2E1BNrh 
Наука
Great video!
I've actually been working on a very similar video myself, and wanted to make a couple constructive comments on the methodology.
The main thing is that audacity due to how it handles alignment isn't a particularly ideal software for this. I initially tried using it, but even in control tests (where playback was done multiple times with all hardware the same, to check if variances occurred) there was significant variation. If using high sample rate this is less of an issue, but with 44.1khz content its quite important. Looking at some information on fourier transform you can see how visualisation of the samples alone without proper interpolation can often greatly mis-represent the actual waveform recorded.
I switched to DeltaWave, which is a tool designed specifically for digital audio comparisons. It has a few advantages including proper sub-sample alignment down to 1/1000th of a sample. As well as correction for long term clock drift and DC offset and gain offset (which is especially important when using an ADC with an analog pot like the focusrite). This allowed me to get MUCH more consistent results as well as in-depth views of spectrogram and null depth.
Clock drift is also a good reason to use a shorter piece of music instead of a longer one. With transients in particular, even the tiniest amount of clock drift or jitter in either the source or receiving ADC can show a difference in the compared/nulled result when actually it would not be there if long term clock drift was corrected. (In my testing for an upcoming video demonstrating DAC differences, even very high end dacs with low jitter, and additionally chord dacs specifically, can exhibit more than enough clock drift to throw off a null test after only 5-10sec)
Additionally, it would be very important to include a control test. Because if using your testing methodology, the same usb cable recorded twice, had variations as well, then that gives reason to believe that the differences shown might be due to other factors or run to run variation, rather than the USB cable.
My thoughts exactly. There should be either a time alignment issue or an amplitude issue with these results. Clocking of a Scarlett is also not what I would call reference quality. Preamp can also cause significant difference and the Scarlett has no pure line level AD.
Even if we assume with an open mind difference in the USB cables, if the differences were that big it would have been abandoned by the professional audio a long time ago.
@@andrasnelhiebel6726 Yeah. Its certainly great that some objective testing has been done, but unfortunately I wouldn't necessarily say these are controlled enough to call conclusive.
I'm using the ADI-2 Pro FS R ADC for my tests and have been able to get controlled nulls to over 100dB once clock drift is corrected for. So plenty of performance there.
Its going to be interesting to see the results
Looking forward to seeing what you find, GoldenSound. I was aware coming into this that the setup want ideal, but given the lack of anyone else testing USB cables with actual music (not test tones) I figured I should give it a shot.
I'm keen to see your version of this and hear your conclusions. Thanks for commenting!
CONGRATS! You are so ahead of Passion For Sound in your approach, we can't even compare them. You are far too nice. This test is frankly so poorly structured as to be useless. You have identified the large, and the small issues. Well done. Two biggest issues on a short clip ran DAC-ADC are going to be sub-sample alignment, as you have noted, and thermal drift of the DAC/ADC, which would easily account for the illustrated difference. Of course the same cable cancels more .. no time differential for thermal drift. I see you noticed that issue with Chord DACs. Benchmark is more stable. The other suggestion I have is that you sample on the ADC at a higher rate than your DAC, and the DAC better not be NOS, or some other DAC with intentional (or unintentional) high frequency noise that can alias into the ADC, but may be intentionally audible or may be unaudible but picked up and aliased by the ADC.
Exactly the questions I had after watching the video. Interesting test, but we needed that control test. Actually multiple control tests I think. Source variation and DAC variation could be of more importance from day to day or time to time than that of the cable itself ...
Dude, this was a brilliant test
Glad you liked it, Bob. Some valid questions have been raised in the comments, but hopefully it's gotten people talking, thinking and asking questions.
Thanks for investigating this. I think the best way to assess USB cables is with a digital audio interface. Preferably one that has digital out and in. Then you can run a loop back test with different cables.
Years ago there was freeware called Diffmaker. It's premise was to compare signals and subtract the difference.
Ideally you want a file that has a known waveform rather than use visual wave alignment in Audacity
Excellent study here and thank you for simply showing the data along with the sound. So there's enough here for the measure and the listen crowds. This does make me adjust my previous thinking somewhat. A lot of questions remain, specifically since we still don't know which one was the closest in fidelity to the original file, but the most revealing test was that even between the same cable, just a day later there was slight variance.
I think there are problems with my equipment that might have lead to some issues, but I'm glad it's adjusted your thinking and I hope it leads to people questioning and exploring, regardless of what the final outcome might be. 🙂
I love this! While watching I kept wondering what would happen if you recorded through the same cable twice and did the same process, and then you did! As always, your approach, explanations, descriptions of limitations, and calls to keep things civil are really appreciated. I love your channel.
Thanks so much, Jeff! There are significant limitations in my tests, but at least the conversation continues until we can find definitive evidence one way or the other (using music, not sine waves)
What a heroic effort to get some real data about a controversial subject! Thank you!
As others have pointed out, unfortunately, the DAC/Analog/ADC sections are probably confusing the test somewhat. (No, this wouldn't explain why different cables had different results; that's true.)
I have a unique set of gear that might help... I'm using a MiniDSP 2x4 HD, hooked up with a USB interface to a Raspberry Pi 4. To the Pi, the device looks like 1 stereo output and 4 microphone inputs; so I have digital access to the stream both BEFORE and AFTER it passes through the DSP. I can null out the DSP functions to just pass-through the signal, and record the stream that passes through. This would have traversed the USB cable "twice" as it were; out and back in.
Additionally, to keep the noisy Pi from messing with the reference levels in the DAC of the DSP, I have an optical USB isolator that completely electrically disconnects the Pi from the DSP.
I don't have a collection of "audio" USB cables, but I will run a test as I described and see what I get and comment back here, maybe with hosted files if anything interesting pops up.
Again, thanks for this project.
Hi Matt, thanks for your comment. I'll be very interested to see what your system shows and might be able to arrange to send you some cables to test if your first run through shows promise.
You can email me at passionforsound@lachlanfennen.com if you want to discuss further.
Unless you are using software that well correlates audio levels, you must tightly control temperature as thermal drift will easily make for significant level changes. Also need to correct for subsample timing differences. Things like audiodiff maker do this.
Thanks very much for all your work, Lachlan. I really appreciate your channel. I have mid-range gear, Burson Conductor V2 amp, chained with Modi Multibit (USB), Burson Composer (rPi, coax, Roon).
Upgrading my USB and interconnect cables (Forest USB> Modi>Forest RCA> Conductor, and Rpi>Cinnamon coax>Composer, Golden Gate RCA interconnects from Composer>Conductor) has made a stark difference. Not able to say why, but it has. It’s not subtle in any way, it’s a huge difference.
That's great to hear, Heath. I'm in the same boat as you with no explanation of why, but plenty of experience of changes (for better and worse). Happy listening!
I enjoyed this alternative review of RSB cables.
I so happen to have the Supra 2.0 USB cable and have been pleased with its SQ.
i have just wrapped the cable in strips of aluminum foil hoping to reduce any RFI from my router , which is in very close proximity to the DAC and USB cable. This small tweak seems to have given more clarity .
Glad you liked it. Interesting idea with the foil. Glad you're liking the sound
One thing to try is this...If the USB cables are affected by the RF noise or DC/AC, always test the cables with a laptop that's not plugged into a wall outlet. That way, no hum or power issues should be introduced because it's running from battery power. You do have to ensure USB outlets don't fall asleep and are at maximum performance in Windows. It would be interesting to know if power is the cause of this or not.
Wow - thanks for taking the time to run this experiment! This was really fun to watch =)
Glad you liked it! There were some significant issues with the test setup that I've since discovered so take the results with a large grain of salt. I stand by the subjective comments and will do further testing in the future, but the objective part of this video is a little flawed unfortunately
@@PassionforSound That's fair enough! And I suppose that's just part of the pursuit of science - always good to refine the experiment =) I'm looking forward to further tests!
You have made a good effort here. Alas, you have missed the most fundamental problem in null testing: the output and hence the recording can drift causing differences that show up in null test even if nothing else is changed! DACs have local clocks and jitter which can varying the timing/phase down to sub-sample level. You will NOT be able to make such alignment visually.
In addition, the analog output of DACs is not very stable. They can drift up and down during the play or from time to time due to many factors (temperature, stability, etc.). Ditto for ADC capture. This must also be compensated for. Otherwise you will hear the level difference as sounding like the source.
Net, net, digital input was consistent. What was not consistent in your methodology was analog output of a DAC. By definition analog is subject to noise, level and timing shifts. You must remove these extraneous factors that dominate such tests. Without it, you are just seeing/hearing effects of errors in analog waveforms, not anything to do with USB.
