Entreq Tellus grounding,in england

[Fiddle Faddle]

BE718,

You yourself are completely missing the point and inventing your own red herrings. The more hard and anecdotal evidence that comes to light that these boxes actually do something beneficial (the opposite of which has been the entire foundation of all your arguments set fourth in this thread), the more you come up with diversionary tactics to lead the less astute astray. First you complain about the source I used, now you are complaining about the outright performance of my system. I never made claims about it being a high performance system - I said that right from the very beginning. My whole point is everything I have ever said about Entreq in these forums is that regardless of how bad or good your system is, the products will potentially improve it.

My Corolla has an engine "problem" too. And the "problem" is that it only develops about one quarter the power of rockitman's Ferrari engine. This shortcoming absolutely devastates me every time I have to take it down to the local shops knowing I don't have close to half a megawatt of power under my right foot, but I somehow manage to scrape by and live with its lowly performance, just like I manage to live with my simpleton hifi gear. Perhaps if I had not had to retire early due to health issues I would own a superb system by now that sounds wonderful and measures brilliantly, but it wasn't to be and I live with what I have and accept that it is not perfect.

Perhaps you need to be reminded that the one of the fundamental claims made of the box is that it reduces noise. Do you not see a difference - any at all - in the noise floor now that I am actually recording a live output between the entreq and non-entreq files (as opposed to previously when I merely hit the record button with no live input source at the other end at all)? I do see a difference. To me that is now irrefutable and absolute - at least in my system. Everyone with a pair of eyes can now see it. So now you don't even need ears. Eyes can just do the job too now. I would have thought that would satisfy those who prefer not to listen and instead rely on measurement.

You may think that pointing out the noise limitations of my particular system (which in the case of these particular tests is computer based so is always going to struggle more with noise than expensive dedicated components) discredits the Entreq product but the truth is it is precisely the opposite. We are seeing what it is capable of doing on the sort of mass market consumer gear that people may often use (given that the boxes I am using are very inexpensive) and most importantly we are seeing an improvement.

The fact that people here using exceptionally quiet, power-conditioned, state of the art balanced equipment with Entreq is something you continually avoid addressing and it seems you are unable to accept that a relative improvement in performance is an acceptable outcome in terms of establishing prima facie Entreq efficacy - not an absolute one. Hopefully intelligent, un-biased readers of this thread will understand this.

Would you dismiss the efficacy of an ECU upgrade to my humble Toyota Corolla just because as a result of that upgrade, it still only develops 28% the power of a Ferrari as opposed to 25%? Your comments above would only be valid in the case where the existence of the Entreq box showed no relative improvement to measured performance, no improvement in actual sonics and the boxes also show no improvement when deployed in extremely high-end balanced gear such as dCS. Given that the boxes tick all of those attributes, as per usual I will have to dismiss your latest criticisms. Perhaps if users of top of the line balanced dCS equipment saw no improvements when using Entreq products, then we might be able to sincerely examine the possibility that Entreq only works with "significantly flawed" budget systems such as my own.

As for the specific noise issue, I am not really hugely surprised since apart from the system being computer-based, the DAC used for this test was a fairly unpretentious Meridian 7.1 2G card. I could not use the main Xonar ST card (which has better measured noise performance but in all likelihood still far from satisfactory to satisfy yourself) as I had to use that card to do the actual ADC and capture from the outputs. The only other way of doing the tests I have done would be to burn all the files to Blu-ray and use a Blu-ray player, but I do not have the space to use one in here, nor do I have the appropriate cables. It may not even have better performance than the Meridian card anyway. I do not own an expensive power conditioner as I cannot afford one. I just have a basic PS Audio Dectet which feeds into a Corsair AX860 power supply. I also have SOTA brand noise filters installed to each of the fans, the hard drive and the Blu-ray burner.

I regret that you are continually unimpressed by everything I have tried to do which - despite that - in my humble opinion is probably a lot more useful to the audio-person-on-the-street than what certain others have done within this thread and elsewhere on these forums. It is clear as I have said before that nothing will ever satisfy you - not even the measured and sonic results of an Entreq loom used in a $200,000 dCS system. That is why I have put the files out there along with my comments for other people to digest other than yourself. So I guess we are at permanent impasse at this juncture, since I do not own any better quality equipment and I will not do so in the future.

