Fancy Interconnects? How about a potato, or even mud?

[Pano]

I did some testing this afternoon thru 6 feet of tap water, salt water, diet Coke, wine and coffee. The beer - I drank.

Conclusions:

• Tap water: Very bad, almost no signal
• Very salty water: OK. About 6dB attenuation and some noise.
• Diet Coke: Had to boil it to release the carbonation, otherwise it was just fizz. Terrible conductor, heavy attenuation.
• Wine: A little better than Coke, but not much.
• Coffee (black): Similar to wine.

Don't try to make liquid cables, unless it's mercury. The things I tried make awful conductors. Maybe over a few inches, like the banana, they would be OK - but not over 6 feet.

I'll get some files up pronto.

 

[fas42]

Just tried an interesting exercise of more vigorously comparing the FLAC versions of the track, which I hadn't done as yet now knowing the results -- which for my setup points out there are 3 factors at play: the variations in the recorded tracks, the impact of mp3 encoding, and the relatively cheap and nasty quality of my playback hardware. These are very roughly of similar levels of impact in degrading the sound, making it a more obtuse exercise to pick rankings.

By only playing FLAC, the mp3 bits are discarded; and now the easily discernable last place is D, the mud version - this has nothing to do with the glitch, the cymbals are clearly rendered more poorly. But the others are harder, and this is due to junk connections and components of the PC speakers; reseating the plug, and refreshing the volume pot setting makes as much difference as switching between versions. To get an authorative assessment would require taking the quality of the playback mechanism largely out of the equation, by transferring all tracks to a significantly higher quality system ..

This is where blind testing can get into trouble -- are you hearing the real underlying differences, or are the inadequacies of the testing environment masking these qualities ...?

 

[Pano]

Once you know what is what, it's not a blind test. You don't like D because you know what it is. I'll randomize those files again in the FLAC version and let you listen.

But don't worry, you'll have no trouble picking out the awful wine and coffee recordings.

 

[Pano]

The goofy set up

Here are photos of my goofy test rig - just so that you can see what I'm doing. It's not pretty, but science rarely is.

The test leads are some unknown mic cable with 2 conductors and shield. The two conductors are tied together into 1, the shield is connected only at the sound card.
Sound card is the M-Audio Fast-Track Pro USB running at 44.1KHz, 16 bit. Player software was Foobar 2000, recording done with Audacity. Editing done with Goldwave. Computer is an HP netbook with Atom processor, Windows 7.

 

[fas42]

I don't get fussed about the blindness, what disturbs me is that the quality literally alters as you listen, call it 'memory distortion' of the system if you want - I've been doing this type of critical listening for decades. If what you're listening to is of sufficiently high quality then this effect is relatively low level, you don't notice it happening - depending upon everything, the quality change can go up or down over a period of time - speaker drivers typically steadily improve as they warm up, say.

Knowing how good something is supposed to be doesn't do for me, the amount of crap sound I've listened to from 'good gear' over the years is a total turnoff ...

 

[sakellogg]

I wonder if the coil of fluid has diffrent inductance? Not that it matters for your test just got my weels turning.

 

[SY]

Keep dancing, Frank.

 

[fas42]

Trollin', trollin', trollin', keep them ...

 

[fas42]

I feel I've got a decent handle on separating the FLAC versions now: the DAC is reasonably well conditioned after a couple of hours running, and I needed a section of the track where the quality issues of the playback can be overridden. So, using Audacity on repeat for the section from about 17.5 to 20.5 secs makes it quite straightforward, it's a good conjunction of a peak of the voice, drums and piano, where one can compare the clarity of each element in the presence of the others.

 

[Pano]

New clip order

OK, here are the same files again, but the mud is gone replaced by steel wool.

EDIT: I had some glitches. Will fix and upload again. Sorry.

 

[fas42]

Gremlins are working overtime, Pano - major glitch in track 1 ... correct it before too many people try?

 

[Pano]

Thanks, I'll check it.
EDIT: Yeah, looks like I got some glitches. I'll pull then and double check.

 

[Pano]

Liquid Loops

In the meantime, here are the files I recorded today.
We have:

• Original
• Copper Loop
• Steel Wool
• Wine
• Mud

(The order may not be as listed above)
https://www.dropbox.com/s/9jilbq1u712yoqm/Clip_A.mp3
https://www.dropbox.com/s/v67ye2zcdk2nlxg/Clip_B.mp3
https://www.dropbox.com/s/8knhjn7mj4ifszs/Clip_C.mp3
https://www.dropbox.com/s/2635woqlqk2e7u5/Clip_E.mp3
https://www.dropbox.com/s/1637i5wem9r2w4u/Clip_F.mp3

EDIT:
List reduced to 5 materials. Should be easier now.

 

[Mooly]

Thanks Pano. Downloaded and ready to be put on a flash drive and plugged into the Pearl-Lite. I'll give 'em a good listen later today.

Edit... great selection "Mit Lebhaftigkeit und durchaus mit Empfindung und Ausdruck" Who is the pianist on the first track ? Really slow tempo for that piece.

 

[Pano]

Thanks Mooly. I don't know who the pianist is, the sample is from an EBU test disc. It includes many different instruments playing single notes, arpeggio and a brief tune. It seemed like a good, clean recording of piano - which is why I chose it.

Here are the FLAC clips (about 10 meg each)
https://www.dropbox.com/s/siv7xywl0eh65f0/Clip_A.flac
https://www.dropbox.com/s/cb7cwh8ec9rt34i/Clip_B.flac
https://www.dropbox.com/s/csif7aj8blqowlh/Clip_C.flac
https://www.dropbox.com/s/9pmi5df1ii49cv6/Clip_E.flac
https://www.dropbox.com/s/zwvneuzn1dipwjg/Clip_F.flac

EDIT: Now only 5 files.

 

[marce]

What amazes me, is how the GEB can always accurately confirm the best and worst recordings, cables, amps etc. when presented with the results or being able to see the set-up...And there was I thinking that their abilities were a little over exaggerated.
For myself I could barely listen to the tracks due to the extreme digital artefacts, present in all digital recordings and playback systems...
AND I instantly could tell which recording was which though it did take a prolonged listening session, the revelation of which was which came to me on the 30th May at precisely 11:04pm...

 

[Mooly]

Lovely piano sound, wish I could play like that.

OK, here we go again, MP3 files listened to. No PC's, no software...

There was one clear winner, and one that was close but with a fractionally higher audible noise floor... do want me to say which ?

Three tracks had quite noticeable "distortion" with one of them having something reminiscent of how I remember AM and SW receivers sounding after dark, the noise frazzling away in the background. The other three were quite good in isolation but they had a higher noise floor and were less involving.

 

[Pano]

Yes, one was recorded very low, and the post production boost didn't help it. I had to bring it up over 30dB. All the liquids needed noise removal and I did that in the digital domain - very effect actually. They were humming and buzzing like mad. It's amazing how much mains noise is just floating thru the air of an American suburban neighborhood.

Hold off on your results as to which you liked the best. We don't want to influence anyone else.

 

[pinkmouse]

I wonder if we could Jason to rig up a "hidden" poll, so we can vote, but don't get to see the answers 'till the reveal in a week or so?

 

[Pano]

That would be fun. There are some very bad ones in there, they should be easy to pick out.

FWIW, Mooly blind picked the original file as best, the mud as second best.