The above is why when you captured the same cables two days in a row, you got two different outputs. The input had not changed. The output naturally drifted as I explained. You should repeat this test over and over again and with it, you will see that there are variations as great as what you thought was attributed to USB cables. Since no cable is changed, it invalidates your cause and effect conclusion.
Fortunately some tools have been developed to deal with these issues. The best one is deltawave: www.audiosciencereview.com/forum/index.php?threads/beta-test-deltawave-null-comparison-software.6633/. As you see in the thread there, it compensates for the issues above. Even then though, the nulling won't be infinite. There will always be some error even in identical sequential captures. But if you know what you are doing, you can extract useful info from it.
If you want to learn more, there is a paper from Audio Engineering Society on this very topic: www.aes.org/e-lib/browse.cfm?elib=14680 The tool that was developed as a result of that paper doesn't work very well so I suggest using deltawave.
Bottom line, you have not discovered changes in USB input to a DAC. You have simply discovered that null testing requires specialized tool and understanding of the methodology or it produces garbage results.
Yes, what is being measured is not the usb input to the dac but output of the dac+adc. strictly speaking, unless we seperate out which component is causing this difference, we cannot conclude the effect is caused by the cable ! The conclusions in the video have simply jumped the gun. At best we can say further analysis needs to be done.
An unending source of negativity. Your response was the first thing that popped into my mind when I saw the topic. And, right on que....
@@ericharrelson32 yours as well.....
@@AudioScienceReview without doubt Sir, negativity begets negativity. Seems as though the pot is perpetually on slow boil. Reading through the comments more often than not ignites my ire. Sometimes it's hard to resist putting in my worthless two cents.
Guys, please let's keep things respectful here.
Amir, I appreciate your input and respect what you've said. I was very aware of the imperfect nature of my setup, but not to the technical levels you've described.
The challenge I see is that many of us hear differences between cables. These differences do not always align with our expectations thus ruling out confirmation bias. My next step is to try and arrange a true blind listening test with local audiophiles with controls built-in to account for critical listening skills, chance and the like (I haven't planned it yet and have lots of thinking to do).
Is there any way that you are aware of to accurately test USB cables and their performance with music waveforms rather than just sine waves to test for jitter and noise. What if the rapid fluctuations of signal levels in music are leading to issues with capacitance, inductance, noise levels, etc. all combining to affect the accuracy of the signal being processed by the DAC? I'm not qualified to get into depth on it, but would love to hear from people who are with solid evidence based on measurements using actual music (i.e. not theory alone)
Recently, I participated in an Oyaide giveaway and received a cable from them (Neo USB C to C - Green). My colleagues were a little surprised when they heard the price and asked if it was worth buying one. I had no reason to tell them to buy because, sonically, I found no differences. So, I thank you very much for having done this test that proves that not all cables are used for everything! You were detailed in your convictions and always said that it was your opinion, not the rule.
Idea for video: With the increase in the use of digital platforms for data communication between the various audio equipment, I would like to suggest to do the same test, but with network cables (RJ45), since the use of the same with the Dante protocol it is increasingly present in the audio market.
Best regards
Hi Leandro, thanks for sharing your experience and congratulations on winning the cable! Nice idea for the RJ45 cable. I need a break from these monster reviews for a while, but might come back around when I have some more time 🙂
Really enjoyed the video and thank you for making it. Learned a great deal and need to rethink my cable choices.
No problems. There were actually a few flaws in my approach so don't take the objective results too strongly into account. I'd recommend watching my recent Supra Excalibur review for an update on the progress of my understanding of USB cables so far. 🙂
Really great testing! Very well made. My view is that differences come down to impedance missmatching along the cable including connectors and its RF performance (assuming same length of cables). RF i terms of its efficiency as an antenna but even more how much noise is transferred from the source via its screen and signals (from the laptop in this case) to the receiver (the Chord DAC). All this noise will add jitter at the receiver end (the dac). Depending on the DAC design jitter affect clocks and/or the digital to analog conversion itself will more or less result in audible distortion of the audio signal. And to me, I think this is what we hear in your tests. If timing is not exactly correct for every sample in the conversion to an analog signal we will get a distorted output signal. So this also means that different DAC designs and architectures will be different sensitive to different USB cables.
By the way. I have also chosen the supra USB cable for my system. Good value, sound and quallity.
Thanks for your comments, Anders. I'm still not certain that my equipment setup didn't introduce some variations as well. There is evidence that jitter levels vary little between USB cables (that meet specifications), but something (maybe RF noise as you suggest) is leading to sonic differences based on my subjective experiences. Hopefully this conversation will lead to others doing further testing with music and not just sine waves.
I this is is most likely happening. I am quite sure that data packet transfer can ve only tre or false. USB communication protocol or any other data communication protocol contains method to transfer packets 100% correctly. If nit, receveir ask to sent packet again intil it is OK. I believe that signal coming out from DAC can be somehow affected by type of cables connected to it. But not data packets.
Hi Pavol, digital audio protocols don't use the checksum processes due to the importance of timing on audio
Thanks for all the effort you put into this video. Just a suggestion for testing your hypothesis: you could measure every cable against itself with the exact same method.
Hi David, that would help prove if the system is introducing issues as some have suggested (and are likely correct). I think ultimately someone with better equipment needs to pick this up from here. I'll move onto a controlled, blind listening test with volunteers once I can setup the required equipment and location.
For the benefit of others I am using an 8 metre length of this Supra USB cable and have no technical issues at this length and it sounds really awesome. So if you need to run a long length then fear not.
Amazing comparison! Great job
Thanks Evgeni! A few viewers have identified some flaws in the approach I used so I'm designing a new test soon. Stay tuned 😉
This is brilliant. Utterly brilliant. Honestly I don't think anyone's come up with a test this ingenious to prove digital isn't flawless. I can't get over how brilliant this is.
Haha. Thanks. There are signs that it might be flawed due to the non professional equipment I'm using (i.e. it's not specific test equipment), but hopefully others with that gear can take it the next step
I appreciate the effort you put into this. My only comment would be to question the use of the Scarlett, which is far from an exacting piece of gear. I've used the Scarlett and found it colored and lacking in transparency. I've also used the Motu M4, which does sound better, but also has its own issues with its draconian filtering above 22KHz.
It's a fair point, John and one raised by others. Hopefully this has at least inspired some people with professional test equipment to see what they can find and therefore continue the conversation.
Possible mains voltage swings affected clocking/timing/amplitude or whatever in the Supra/Supra2 test? Do you have a kill-a-watt or something equivalent to show your current mains voltage?
I decided to try the Supra USB cable. After the proper burn in period, I had it in my system for 2 weeks to evaluate it. This USB cable has a relaxed and spacious sound, but with a more prominent bass. That may work in some bright systems, however, it also has less detail and a less controlled, more prominent bass. This overwhelming bass was distracting, I just could not get used to it. I prefer a lean, controlled bass and more detail in my system, that's what the wireworld Chroma is all about. I'm listening to the Chroma again, and the pleasure of Music listening is also back. Yes, it's not perfect, but in my system it works like a charm and giving me more listening pleasure. I think USB cables are system dependent. Try them out and see if it complements your music system.
I think system synergy and personal tastes are a big factor in cable choices. Glad you're enjoying the Chroma. Thanks for sharing your experiences with the Supra
Between the Diamond and the Oyaide inverted it was very subtle difference in sound it didn't allow to much sound to get through the supra and the wireworld allowed too many transients to get through like it was very noticeable they couldn't keep a good fight between the Diamond nice review I just order the Oyaide neo d+ thanks to this video.
Hope you like your new cable!
Super interesting content and worth looking into and discussing. Thanks!
Thanks Casey. It's an imperfect test, bit hopefully it sparks further conversation and exploration.
Quality review. I love the way you did. Thanks
Glad you liked it. There were a few flaws in the methodology for the test, but I'm pretty confident that I was hearing a difference 🙂
If there was a way to see the comparisons between any two signals the difference of the waves and the audio outputs, I would feel that that information is very viable. We don't get the see each comparisons outputs, only the raw L/R output. Also, if we could see what signal is left of each would be quite interesting. Thank you for the awesome video!
Glad you liked it. Yes, comparing digital streams would actually be the most accurate approach to this.
I did the exact same test but for a different variable a while back and abandoned it because they didn’t null very well. I’m pretty sure its because the ADC is taking the snapshot of the analogue waveform at slightly different times, so you would need some sort of rig to perfectly match the start of the playback and the recording process. Let’s say hypothetically that between sample 1 and sample 2 in the original the signal jumps between zero and one volts. If you hit record it might be right in between in which case it’d register as .5 volts. Try it again and it’ll be at another point, say .25. Try it a hundred times and you’ll get a hundred different answers for what the voltage is. Of course this doesn’t mean that the signal is different since with multiple samples the dac will calculate the exact same wave, its just got to be the timing differences. I tried my test again but recorded at the maximum that audacity could (384k if memory serves) and the null was substantially better but still not perfect. If you save the null result as its own file and do a frequency analysis of it you should see that the lower frequencies null out better than the higher frequencies. It was a great thought, lovely to see, but I think the only way out of it is to try to get some sort of automated sync going between playback and recording or perhaps do something like 100 runs of each cable and then calculate the averages (yuck).