Please remember one thing above all else. I have never claimed that buying my Entreq gear turned my systems into a dCS killer. I have only ever claimed that it improved the performance of my systems and that from a sonic point of view, better than any other upgrades I have made for the price.

 

[BE718]

Sorry Fiddle, this is a precise example of what I referred to earlier in the thread. Your system probably has an issue with its grounding regime. There is no way that mains components should appear at that level even with modest computer based systems. BTW if you want to blame the computer system then it should be self evident that its not adequate to perform recordings that are adequate for analysis / comparison of subtle potential improvements that the entreq might make. Also did you notice that although the mains harmonics reduced when you connected the entreq, the higher frequency noise increased?

You can choose to not believe me if you wish, its of no consequence to me. I demonstrated that I dont need expensive boxes of bent copper and granules to achieve exceptional noise levels. I dont have anything in my system such as power conditioners, filters or the such like. My MDAC is nothing esoteric.

Im not unimpressed with what you are doing, its excellent that you have acquired measurement data thats indicated there is an underlying problem.

If you want to try an experiment try joining you kit together with ordinary bits of wire in the same way as your entreq and perform the measurement again.

Beyond that, look for multiple connections to earth

 

[Purite Audio]

It might also be worth removing the PS Audio and Corsair products and just plugging everything directly into the mains.
Keith.

 

[jkeny]

Sorry Fiddle, this is a precise example of what I referred to earlier in the thread. Your system probably has an issue with its grounding regime. There is no way that mains components should appear at that level even with modest computer based systems. BTW if you want to blame the computer system then it should be self evident that its not adequate to perform recordings that are adequate for analysis / comparison of subtle potential improvements that the entreq might make. Also did you notice that although the mains harmonics reduced when you connected the entreq, the higher frequency noise increased?

You can choose to not believe me if you wish, its of no consequence to me. I demonstrated that I dont need expensive boxes of bent copper and granules to achieve exceptional noise levels. I dont have anything in my system such as power conditioners, filters or the such like. My MDAC is nothing esoteric.

Im not unimpressed with what you are doing, its excellent that you have acquired measurement data thats indicated there is an underlying problem.

If you want to try an experiment try joining you kit together with ordinary bits of wire in the same way as your entreq and perform the measurement again.

Beyond that, look for multiple connections to earth

Good suggestions
I see in previous measurement plots in the Measurements Section that no one, not even Amir, produced such clean & low level noise plots as you did. It might be useful to all, as an example of exemplary multi-device grounding, for you to detail your setup in this regard!

Edit: It would be interesting to know just what percentage of audio systems in the field have less than optimal grounding & just how typical Fiddle's system is? If there's a quick fix to the grounding of these systems then this would also b worthwhile info

 

[amirm]

"Digital" Entreq comparison download link:

https://www.sendspace.com/file/gcquhk

Thanks for all the hard work Fiddle. The tedious aspect of this is the reason I don't constantly do these tests either .

Given the work you put in, I thought it would only be fair that I make an attempt at trying to hear the differences between the files. I spent half hour doing this (had to walk the dogs in the middle ).

Good news. I could easily hear the difference between the entreq and no entreq files when I just played them.
Bad news. I could not remotely do that in blind ABX testing!

Here is one sample results:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2016/02/14 08:50:25

File A: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_entreq.wav
File B: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_no_entreq.wav

08:50:25 : Test started.
08:52:22 : 01/01 50.0%
08:52:30 : 01/02 75.0%
08:52:43 : 02/03 50.0%
08:52:51 : 02/04 68.8%
08:53:03 : 02/05 81.3%
08:53:32 : 02/06 89.1%
08:53:58 : 03/07 77.3%
08:54:12 : 03/08 85.5%
08:54:27 : 03/09 91.0%
08:54:31 : Test finished.

----------
Total: 3/9 (91.0%) [ only 3 out of 9 answers were right with probability of guessing of 91% ]

I would play A and then B and easily "hear a difference." But then when I listened to X and Y which are the the same two but randomized, I could not at all duplicate what I was hearing in A and B. With almost no trying I could read the difference i thought was there in A and B, vote and then get told it was wrong.

So in some sense my results both agree and totally disagree with yours. Sighted I could "easily" hear differences. They were beyond obvious. But the moment the identity was taken away, my ability to identify and hear them would go away.