Thanks for sharing your experiences, drubin. I think my next step in this journey is to come back to human hearing and setup a blind test with a bunch of local volunteers.
Very informative. Thank you. If I may ask a question: what would be a better cable output option from my MacBook M1 chip to my Gustard X18 Dac. Either: US-C Mac to US-A Dac. Or: HDMI Mac to HDMI Dac port? Thank you in advance.
Normally the HDMI inputs on DACs are for I2S which isn't a standard HDMI connection (same physical connection, but different data/signal). That means that you'd need to stick to the USB connection, but stay tuned for an upcoming review of Digital to Digital Converters as I've discovered some very interesting details that are relevant to your question, but too long to explain here.
Enjoyed watching your video and reading the responses. I think both the science and listening tests are important, when it comes to audio gear and cables. As you say you are hoping to open the door to further discussion and maybe someone doing a more thorough scientific test on the matter. It was also interesting to read the comment by Audio Science further down bringing into the equation other possibilities.
I haven't done any listening tests with usb cables but I did many years ago listening tests with cables for analogue gear and can remember at that time ( about 15 years ago) many audio and electronic engineers refuted the idea that cables can make a difference to the sound in an audio system. They relied on their audio and electronic scientific knowledge without recourse to doing A/B listening tests and using their ears for measurement too. Or if they did some test they wouldn't hear the difference. When I was doing my listening tests I decided to go on a sound engineering course to marry my skill of listening to sound and just gain more knowledge.
In sound engineering the human ear is the determining factor to assess the objective and subjective quality of a sound, a sound recording be it a mix or mastering. So, the test measurements here is the human ear. Also, when doing an a/b test with cables or gear I found it important to have the loudspeakers set up properly/precisely in a room so as to hear depth of sound staging, stereo imaging etc. I found this crucial to hearing the differences between cables and their effect upon the sound (eg. sound stage size, width, height, depth, space, tone etc). Also I think you have to take into account the complete system from front end to loudspeaker as the "The whole is greater than the sum of its parts".
As you suggest, It would be interesting to see a test result with a more real world musical signal with all its harmonics through a usb cable rather than just a simplistic sine wave, if that's possible on proper test equipment. From a another of view, it will be interesting to see results of your a/b listening tests of usb cables.
Thanks for the great comment, JMT. I'm in the process of devising a blind listening test that will be my next (and final?) attempt to answer this question once and for all. Ultimately, I believe I am hearing differences because what I've heard has been consistent and often gone against expectations, but some solid evidence in some form would be great.
Same result for me.I have tried various expensive cables, audioquest, oyaide but the best audio came trough the Supra. Transparent, open, deep bass and natural. Great review
Thanks wakesfield. The Supra really surprised me with its sound
Same here although I also buy one generic cheap but 24awg cable just to try it, and its almost the same as supra but again with supra it was little less harshness
@@filipkrstevski5449 I also think that the key for a good sound is a separate usb card. In my case it's a pcie usb3 card that made a significant improvement. I'm on pc
I'm about to review the Matrix Element H card soon and it definitely seems to be an improvement
What would the result be if you did two separate captures using the SAME generic usb cable on your Focusrite ADC and then inverted one of them? The way you do it fail to control for differences potentially introduced by the different ADC captures...
Great test Lachlan! Really fascinating. I think your null test of the cable against itself, particularly Supra 1 vs Supra 2, shows that whatever is causing the differences is definitely cable specific. Shame you didn't share your own ideas as to what might be going on. That is important as there may be people out there who might have an idea for a more specific test for one particular factor (RF interference, timing differences etc).
There may have actually been some flaws in this test so please take the results with a grain of salt. I'm pushing my way through learning about why/how USB cables make a difference with very little support because I'm yet to find a cable manufacturer who can/will explain it in detail.
Great video. I wonder if that might also be an explanation for why ic amps sound "flatter" than discrete designs. Perhaps they are more sensitive to rf noise due to their compact size and so you wind up with a similar loss of "air" and "space" in the final product depending on the noise in your chain.
Possibly, David. I'm still trying to get to the bottom of that myself. It could also be the way feedback is applied with ICs - I'm not sure.
Great video! Big respect for taking the time and courage to tackle controversial topics.
For me cables matter but it's much more fun spending money on headphones, amps, or even DACs. So the cables that go further than the generic options but at a good price are very attractive! I have a Supra USB cable in one of my setups, as it fits that balance for me.
I agree that they're a low priority addition and I'd be super happy with just the Supra - I think it's excellent.
Great work. Would be interesting to compare the differences from source to destination. That is compare the original source recording against he inverted file from the destination, rather than comparing the audio from the generic at destination with audio with other cables at destination.
Hi Nepherim, I considered this and discussed it in the video. The issue there is that the audio from each cable HAS to pass through the digital decoding and analog output stages of the DAC which will always introduce some sonic character not in the original file so you wouldn't be able to separate what changes are from the cables, what is from the decoding and filtering, and what is from the analog output stage.
Excellent work!
Some things I've thought of, which would be interesting to test:
1. The same cables, 5 times, to test consistency.
2. The same cables, but at different times of day, when there's more or less electrical noise from other things in your house or apartment.
3. Do the differences decrease if you use USB and/or power filtering with components.
4. Is there a difference between one of those USB A-to-B short adaptors and a regular cable. Maybe the adaptor could serve as a benchmark, considering it is near only just the plugs.
I love those suggestions! I think there were some other issues with the process I used meaning that the exact timing of the samples by the ADC might have been different each time and this might have altered the results. I've got some new software now that might help so maybe I can do this again one day and add in your ideas.
So glad your doing this and not me... That said while waiting for my local hifi shop to get my Chord Silver USB in I popped into the yellow HIFI? shop and grabbed a $6 cable. I use both still in various DACs (Audiolab M DAC plus and a MF V dac 90 and Cambridge dac magic and when not thinking of what's connected to what and just enjoying the music "the difference's" never enter my head. Sometimes with the $6 cable through the MF i think it sounds great and i look at the 100 plus chord and feel stupid. Love your shows.
Glad you like the channel, Andrew. I still hope to find a definitive answer one way or the other one of these days... Not sure how I'll feel if we prove definitively that your $6 is really just as good... 😂
What I could hear in the recording was that the song played when aggregating a source + an inverted recorded one, the song played flawlessly but very thin and low volume.
This indicates that the difference was not in offset nor was it in frequency shape, but only in amplitude. And since it passes through DAC (the numeric resampling component + the processing component) + xlr analog cable (that will induce attenuation too) ... Thes frequency had a variable amplitude impact where some frequencies where attenuared more than others.
To my ear, the test itself is in favour of numeric cables equity.
The scenario of the test itself is inconclusive to me.
(I am a radio engineer).
But I appreciate very much the maaaassive effort.
I enjoy very much the channel, very insightful !!
How would you tell the difference between simply differential aplitude and some extra audio information that when isolated resembles the original track at a lower volume? Maybe a control where a heavily compressed version of the track against the inverted original to demonstrate how that would sound?
Hi R, each of these tracks was recorded with identical equipment including DAC, all interconnects, recorder, software, etc. Any colouration of the sound should be consistent with all files and that doesn't explain the closeness of the two Supra recordings I compared vs the comparisons of different cables.
Things not compensated for:
- Thermal drift due to difference in time of when samples were taken
- Timing differences due to sub-sample clock error
- Clock drift
- Potential issues with aliasing / filtering between DAC output / ADC input
Maybe I missed something. But what happens if you do another recording from the same cable and compare the negative from that recording to the positive from the first recording?
You didn't miss anything and given more time, that's what I'd have done too. I think there are some flaws with the setup that people have pointed out so I will continue searching for a better test for now...
@@PassionforSound Have you done this test because I can't help but think that the issue here is in the analogue to digital stage of the test. Pretty much every other test I've ever seen is done on the analogue output without any conversion back. I wouldn't be surprised that if you ran the Audioquest through this test several times you wouldn't get complete silence on every pass of the initial inverted file but I would really like to see that test being done for completeness.
Food for thought. While comparing speaker wire connections OCC vs OFC what I found was that the wave profile on the spectragrams were virtually the same as wave vs mp3, respectively. So, yes, you can easily hear difference in wave and mp3, therefore, you certainly can hear a difference in OCC vs OFC speaker cables. It's not too surprising to me then that a similar affect has been observed with USB cables since it's even easier to hear differences in USB cables. Are any of these usb cables used in this test OCC?
I don't think the wire types were published and I don't have much of the info anymore - sorry
Well done and thank you for setting up such an interesting and revealing demo/test. The fact that you even illustrated the exact same setup (including the cable used) could produce slightly different playback result is IMO a nail in the coffin for the argument that USB cables do not make any difference period.
Hi W L, I'm glad you liked it. Some viewers have raised reasonable concerns about timing errors introduced by the recording/A-to-D process so this isn't definitive, but I'm glad it's helping to continue the discussion until others with high quality test gear can chime in.
Can you please tell me do you use usb 2.0 or 3.0 port because I have noticed difference in sound from the two thanks in advance
Hi Filip, I'm pretty sure I only have USB 3.0 on my current PC
@@PassionforSound thanks for the responce I have noticed in my case that using usb 2.0 port for usb 2.0 device sounds much better and more dynamic than using usb 3.0 port which sounds more compressed and maybe a bit louder cheers
That's interesting, Filip! Thanks for sharing your experience
Yet again, a brilliant and informative video! Have you tried switching the cable of the chord mojo 2 (from stock micro sub to usb a cabler to a higher end cable like the curious cable)? Do you think there would be a difference in SQ?
I haven't tried with the Mojo 2, but I'm yet to find a DAC that's completely unaffected by cable changes so I would definitely expect improvement with a good cable. I don't have any recommendations for you though because of the micro USB connection. Not sure which brands make those for audio.
@@PassionforSound I’ve ordered two audioquest cinnamon cables (one is micro usb to usb a, and the other is usb c to usb a) and an audioquest carbon usb c to usb a to demo with the mojo 2 via the camera adapter to iPhone 13 pro. Since Mojo 2 has a usb c line, hoping that the cinnamon or carbon will work, although the SQ using a fiio OTG (usb c to lightening cable) didn’t produce very good SQ compared to stock micro usb to apple camera adapter.
Not sure if having the camera adapter as part of the chain will effect the SQ though?
I can't answer that about the Apple setup - sorry. Good choice on the AQ cables though - always a solid choice
That's a real interesting method of testing. Nice job. Have you tried doing even more of the same cable recordings and plugging motions, then invert on themselves? Having more data like that would solidify it to anyone. I've seen some folks do FFT and finding differences too. Doing a delta like this is probably the best test. Maybe throw it through an FFT so we can see what's going on spectrum-wise?
Please do a follow-up at some point to show you can make the data consistently and prove your method is correct under most circumstances.
Thanks for the encouragement! I've got a bunch of other reviews to get through so finding time for more of this testing is difficult. I do have plans for a blind test experiment in the future which I hope will provide some more food for thought/conversation.
@@PassionforSound will be looking forward
Hi,
Greate test!
Please explain how you synchronized the recorded sound and the inverted sound to be in phase 100%
Hi Dan, I showed this early on in the video where I selected the same exact sample to start from in every file. I've since discovered that it might be imperfect because the ADC might have capture those samples and slightly different times on each playback. I'm planning a set of repeated blind tests to do this again some time (or a version of this)
Madisound sells Supra USB and CAT 8 cables. Closer to my side of the planet at least. Very affordable. My Modius takes a mini B 😠, but BiFrost 2 takes the less fragile normal B connector. I'll keep both DACs so I suppose I'll be getting both cable versions eventually. I'm using the cheapest 5m Monoprice USB cable they sell. Thinner than a shoelace. Thanx a heap for your insightful and valuable work reviewing audio gear.
I'd like to see similar comparisons of COAXIAL and TOSLINK cables if they aren't bit perfect either. Jitter also seems to be an issue, but those rabbit holes might have already been explored enough.
Thanks AA. I need to give this rabbit hole a break for a little while, but I am devising a new test for a future video which could easily include Coax, optical and maybe AES or I2S so stay tuned (just don't hold your breath)
I really like this test. However, many things are not ruled out. distortions and shifts may occur during the conversion of the signal from analog to digital... in my opinion, the test should also include a comparison of 2 recordings made with the same cable. to exclude the influence of possible differences arising during conversion.
I applaud your efforts to try finding a methodology to evaluate USB cables. USB cables don’t really sound like much IMO, but I’ve noticed subjective differences on SPDIF and AES digital connections with different cables. Supra is my preferred semi-pro cable brand. I custom build my balanced cables with Supra DAC and use their finished USB. I think their edge comes from the shielding they implement, no snake oil or exotic materials… And priced very reasonably. Even if it is all placebo, I really like the sky blue, makes the cables almost invisible, still a good investment for looks alone… ;-)
Thank you. Have you seen my Supra Excalibur USB review yet? It's just come out
@@PassionforSound yep, I ordered one. But really just because I felt like burning money - I’m not expecting any improvements… My current USB digital chain is very well optimized… I do trust Supra though, solid engineering and genuinely trying to make their products better where it counts. I have no doubt the cable is better as a cable but will it have any impact in my system? I don’t know…
It'll be interesting to hear your experiences if you feel like sharing once the cable arrives 🙂
very well done comparison !
Thank you. It has a few flaws unfortunately, but I've got some hopefully stronger evidence coming in August
Excellent content... Just like many things in higher end audio, unless you do an actual A/B comparison with these cables, you probably would have a difficult time ascertaining significant differences between one versus the other and as you mentioned, you would be satisfied with any of them.. In short, we are splitting hairs here.. But this is a mentally stimulating exercise for sure as I was fascinated by it... I never buy the most expensive of anything.. I try to end up just a couple of notches below as I value VALUE !
Glad you liked it, P G. I'm hoping to setup a proper A/B blind test with local audiophiles in the future as I think that might be the only way I can prove that there are or are not audible differences.
Just a thought, if you play the original song on the pc directly on audacity or whatever then compare that with the usb cables recorded with the audio interface, we can probably get a sense of which one is closer to the original except for what deterioration the focusrite adds to the table.
The deterioration and sample timing differences are the problem here unfortunately. Otherwise your suggestion would be brilliant
Finally, somebody who's actually TESTING the cables. Did you record right off the analog out? One other test that would have been revealing is how consistent two copies of the same cable from the same mfr would be - a test of mfg consistency, so to speak. You might also take the same copy of any recording (or any audio file for that matter) and perform some math on the two after you time shift one of the copies, change the amplitude of one of the copies, inject a random low-level tone into one of the copies, etc, to see what resultant waveform you get, and how that result would reveal what is happening between two real-world samples. Playing them back inverted and looking at a level meter and listening with your ears doesn't provide as much useful information as actually doing math on two sample waveforms and seeing the results in a third waveform. I've been actually considering doing a similar test myself, but being a little more rigourous in documenting everything than you've been here.
My setup isn't really at a level to do this properly and accurately so I'm forming plans at the moment to use my best equipment (my ears) to test some cables with a blind test. That will be the best contribution I can make to the hobby (in terms of USB cables) at this time. 🙂
Would optical cables have similar results?
I only ask because I recently changed from usb to optical and I swear the sound seems different. Probably not in tonality but just the overall volume. My only thought is something about how optical cables aren’t affected by interference from other electrical devices?? Not sure...
I don't really know anything about optical signal transmission, blitz, so I can't really comment - sorry!
@@PassionforSound No worries, regardless very interesting findings indeed. Thanks for your input in this!
My pleasure
It's natural that the audio files recorded in this way differs little by little. Even if the subject moves in exactly same motion, if the timing of pressing the shutter of the camera is different, the result will be different. The timing inevitably changes during the AD, DA sampling process.
Yes, I hadn't accounted for that. We need someone who can analyse the musical signal exiting the USB cable at the DAC.
Fantastic idea for testing this!
So, I had to give this a try but right away I had interesting results. Maybe I did something wrong. Let me know what you think. In my test, I generated a white noise wav file in Audacity, took that file into Logic, ran that white noise file out of Logic through a Modi 2 and recorded it back into another channel in Logic (with no output). Then I duplicated that channel (with no output), did it again, and phase inverted the second recording using the stock Logic Gain plugin (because it has a phase inversion setting). Interestingly, there was a very audible high end hiss (metered to be a slope increasing from about 3k to 20k).
Then I exported the two recording and took them into Audacity to see the individual bits and yes there's a visible difference in the bits (and they're lined up perfectly). Weird, right? Well, it gets weirder. I swapped the USB cable going into the Modi and did the whole thing again and, lo and behold, there's a bigger difference. This time the slope starts metering at 2k and there's a 5db difference!
WTF? IDK. Well, now I have a new hobby. :)
That's really interesting and a nice way to setup the test. The results are curious though and make me wonder if something is going on in the processing of the signal to create that noise, spending on how loud it is without any additional gain applied.