This is the same device and listening protocol I used to hear differences in high-res vs not, MP3 vs lossless, etc. where I generated positive results. Yet here, I cannot.

If there is a difference here, is far, far less obvious than you think. I like to see others try and see if they can find a difference in blind testing.

BTW I only listened to track 1 and 4.

 

[Barry2013]

Well done Fiddle Faddle.
Great work for which we should all be grateful.
Thanks also to Amir.
I hope others will follow his example.

 

[jkeny]

Thanks for all the hard work Fiddle. The tedious aspect of this is the reason I don't constantly do these tests either .

Given the work you put in, I thought it would only be fair that I make an attempt at trying to hear the differences between the files. I spent half hour doing this (had to walk the dogs in the middle ).

Good news. I could easily hear the difference between the entreq and no entreq files when I just played them.
Bad news. I could not remotely do that in blind ABX testing!

Here is one sample results:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2016/02/14 08:50:25

File A: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_entreq.wav
File B: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_no_entreq.wav

08:50:25 : Test started.
08:52:22 : 01/01 50.0%
08:52:30 : 01/02 75.0%
08:52:43 : 02/03 50.0%
08:52:51 : 02/04 68.8%
08:53:03 : 02/05 81.3%
08:53:32 : 02/06 89.1%
08:53:58 : 03/07 77.3%
08:54:12 : 03/08 85.5%
08:54:27 : 03/09 91.0%
08:54:31 : Test finished.

----------
Total: 3/9 (91.0%) [ only 3 out of 9 answers were right with probability of guessing of 91% ]

I would play A and then B and easily "hear a difference." But then when I listened to X and Y which are the the same two but randomized, I could not at all duplicate what I was hearing in A and B. With almost no trying I could read the difference i thought was there in A and B, vote and then get told it was wrong.

So in some sense my results both agree and totally disagree with yours. Sighted I could "easily" hear differences. They were beyond obvious. But the moment the identity was taken away, my ability to identify and hear them would go away.

This is the same device and listening protocol I used to hear differences in high-res vs not, MP3 vs lossless, etc. where I generated positive results. Yet here, I cannot.

If there is a difference here, is far, far less obvious than you think. I like to see others try and see if they can find a difference in blind testing.

BTW I only listened to track 1 and 4.

Amir, can you explain what specific differences you focussed on in this test Vs what specific differences you used in your previous successful ABX tests?

 

[Fiddle Faddle]

Bad news. I could not remotely do that in blind ABX testing!

I used to be a fan of blind testing but I've changed my mind in recent years as more and more respected people in the audio world have brought it into question. I'm obviously not a brain expert but I know even with myself if I am deliberately switching back and fourth between two things that might sound very close but not identical, quite soon my brain just starts to made one sound like the other and vice versa. It always seems to go down the same way - the first few tests are easy. Then fatigue sets in and the brain starts to shut down. I can emulate that problem with colour perception. I can take two shades of white - one with a green tint an one with a cream tint and they are clearly different colours. But if I start formally trying to switch between the two after a short while the brain gets muddled and often they no longer even appear to be the original colours at all! I sometimes suffer the same problem if I am looking at a web page with a certain background colour. Suddenly my eyes stop working properly and I can't distinguish certain colours afterwards. I even had a professional painter in my house some years ago and he got his brain in a knot over the colour of my wall versus the ceiling. They look like the same colour but they are nothing like it. One is green, one is white. You'd think it wouldn't be hard to distinguish. But that's the brain for you.

What sold me on Entreq was during the audition when the cable fell off the back of one of my components. For a day or so I was wondering what had happened to the sound. It wasn't till I went around to the back of the cabinet and saw the detached cable. And only yesterday again I was wondering why the sound had gone downhill since the day before. Then I discovered I had accidentally turned off the Isotek wall plug right next to the power point where the Gigawatt power strip plugs into. For me, these things are more important than blind testing.

Mind you, with my best headphones and best headphone amp, blind testing is quite easy for me. If anything it is annoying because the slightest differences between things are so obvious. But on my workstation, it only has the headphone amp output from the Xonar card and I use Sennheiser PXC300 headphones as they are great to isolate from the noise from the computer. But they aren't that great for serious blind testing.

The way I test anything new these days is quite different and does not involve blind testing. I will test the new component / tweak in-situ and simply go back once per day, listen to a specific set suite of test tracks, make notes, then come back and do it all again two days later. I will do this over and over again for a few weeks. Then I check the notes for consistency. If I consistently liked the sound over that period compared to before, the component / tweak stays. If not then it doesn't.