I should add that some people pointed out that the exact sample timing in the recording stage (analog back to digital) could be introducing some variance between each recorded sample.
@@PassionforSound Yeah. Thats what I thought too but I'm thinking that if it was a timing issue, you'd get a chorus effect and not a filtering. I exported the white noise recordings from Logic and opened them in Audacity so I could line them up at the sample level and ... get this ... there are subtle differences in the waveforms of the two recordings. So I'm left thinking that the bits traveling through the same USB cable are being interpreted differently at different times (even a few seconds between recordings) OR the analog conversion of those bits in the DAC is fluctuating slightly. Analog would, in theory (maybe), have more natural fluctuations based on fluctuations of temperature or power surges or something but that doesnt explain why there's such a difference between the results of different cables. Right? IDK
Edit: when I say there are differences in the waveforms, it APPEARS to only be fluctuations in amplitude. The basic shapes of the waveforms match but the visual height of the peaks and SLIGHTLY different.
@@PassionforSound another thought: I'd like to run this test again with the Bifrost to compare results. Have you tried other dac setups?
Ok. Sorry for all the replies but my first tests were with Belkin cables. Just ran the same test with the Amazon basics cable and there IS a chorus/flange effect happening on that. Checked my setup to make sure I didnt do something strange, did it again, and got the same result. Even retested the first cable to see if I would somehow get flange with it and still got the original result. IDK what's happening. LOL
Hi Lachlan 🙂, I have read ur big blog review of usb cables (AQ, chord and ifi) from 2016. It seemed that the best value for the money was maybe the AQ carbon? Where the supra 2.0 would be placed performance-wise between the AQ pearl, forrest, cinnamon and carbon? If you can compare just from the memory, even broad estimation would help 😃 thanks. Iam trying to find something that would increase the sound fidelity of my usb portable AMP/DACs for instance the centrance dacport HD. 🙂 AQ seems to be the only option for usb A to usb micro and I would like to have something atleast on the supra 2.0 level which I really like. So the supra is it more level of the forrest or the cinnamon?
That's a great question and I don't have the Carbon or Cinnamon here anymore. I think probably the basic Supra cable in this review is probably at the Cinnamon level while the Excalibur USB from Supra is more like the Carbon I think.
@@PassionforSound thank you very much, I bought the pearl vesion ( it arrived today) for trying and I hear it exactly like u described it in the review 2016, The sharpness around the edges across all freqs. is now gone and the sound is within the bigger space with better imaging and separation. It is more open sound. I will return the pearl and try also the carbon (I am looking forward to it how i will upgrade tp sound again ) 🙂. And propably I will keep that one.
Excellent work, as ever. Thank you. I use both wireworld and supra cables. Your conclusions fit with my experiences. One thing I do know is that usb data transport to a DAC is a frustratingly tricky thing. Network audio adapters, such as the Pro-ject audio S2 ultra DO make a difference. USB cable construction DOES make a difference. Beyond that, I do not have a clue 😂. Anyway - thank you for attempting to benchmark the effects.
Thanks Tim. It's an imperfect test to say the least, but I'm with you in believing that there are differences. Now we just need to find a way to identify what they are and why.
Great video.....my question from my own recent test comparison is the affect of the ferrite noise blocks. I found that my USB cables with ferrite noise blocks on both ends significantly improved sound in my unscientific test. The result was significantly more range, depth, & separation. Maybe you can test a basic USB shielded & unshielded cable and then use ferrite blocks on the cables to see if any affect can be detected. 😳👍🤔💯
Ferrites are interesting because they can be good or bad based on my general experiences. It's unlikely I'll get a chance to do that test for you, but it would be curious
Good point. My Chord Qutest came with a generic USB cable, but with a ferrite block on it. Lachlan, do you have Rob Watt's views on this test? I would be interested to know his views.
I don't know Rob's views on ferrites, but the Wave BNC cable that he uses is covered in ferrites. I'm just not sure if a single ferrite on a USB cable makes a big difference
Would it also be worth conducting a similar comparison test, of the controversial, audiophile network switch? Does a network switch enhance signal quality or not?
Potentially. I have recently reviewed the JCAT USB and network cards so there's that...
very nice video, and yes it is a full scientific one 😊, because you have used positive and negative control and more than one measurements 😁, ideal would be 3, but this is still enough to prove💙. Iam curious if those cables would make even better results when combine with the Audioquest jitterbug noise filter or not ...🙃
The JitterBug is an interesting one. I own one (or maybe two?) but don't tend to use them anymore as I'm not convinced that they improve the sound. Maybe I can do a comparison of filters one day
I have two Audioquest Forest USB cables. They are plugged into D90 and ADI-2 DAC.
They don't sound any different than generic cables.
That's fine. I've differences between a number of my cables (and often against my expectations so it's not confirmation bias) and that's why I believe there is more going on than what sine wave measurements are telling us.
Very interesting stuff! Is there a difference between using a laptop or a phone to send the music to the dac? Thanks!!!
Hi John, the quality of the USB source will definitely make a difference. I haven't really delved into phone vs computer outputs though and it probably varies a bit from one devic to the next
@@PassionforSound thanks for the response!
No problems 🙂
Awesome test. Love the new Supra cables!...Anyone know what the test music was? trying to find it.
Glad you liked the video. In hindsight there were a few issues with the test so it should be taken with a grain of salt (a large one perhaps), but I'm working on something a bit more scientific so stay tuned!
I can't recall the track I used for the testing, but it will be from Epidemic Sound (link in description). I can see if I can find the track name for you if you'd like.
Can i open up another can of worm for you, would you do another review like this but with RCA and XLR cables ? thanks
Hi Steve, I certainly might at some point, but I'll wait until I can get some better test equipment first I think (see your other comment for more expanded explanation)
Hi Mate, have been viewing your videos for awhile and just recently subscribed to your channel. Viewing this for the first time Feb. 02, 2024! Great video, review and effort!! 💯👋👍🎉 One very important thing all should remember is system synergy!! My personal experience being in this hobby... which can drain you wallet easily if you let it... is to do adequate research and be happy with YOUR audio HiFi system. In the end our personal taste and what we hear is all we need to satisfy. But if I may say again system synergy, I believe plays the most important part to end sound AND if I may also add is room treatment (IEM/headphone not included). Concerning digital clean accurate clocking/timing of the data transfer for processing is soooo important. Consider what DDC's have done for improved sound quality!! Take steps to reduce/eliminate noise, thus having a lower noise floor and "watch," or should I say listen to YOUR systems sound improvement!! I'm interested in the Supra brand/model you highlighted. Please provide website and model used in this test. Cheers from Northern California!! Thank you.
Hi and thanks for the sub!
I'm glad you've enjoyed the videos. The test in this one was actually a bit flawed, but I've left the video up because it still sparks discussion and curiosity for people.
The Supra cables aren't particularly well advertised or distributed so I don't have a site to direct you to, but the model name is just the Supra USB 2.0 cable. If you search for that, you should find a local online or storefront retailer.
Thanks for this info, I have the OYAIDE and want to find a better natural and more background sounding cable,...Any suggestions, is the Excalibur better overall over the Oyaide ? my system is very revealing.
Thanks !
I think the Excalibur is a nice step up from the Oyaide without spending mega bucks. 🙂
@@PassionforSound thank you 🙏 the one thing is that Is it lacking bass extension ?
Not in my experience. It has a nice sense of energy through the transients that might draw attention away from the bass compared to other cables, but I've never felt a lack of bass with it
@@PassionforSound I apreciaste your input very much, I used to have a Mad Scientist ultra magic USB that receive at the beginning of this yer as the ultimate usb cable connected to my Lampi Baltic 4 DAC from a Jcat XE usb card from a Tweaked Audio PC, but as time was passing, I found out in a forum that the brands of a knockoff of AliExpress Odin 2 USB, pure silver and XANGSANE USB 4N pure silver, and the set of tubes that I had at that moment in my Lampi Baltic 4 apparently this cables were more resolute and more balanced than the ultra black magic while the OYAIDE was left for dead - resting somewhere in my Audio room at least two years, then I pulled it out and for my surprise it sounded the tone right with my last upgraded rectifier tube with an adapter and all came to show me that the more tone balance USB cable was the OYAIDE and not Odin 2 or XANGSANE , so System synergy is always present, now I know because of these changes in my system among others that are showing its true colors that I could even find the better usb cable, OYAIDE is the best with my system so far but it could be more clean and resolute, …aiming to Supra Excalibur, Intona pro reference or Tubulis Colncentus, hope you be able to test those
Thank you so much
Yeah the supra cables are hard to get . I went with wireworld for now because I am fairly new stepping into the audiophile scene , and they seemed reasonable in price and availability .
Yeah, it's one problem with them. Hopefully distribution improves, but the WireWorld is a good choice as a start 🙂
What if you need to use a 8m cable? I believe the longer, the more issues arise right? Any advice which cable would be suitable? Or do no longer cable versions exist?
Hi John, that's an interesting question and I believe that 5m might be limit, but don't quote me on that. I don't have any awareness of who would make longer cables like that - sorry!
@@PassionforSound i just ordered the supra 8m Version gonna check it out saturday :)
Interesting. Please let me know how you go
@@PassionforSound it works! It sounds veeery close to my XLR recording, and far better (more Highs, less muffled) than with my cheap printer cable!
Glad it sounds better, John. I need to doore testing in the future to prove this whole USB cable thing, but I believe they make a difference and I'm glad you're hearing improvements
first thing you have to do with usb cable is... remove the +V from the cable that transmit the signal, DIY a splitter and connect the +V (if/when required) via an additional cable running outside the usb cable... if not required do not supply the +5V at all, insulate the +V contact at the source side... a HUGE amount of noise into the usb is from bad power supply
The differences we hear are related to voltage differences and magnetics. The voltage differences, that run through the electronics, change what we hear, even at millivolt levels. The magnetics, from the current going through the cable and electronics, create distortions in the signal as well. I've used a digital volt meter on my power distribution box and have seen 3 to 5 volts difference all through the day. Even 1 volt difference will change things at the millisecond and millivolt accuracy of digital audio.
I was surprised that the recording you did the next day wasn't more different. Thank you for doing this video and for showing how much the cables and electronics can make a difference.
Thanks Dan. Glad it was interesting. I'll try and do a multi-recording test with the same cable sometime soon and see what it shows us.
Did the temperature of the dac chip change during the test ?
I have no reliable way to measure/know, Dave
Excelente video, muchas gracias!!!
My pleasure
Really good and informative. Well I just got a audioquest usb b cable for bifrost 2. I think I hear a difference or it might be my mind telling me there is a difference.
It's a tricky one, Kam. I believe you're really hearing a difference, but I'm still searching for the explanation.
Wondering if this is electrical noise (analog) from laptop getting transferred to the DAC through the USB cable, and affecting the analog output of the DAC? So probably no issues on digital side, they're still bit perfect, but the issue is more of electrical noise leakage through the USB cable?
Great video, we needed this and the other video on the USB cable subject.
But this was the digital end, how about the analog from the amp to the HP?
Hi Gordon, that's a nice idea. I think there's less debate about analog cables, but I can probably manage something using a range of IEM cables that I own.
tho is your soundcard bitperfect and do you use good usb port like from special usb pcie card?
Hi lex, there's no soundcard involved and the USB port is a regular USB port just like most people would be using with these sorts of cables. That's the point of this test - to start a conversation around the fact that the digital audio system isn't bit-perfect in the sense that my experience suggests that things CAN sound different through different USB cables and thatay reflect any number of factors including how the USB socket performs and how that might affect different cable designs, etc.
@@PassionforSound yes kinda crazy that 1&0 would be loss or corrupt in any of the digital transport layers in these days and age.. maybe optical was the way to go if thats the case..
I think optical might have its own issues ☹️
Did you consider to make two identical recordings with a same cable, and do the 180degree off phase test?
Yep. It's in the video
@@PassionforSound no it is not. You do not understand, that you can not align two analog tracks perfectly. Never ever. It is impossible. And that is why you get this result. If you use only digital signal and do this same test the result will be perfect silence every time. Every single time.
Thank you!
So, i would love to see you compare the Audioquest diamond with all the cheaper cables and see which one is the closest to see which one is closest in performance to the high end cable
Hi Ronald, I'm working on a blind testing setup at the moment so stay tuned... 🙂
While inverting a file may perfectly cancel that wave out. There are other variables that could affect a specific pass through analog. Try doing a capture from the same cable a number of times to see if there are just variables in the process. If perhaps 3 passes from a cable dome with 2 different cables? If each came back null for each cable. Then that variable could be reasonably ruled out.
Hi Glenn, I agree and will try to do that test in the future and share it on the channel. It will be interesting to see if the comparisons of files from the same USB cable are consistently more similar than comparing files played through different USB cables
@@PassionforSound I was thinking the same. I don't think it's fair to call your test invalid. The test of the cable against itself recorded the next day seemed to have a lesser difference vs. the other cables suggesting that independent of dac and adc variabiliy the cable swaps could be a factor. Repeat testing with similar results would strengthen that position.
@@PassionforSound That will be very interesting. Even perhaps take two cables and do three transfers with each switching back and forth between them. This should even include testing the individual physical connection each time, as opposed to just doing three captures without any physical resets. Connector contacts being wiped cleaner by reinsertion.... and any "learned" interfacing like clock cycles matching more over time.
As a long time tech, from old school audio to IT, I am hoping similar differences will be noticed so we don't have to try to find technical explanations for USB cable sound.
Fingers crossed. It might take me a while to get to this one so stay tuned...
Thank you for this information. Very interesting. At one time I started an audio cable company and fro fifteen years it was my Walter Mitty second job. Great fun and I learned a lot. Cables do make a difference. I think the issue of cables making a difference and cable companies indulging in hype and pseudo science to sell expensive cables gets mixed up in peoples emotional opinions about cables. Yes there are grifters in the business selling expensive cables with no or little real sonic improvements of more modestly priced cable. They make them exotic looking and very expensive hoping to snag people with more money than sense. That said, the fact that some people are abusing their customers for profit does not change the objective fact that you can hear differences. I have also found that some of the most adamant cable haters have very strong opinions based on little experience and a lot of conjecture. I suppose if someones hearing isn't good and they have such modest equipment that the differences are not apparent, then they cant benefit from the nuances. If you have any kind of revealing system and your haring has not been damaged by the enemies of healthy ears like power tools, motorcycles, race cars, firearms, the military (explosions and firearms), rock concerts etc, then you may benefit from checking out some alternatives cables and see if they benefit your system. You can start by getting your hearing checked.
Thanks for sharing your experiences, Stagger! I completely agree that there are poor examples of cable manufacturers out there doing a disservice to all of the excellent manufacturers that really care about making a difference for their customers.
@@PassionforSound I am sure, SURE that there are scammers who make astronomical profits with no real benefit for the buyer. I also know that not all dacs sound the same, not all cables sound the same, and so on.
Some people claim they can hear nonexistent sound quality improvements just to show off their 20K $ cables and bionic ears. That's human nature, we can't do anything about it.
Envy is another interesting trait of human nature. It was already known in ancient times, see here: ruclips.net/video/bw_5JfrQ-aE/видео.html
How many people are eager to call those who can actually afford a 30k $ amplifier a fool? Too many! But this doesn't mean that the amplifier sounds the same as a 300$ Chinese box. BTW: there are people who easily spend 5K $ / year in smoking, alcohol and gambling. In twenty years that's a magnificent hi-end stereo set. And yes, it sounds much better than the Chinese combo. Different Choices.
Oh, btw, it is clear that some measurements (not thinking of anyone precisely) are done exactly with the aim of "proving" that hi end does not exist and is all in the imagination of some "audiophools". This thing has a market: there are thousands of people who love to hear that because it makes them feel good and smart. That's what I think.
where did you get that light blue "usb cable" hoodie you are wearing in the intro? Looks cool!
It's from my merchandise store, Kawshun. You can now access the store from directly below the video or via a link in the description (to Spreadshirt)
I boutght the Curious Evolved for my main speaker system a couple of years ago. Surprisingly it sounded like a downgrade vs the Audioquest Coffee it replaced. I put the Audioquest Coffee USB back in and put the expensive Curious into my desktop headphone system to replace the generic. The result was that my desktop was dramatically improved. I would never have invested so much for my desktop, but I am forever grateful that I did, given the amount of time that I spend at my desk. I cannot explain either result.
Thanks for sharing your experiences with the cables. Sometimes, system synergy comes into play so I'm glad it worked out for the best!
I neglected to mention that I preferred the Curious to the Audioquest Coffee on my desktop rig. It is currently attached to Xduoo TA-22 w Apos Ray Tubes. The sound is clean, smooth, dynamic and engaging. On my speaker rig it was plugged into an MScaler and introduced an unpleasant higher frequency edginess. In contrast it just seems to disappear with my headphone DACs and amps. It required no breakin plugged into my RME dac when i first tried the Curious. It sounded quite good immediately.@@PassionforSound
Very interesting video but I'm still inconclusive. You've convinced me that with this testing method we do measure differences, but to really convince me there is a difference between cables would really require more scientific testing.
I also understand that is easier said then done.