 

[amirm]

Amir, can you explain what specific differences you focussed on in this test Vs what specific differences you used in your previous successful ABX tests?

The differences that I thought I heard sighted were that of level overall and amount of high frequency detail/level. So I focused on those but those determinations were faulty and I could not correctly categorize them to pick X or Y as matching A and B. That categorization had helped me in other tests. But there, the categorization was not possible to any kind of reliable level.

 

[amirm]

I used to be a fan of blind testing but I've changed my mind in recent years as more and more respected people in the audio world have brought it into question. I'm obviously not a brain expert but I know even with myself if I am deliberately switching back and fourth between two things that might sound very close but not identical, quite soon my brain just starts to made one sound like the other and vice versa. It always seems to go down the same way - the first few tests are easy.

Not so here. I went wrong right at the start.

Note that when I start the test A and B are unknowns to me. They both sounded "easily" different. Yet I could not match X or Y to either. And when I went back to A and B, I could not replicate the reliable difference that I thought was there.

Since the conditions for evaluation of A and B versus A and X were identical, and my mood just the same, none of the typical arguments about these tests hold water to me. I was reading differences into A and B rather than hearing what is there.

One has to have a way to verify what they are hearing to be there or imagined. For me this is the scheme. Without it, you are taking an exam and grading it yourself. You will never know what you answered wrong.

 

[Barry2013]

My technical knowledge of electronics is limited and hopefully I am not alone.
I am puzzled by BE718's post no 360. The system noise he refers to is common to both files so why should that be an issue in the comparison.
If something remains a constant it is not not normally a factor.
Can BE or anybody else enlighten me?

 

[esldude]

My technical knowledge of electronics is limited and hopefully I am not alone.
I am puzzled by BE718's post no 360. The system noise he refers to is common to both files so why should that be an issue in the comparison.
If something remains a constant it is not not normally a factor.
Can BE or anybody else enlighten me?

Yes you could compare the two results with the noise, but is the noise obscuring the actual difference if any?

Looking at fiddle faddle's graphs his system had 0-20khz noise floor of maybe 80 db or a touch less. That isn't stellar. It is okay, and depending on the particulars you might not notice it. Does the Entreq help or not? Maybe, maybe not, but usually you can get lower noise in a system without needing the Entreq for that.

My guess would be the noise is in the sound card he used to record and probably doesn't represent the noise level of his system. Not meant as a criticism or complaining after all the trouble he went to for our benefit. Just pointing out the facts.

 

[esldude]

Thought I would add to the previous post. Someone might be wondering where I get the 80 db number from if you aren't familiar with FFT graphs.

Let us suppose we measure the total noise, and assume it is white noise evenly spread thru the band we are measuring. 0-24 khz we will assume. Well we measure that and get a number like -80 db. Now if we split the band into 0-12khz and 12 khz-24 khz the total noise in each band drops by 3 db. So each part would read -83 db. Now split it into 4 equal bands and each band drops 3 db more and would read -86 db for each of the 4 bands.

An FFT splits the bands this way. The original graph was a 65,000 bin FFT which would split the 0-24 khz band into 32,500 equal bands or bins (ignore why this is so for the time being). So the graph that looks okay with the levels bouncing around -120 db are such that added together for the whole 0-24 khz band would measure in the vicinity of -80 db. Again this isn't stellar for noise levels in a modern system. Normally you can get down to -100 db while a few systems manage -120 db.

Yes for those who know about FFT I took some shortcuts and liberties to get the main point across. So don't crucify me for that. For someone not familiar with FFTs this hopefully helps understand what the FFT is really showing.

 

[jkeny]

My technical knowledge of electronics is limited and hopefully I am not alone.
I am puzzled by BE718's post no 360. The system noise he refers to is common to both files so why should that be an issue in the comparison.
If something remains a constant it is not not normally a factor.
Can BE or anybody else enlighten me?

I believe what BE is saying is that the spikes seen in the first plot are as a result of mains noise & harmonics - these are the 4 spikes that touch on 105dB amplitude.
These mains spikes disappear in the second plot (with Entreq inline). His conclusion being that Fiddle's system has some ground noise issues which the Entreq is clearing up.
Further to this his suggestion is that this is not a "properly grounded" audio system & in post 362 suggests an experiment attaching ground wires to his audio devices, in the same configuration (presumably a star ground arrangement) as is used by the Entreq, which may well provide the same measureable improvement as is seen in these plots.