Hi Richard, I'm right there with you. I subjectively believe there is a difference between cables, but this test hasn't given me definitive evidence due to the basic setup I used for the testing. I would love to see some of the people with high end test equipment do some testing using music rather than sine waves, but I'm not sure if that's even possible.
Fantastic video!
Thanks Themos!
@@PassionforSound did you put your Curious Evolved through this test as I understand you did keep that cable?
Hi Dave, no I didn't, but I have a new test coming up soon with the Evolved.
Hey Lachlan I know this is an old video but you have heard and compared two cables that I'd like to hear your opinion on. I just purchased an audio quest coffee which I know you enjoy. I really enjoy it except for the fact that I'm treble sensitive and I find it has a glare or brightness that bothers me. You have said that you prefer the Supra to the Diamond but what about the coffee? Would you say it has similar characteristics with less brightness? Does it mellow over time? I'm attributing brightness to silver content in my system with regards to interconnects and speaker cables- and have eliminated any silver. Is the Supra worth a try or is the coffee something I should let settle in/get used to?
I don't think the Supra is smoother than the Coffee. Perhaps a Curious Evolved might be a better option.
Interestingly too, silver cables don't seem to influence the sound in digital cables like they do in analog cables so don't shy away from a silver USB cable (I'm not referring to the Curious Evolved with that statement as I have no idea what's inside that)
@@PassionforSound Thanks for the reply. That's what I was thinking when I bought it but I still find it has a brightness in my system. It replaced an audioquest pearl and switching back and forth I can't go back. I'll leave it in for a month. Maybe less silver :P or yeah a different brand. Thanks again.
Just an update- took my ifi igalvanic3.0 out of the chain between my PC and Gustard A22 DAC and the brightness is gone! The sound is much more natural and immersive. Audioquest coffee sounds great on its own and terrible with that specific USB isolator.
That's really interesting! Thanks for sharing
I recently purchased the Oyaide class A usb cable and it easily dethroned the Audioquest coffee from my system. The absolute sound review was spot on about the Oyaide Class A usb cable . In comparison the Audioquest cable artificially expands the soundstage and voices and drums can sound hollow. The Oyaide usb sounds more accurate in timbre quality and nails the imaging and soundstage . High frequencies on the coffee are more detailed BUT can sound bright and artificial at times . Oyaide class A is the best Usb cable I’ve tried for my system . It really sounds analog to me. Cables I’ve tried: AQ Coffee, pearl, carbon, Straight wire usb , Supra usb, monoprice monolith, Oyaide Class A, blue jeans usb, and other generic usb cables.Usb cables make a significant difference to my 👂. One thing I’ve learned is that higher price doesn’t mean it’s better, it ultimately comes down to preference and taste. Another thing I’ve learned is that I prefer the sound of copper cables over silver or silver coated copper. Silver cables can sound more detailed but artificially so.
Hi Aztek, I agree that expensive doesn't necessarily mean better.
It's interesting that you prefer the Oyaide to both the Coffee and Supra. I really like both of those and I think it's a bit curious that people can disagree over which digital cable they prefer. That says to me that there is a definite subjective element even when it comes to digital audio
I think bit-perfect was never really an accurate description of how DAC works: it can be bit-perfect up to DACs DDC (CRC is verified as ok), but as soon it enters the DACs digital processing pipeline (whether that be DSD or R2R implementation), DAC will do some interpolation, upscaling or some other type of modification of digital signal to adopt it for internal usage. So we always lose that "bit-perfectness", or we should say its only in the context of the source (PC that uses ASIO/WASAPI).
Measurements of USB cable against original inverted signal is also showing influence of amplification and maybe even ADC converter (we saw that in the video as much lower signal), so very interesting video!
My only concern would be an alignment of recorded signal vs inverted original signal - if its just one sample off (we have 44100 samples in 1 second, a lot!), we will hear it as a difference even though there's none maybe. You should probably move one signal left/right few samples until you find the one with the least audioble difference (sorry if you did this already)
All great points, Hrvoje. I attempted to align the files manually at first, but was having much more trouble than by using the graphical representation of samples. I recognise that this might also be slightly inaccurate due to the timing of the ADC in the Scarlett 2i2. You're right about the bit perfect signal being changed once it enters the DAC, but my concern is more the assertions that the data reaching the DDC is the same regardless of the cable (as long as it meets specs). I'm keen to see this tested further to prove/disprove my experience and the experiences of others when listening to the rapid fluctuations of music rather than the predictability of sine waves.
I think that this is a great misunderstanding of what "bit perfect" means. When looking at players/sources, the claims of bit perfect are valid but it's not what you think and we don't lose bit-perfectness. Example; Roon is a bit-perfect digital audio player. This will mean that it will read the file and send on the target media (SPDIF or USB) to the target interface (the target DAC) without downsampling or changing bit-depth. That is the bit-perfect claim.
Bit-perfect doesn't mean that there is a 100% perfectly square digital signal of the exact representation of the 1s and 0's received by the DAC processor. ......and no one can make that claim either.
I read several of the comments and I'm surprised I haven't yet seen anything having to do with optical formats. I wonder if there would be any advantages to converting the USB to Toslink at the source and going directly into the DAC's Toslink port? Of course, few (if any) computers will have a Toslink out, so a conversion would be necessary. Maybe in all your spare time, you could do a comparison.....?
There are both pros and cons to toslink based on my experiences, reading and discussions with people. Toslink still has conversion at the source and the DAC (into light pulses and back into electricity) so the quality of the toslink interface will play a part in its performance. I've also still found toslink cables to make a difference so there's that side of things too 🙂
The variaion between supra and supra 2 probably comes from 2i2?
But no matter how, the chard output with different USB cable used is definitely different. Good test. Thanks.
I think the ADC process was creating some differences due to the exact timing of samples. Hopefully it's continued the conversation and curiosity for people to explore and understand USB cables for themselves though 🙂
Companies have been discussing the merits of better cables however most do not show details as to the justification for their claims. Thank you for the more scientific approach
No problems. There were actually some flaws in this test that I discovered later so it's not entirely valid. I think part of the challenge is that there are so many unknowns about how we perceive sound and even exactly how the perception of sound can be influenced by the factors altered by different cable designs (metals, structure, etc.)
I've done this invert test with lossless vs lossy formats and the usb cables here eapecially the wireworld and the diamond have a bigger difference to the generic wire than mp3 320 vs flac files
Unfortunately, some variation may have been caused by subtle differences in the sampling running of the ADC as it converted the playback to the digital recording file.
Very clever. Great video
Thanks Andrea. It actually had a few flaws in hindsight so do take the "evidence" with a grain of salt, but there's no doubt that cables are somehow influencing the sound - now we just have to understand why/how
I think the dominant factor in most cases is cable length. Would be interesting to do this with different lengths of the same cable
Length actually makes no difference at all up to 5m for USB audio. Which is what the spec is.
@@phototristan fair point but more precisely IMO is that there will be differences even between various cable lengths less than 5m but these differences are deemed to be within a predefined tolerance.
Measurements performed by Amir at ASR confirms that length of USB cables are indeed what matters.
Hi HP, I'm pretty sure Amir was focused only on jitter/noise if I'm remembering correctly.
Jos, this test of mine may have some technical issues due to the equipment used so don't hold it up as definitive proof. I'll keep working on new ways to test what many of us believe we are hearing though
@@PassionforSound than you need to test in the field of psychology and not physically. It doesn’t exist. It’s been long proven.
I would echo the comments about the testing below. You would not expect to hear the song so clearly, it should be random noises. To me this suggests a difference in amplitude between the samples. Have you tried (manually) tuning the amplitude to see if you can reduce the difference? The smallest possible difference would be closer to the real difference in sound signature. It might be worth trying to stress the cables with RF interference to see what happens, eg run an unshielded power cable alongside them. Great idea though, thanks for doing this.
Hi Barton, if the amplitude is changing it speaks to influence from the USB cable because that's the only variable in the "experiment". This definitely needs a repeat/improved attempt with higher end testing equipment, but I do believe it's beginning to show us that audio wave forms (as opposed to sine waves) may be altered by USB cables due to the rapidly fluctuating nature of the signals.
I have to say that this is a fantastic test. I don't think anyone can really argue with this test, that's a pretty scientific and clever way to go about testing quality of the data transfer efficiently and easily in home environment. It's of course not perfect (goes for any single scientific test tho) but as far as what the hypothesis is and what the results are, unless there's something seriously wrong with the process (which is so beautifully simple that any serious wrongdoings should be blindingly obvious - so I really doubt it) this is very eyeopening and I have to say I was extremely skeptical. As to whether or not the difference in listening can be that staggering, I guess as most things - it depends. But this for sure does show that digital audio is still audio, and as such is still susceptible to tiny differences in the live bits that can come out altered on the other end. It's really fascinating that it's even possible though.