As regards the noise floor level, I would agree with esldude - that this is likely due to the sound card used for recording

 

[amirm]

OK, I did some analysis of the starting black segment. First test I did which I am not showing is comparing the captured noise between different trials. This showed some differences so even in digital domain, variations exist as expected.

I then compared the with and without Entreq in track 3. To make sure I have the same analysis applied to both, I compared the segments from 0.5 seconds to 1.5 seconds. Here is the outcome:

As noted, the red graph is without Entreq. Addition of Entreq has substantially reduces harmonics of mains power at 120 and 240 Hz. The picture reverses above 500 Hz however with all the noise peaks being lower without Entreq.

So there is measurable difference. But the cause and whether it is considered an improvement requires further analysis.

 

[amirm]

Thinking about it , the wire in Entreq may very well act like an antenna picking up equipment noise at higher frequencies as radiated by the equipment. If so this explains the worsening of the correlated noises above 500 Hz.

 

[esldude]

Thinking about it , the wire in Entreq may very well act like an antenna picking up equipment noise at higher frequencies as radiated by the equipment. If so this explains the worsening of the correlated noises above 500 Hz.

The first time I saw the teardown pics of one of these that has been my opinion of what would happen. The plates and extra wire would act as antennas. If the litter were the correct substance it might absorb RFI which would possibly be the purpose of the plates. If it isn't an RFI absorbing material or if there is not enough picking up noise seems likely.

So do a proper star ground only and you might reduce the mains pollution without picking up noise above 500 hz.

Even with these differences when you null these files out, you see mostly the mains difference and the above 500 hz noise difference. The profile looks very similar. Playing the nulled files the differences while measurable result in no sound from your speakers at a normal or even elevated level. Amplify the residuals enough (50 db) and you hear low level hum and higher level buzzing noise.

 

[esldude]

I took the silent portion of the track 1. Amplified it digitally by 50 db. You can hear the same mains level and higher level buzz difference between the Entreq and without. I then EQ'd a steep filter at 200 hz. The Entreq almost banishes the mains noise while you hear it though still at a low level in the track without the Entreq. Doing the reverse filtering out everything below 200 hz you hear more buzzing and at a higher frequency with the Entreq vs without.

So Amir's analysis is easily heard as true this way.

I have uploaded a zip file with mp3 of this. Two files took the silence at the beginning and had a steep low pass filter at 200 hz. Two had a steep high pass filter at 500 hz. The differences were also amplified 50 db making it easy to hear.

 

[spazmatron]

Anyone done any testing with a Ben Duncan earth Henry ? I have one, it's a ground cpc noise reducing device. It's super cheap and simple but would be nice to know if it makes a measurable difference.

 

[Fiddle Faddle]

As regards the noise floor level, I would agree with esldude - that this is likely due to the sound card used for recording

I'd bet on it having a lot to do with the very poor quality cable I used. Infact, so bad was the cable between the DAC and the ADC I almost made the decision not to even bother doing the tests at all. I mean it really is an extremely poor cable - the cheap thin type you get with a $100 portable audio player. The best thing you can say about it is that it technically is an electrical conductor. But I do not have any other type of cable that would work in that configuration - I'd have to go and specifically buy a cable just to do this test or to make one from scratch. Not something I want to do to prove a point, nor could I afford it anyway.

Just an hour ago I was re-testing the system with that cheap cable removed and those harmonics completely disappeared. The noise floor of the recording ADC was extremely close to the theoretical 24 bit limit (I still have no idea of how much noise the card I used to record from produces since I do not own the equipment to measure it). Anyway, the cable is picking those harmonics up definitely. It is not a system issue - it is as I suspected a problem when using a $2 cable in a system that normally is cabled with high quality Wireworld cables.

That said, I am not going to repeat the test because the output file differences are still what I hear routinely whether the Entreq is used on my workstation in the "normal" configuration (with high quality cables) or in my loungeroom system which is a completely different CD based system.

Amir's graph is really interesting. That is the best and most useful technical analysis I have seen so far. It would be good if the same types of test could be performed on a top end balanced system, since owners of such systems experience the same benefits that I do.