Sadly, the precise timing of the sampling as the signal went back into the interface might have been introducing some variables. I've not taken the video down because I think if it helps further the conversation and reduce the black-and-white views of cable influence then it's only a good thing, but it's imperfect as a test. I have some more content planned, but need to arrange some things before I can produce it. Stay tuned... :)
@@PassionforSound I'm sure it's not a perfect test, however comparing the same signal transmitted twice over the same cable, shows that the process is good enough for what was demonstrated. If this is true then this has serious implications not just for USB but other digital audio and video cables such as HDMI often used in high end audio setups. Which is another super complex layer that would probably benefit from deep investigation but that's a rabbit hole so deep it makes my head spin tbh. Anyway, thanks for the cool video I'm subbing.
I appreciate the sub - thank you!
I would think grounding might have something to do with it: if (and this is the question) the usb cables influence grounding noise which to some extent enters the dac thru the usb input, than the analog output signal might be influenced by that noise to some extent as well (perhaps by altering operating point of output stage ..?)
If the cables would influence only the signal part, than i‘d expect mostly white noise as the difference between inverted and non-inverted audio recordings.
Comments appreciated :)
My Matrix Element H review would suggest that a lot of this depends on the DAC.
I highly appreciate your efforts and open mind to the subject, but there is something seriously wrong with the results. The dB levels of difference are bigger than even the difference between 128k mp3 and WAW, which is about -40 dB (I made such a test once for myself). There must be a serious amplitude or alignment issue as most of the time you hear the original song with a lower volume. Keep in mind that filtering effects are hardly audible when the time difference is very small, say under 1ms, which is still a very high value for perfect canceling. The only way to check perfect time alignment is to have the same sample count and aligned up correctly, which is hard to achieve, even with good gear. Let alone with a Focusrite, which is not the most accurate thing in the world, be it timing or amplitude (the mere fact that it has constantly a preamp in its path is suspect). So the results are questionable at best.
If USB audio was that bad, it would NEVER be used in the professional world, and it is used heavily even among the highest levels. Sure AoIP, AES and Thunderbolt is there, but mostly for other reasons than sound quality (better data redundancy, better latency, higher data transfer, etc.)
One thing to note though that it is always mentioned that cables can make a difference because of noise, etc. but it is hardly discussed that what manufacturers should do to avoid those issues. My opinion on the thing is that if a piece of equipment has a notable step up from its USB to other inputs, or cables do make a difference, then it has a lousy USB implementation.
Anyway, I think you are on the right path, but the conclusion here is drawn too early.
Hi Andras, I appreciate the input. My only conclusion here is that the assertion that digital is digital and immune from changes doesn't appear to hold up. I agree that the focusrite is nowhere near pro-grade testing equipment, but I think someone needed to start the conversation about testing USB audio cables with music and not just test tones. There are too many examples of people hearing differences and forming preferences (not always aligned with their expectations so not confirmation bias) for me to believe there is no difference until I see solid evidence using music rather than test tones.
@@PassionforSound Thanks for your response! I agree with you that this topic should be discussed and researched, and I really appreciate the way of your thinking here. Test tones and THD/IMD plots are not sufficient here, and I think you are on the right path. Just be aware not to draw any conclusion too soon! The really hard part of any scientific method is questioning our own results and search for possible influences and eliminate them. Of which there are quite a few in these first results. I am not sure myself how to conclude a proper test, I admit, but the whole DAC/ADC chain seems way less accurate than what you actually want to measure, even with instrument grade converters.
One thing to be aware of though is that some less educated and/or more opinionated subjectivist people come and cite your results as prime evidence of their confirmation bias, and not sseing their limitations. But that is something one has to live with when going public. Not that there aren't any people like this on the naysayers side...
Anyway, keep up the good work, and don't lose your passion!
Thermal drift and sub-sample timing inaccuracies. Audacity is not the software for making such a comparison.
Food for thought. While comparing speaker wire connections OCC vs OFC what I found was that the wave profile on the spectragrams were virtually the same as wave vs mp3, respectively. So, yes, you can easily hear difference in wave and mp3, therefore, you certainly can hear a difference in OCC vs OFC speaker cables. It's not too surprising to me then that a similar affect has been observed with USB cables since it's even easier to hear differences in USB cables. Are any of these usb cables used in this test OCC?
@@PassionforSound you don't understand what Andras is telling you. You were not testing differences in digital side of things, you were testing the analog side and alignment problems. That is all you did. You proved nothing. Your test was not even close to scientific.
You could repeat this test with some sine waves at different frequencies, as well as a sine sweep. That might accentuate the effects of noise, and help isolate where in the frequency spectrum each cable performs differently.
Hi Steven, my concern with this is that sine waves are not representative of music and this is where I think the biggest gap is in the subjective vs objective conversations of USB cables. Too much of the objective conversation is looking at measurements of sine waves that do not fluctuate rapidly like music does.
@@PassionforSound True, I think so, too. But it might be a way to backup and expand on what you've already done. Especially if the differences between cables can be thought of as distortion of some kind.
Meaning, if two different cables can't even produce identical sine waves, then they couldn't hope to produce identical audio signals.
Good points, Steven. It might also be a good way to see if the system is creating issues with timing slips, etc.
@@PassionforSound Love your videos, btw!!
It could be Gremlins! LOL!! Very nice to see this test performed. If physical properties are different there must be some audible difference. However, I am a little surprised the difference is so obviously audible. I use USB cables in the $20 range and physically, that should be a sufficient match with my gear, and my 57-year-old ears. This video is going to have an interesting ripple effect within our audio community.
There were some issues with this test that viewers have pointed out (the timing skew from the ADC recording the music) so take the results here as questionable. However, I still agree that physical properties are definitely influencing the sound. I've got a new test planned soon thought that should be pretty interesting. Stay tuned! 🙂
Nice test. Even if not perfectly scientific for a variety of reasons (see comments below), if opens a quite exciting discussion about USB-cabling - wich I believe may be a bit overrated by some of us, if not many.
1) In general, I do not assume that 'bit dropping' is a major issue in USB music transmission. If you look at the USB protocol you find that most of the data rate is consumed for protocol and data redundancy, and the true data make only a fraction of the overall stream of bits being transferred. As a result, the DAC's error procession should not be a true reason for sound differences between cables.
2) What definetly will make a difference is noise that may get fed into the cable from a poor USB sending electronic circuit, hitting on a poorly filtered/isolated USB input section of the DAC. Or it origins in the circuitry of the (poor) DAC itself and finds its way into the audio stages. In this case, different cables may act as noise cancellers for resaons such as plug material, shielding of the cable, etc. But I am not sure.
3) Even worse, als described in one of the comments below, will be clock jitter and drift on either side that may result in a slighty different analog signal curve being produced in the DAC side, compared to the sending side. This is the major reason for central clocking in studio setups, especially when combining audio and video for broadcasting.
4) To all of us who still believe they need to spend more than 100 bucks on a single USB cable for good music listening experience, at the same time discussing the benefits of QUOBUZZ and TIDAL over e.g. Spotify, I strongly recommend to take a look at the long transmission path the original sound file from any of the streaming provider servers takes to their homes. In specific, take a look at the internet cabling in most of the homes: thin pair of twisted telephony cord - nothing more!
5) My personal conclusion: the USB cable, unless it is defect in any way, is the least priority to go for.
Instead, I continue spending any buck saved here on further optimizing my room with absorbers, diffractors and other nice stuff to make it sound as good as possible. The bang for the buck will be far higher than with ANY cable in my stereo, including RCA, XLR, COAX, Speaker, or USB.
Final remark: How many hours of listening to great music could we (me included) have spent in the past, let's say 3, month instead of seeking the internet for the best sounding cable?
Great points all around JoF! I definitely agree that our focus should ultimately be on the enjoyment of the music rather than losing too much time worrying about tiny differences that may or may not exist. That is of course unless you're a reviewer or you find it enjoyable. 🙂
I've often thought about the full signal path as you mention whether it be the network cables or even the internal wiring of an amp before it reaches my expensive headphone cable. It's an interesting question as to where to draw the line with upgrades to cables and accessories.
@@PassionforSound I fully agree: your audience requires you to go deep into some technical specs and listening comparisons.
And definetely, there is an audience who wants to experiment with different gear, just to be emotionally safe that they did everything they could to get the utmost performance out of ther systems.
And, there is a large industry that serves them with a variety of products, some being solid while others come close to 'snake oil'.
But even if one feels good with a 'snake oil' accessory - hey, that's great for all in the game, isn't it?
This is exactly when you come into play with your great channel, combining technical with emotional perspective, withouth blaming either side - great job, ideed!
I totally agree, JoF. Thanks again for commenting!