Yes, the diet coke recording was very noisy and low in level. What you heard is my attempt to clean it up. The other liquid recordings were cleaner, but not by much. To my ear it was the combination of boost and noise removal that caused the swishing, modulated noise. The bigger the boost, the worse the modulation.
This is where I find music recorded in complex ways makes the 'hearing' of differences easier ... say, a new wave music track with 10 musical elements, all in their own acoustic environment. Better reproduction separates each element, the mind has no trouble 'seeing' the contribution of that item as existing independently of the others; the differences correlate with the ease with which the mind can do this.
Michael Jackson's 'Bad' is a good example of this; a really severe test is the soundtrack of 'Moulin Rouge', which also combines great dynamic swings ...
It could be a combination of several minor differences that do not lend themselves to easy identification in a double blind environment. I've heard digital recordings that, in specific areas of frequency or imaging or dynamics, sounded anywhere from pretty bad to pretty good, but some of these areas would counterbalance the perception of others and would seem likely to make it harder to separate them in a short order blind test.
Digital 'problems' are typically hard to identify - they don't obediently sit up, clearly outlined, and remain there, unchanging, so that you can circle around and around them, getting some nice measurements and working out precisely what's causing them. Usually picked subjectively by the sound being 'right' or not, over some period of time -- does the listening just feel like performing a dreary exercise after a while; or are you trapped into wanting to just keep listening, you can't tear yourself away from the music?
So that's the "noise" you're talking about, mains-related signals pulled out of the air by the "conductor."
My first thought is that it might be better to make up the gain in analog before going to the D/A, though some might argue they're then listening to the opamp you use.
The other is that the digital noise reduction (the FFT-based functions audio editors use) makes artifacts quite similar to mp3 encoding, as it works much the same way (at low bitrate and low quality it's the obvious swishing and underwater sounds), and this would be the things many people hear in addition to or instead of the distortion of the substance under test.
Perhaps the "loopback" recording could be made with a few feet of unshielded wire to get some of the same noise pickup that the others get. Or you could do these tests in an animal cage with grounded chicken wire, or a similar Faraday-cage type testing jig.
How complicated this becomes! We never would have known what it takes to do these tests well if you hadn't done this!
I've listened with wine (actually it was much more often beer or rum) or (in more recent decades) coffee going through my veins and arteries, does that count?I don't really expect anyone to be able to identify which liquid is which. How many of us have experience listening to wine or coffee wires?
Did you boost first and then do noise removal? I'd think that would give best results, but it might be interesting to try both orders.
Yes, I was amazed by all the noise they picked up. There were not near the house, but there must be a lot of nose in the air.
I've thought about that, even using transformers to step it way up. I did boost the output level of the sound card on most liquids to get a decent input level, but for some reason didn't on the diet coke.
I'll just make pipes from rigid PVC with aluminum foil as a shield. Should work OK.
Yes it does, Ben. Yes it does.
IIRC, yes. Boost, then remove the noise. Still it would be better not to have the noise in the first place. Next time.
Only just come back to this thread, great to see I'd got it right
... though I was listening on a laptop with headphones
This pretty much re-enforces the beliefs I've developed (or stolen from other people) that a signal is a signal and will pass through any decent conductor well enough that we'd never notice the difference... or in more plain english, an interconnect is an interconnect.
I do wonder where publications like what hi-fi get off with recommending £80+ digital cables... when it comes to digital signals, I refuse to believe they can make any difference.
What would be more interesting would be if we could figure a way of comparing speaker cables, surprised nobody has suggested that yet...
(talking of which, has anybody here tried using network cables as speaker cables?)
... though I was listening on a laptop with headphones
This pretty much re-enforces the beliefs I've developed (or stolen from other people) that a signal is a signal and will pass through any decent conductor well enough that we'd never notice the difference... or in more plain english, an interconnect is an interconnect.
I do wonder where publications like what hi-fi get off with recommending £80+ digital cables... when it comes to digital signals, I refuse to believe they can make any difference.
What would be more interesting would be if we could figure a way of comparing speaker cables, surprised nobody has suggested that yet...
(talking of which, has anybody here tried using network cables as speaker cables?)
Tried here... caused instability post #2747. The cause of the instability. Using cat5 as speaker cable.
http://www.diyaudio.com/forums/solid-state/3075-jlh-10-watt-class-amplifier-138.html#post3517466
http://www.diyaudio.com/forums/solid-state/3075-jlh-10-watt-class-amplifier-138.html#post3517466
Interesting post, is there any discussion on the performance of the cat5 as speaker cable? from a physics point of view I can see the logic in going for thinner cables over thicker.
It's a pity that thread died - as you rightly pointed out, the amp shouldn't have started oscillating from the cables, sounds to me like something was very wrong at the O.S.
Ditto what SY just said. There have been a number of posts in favor of CAT5 for speakers over the years - single or multi-stand.
Well it looks like this test has fizzled out. I've gotten only 4 people to guess which material is which - three of whom correctly picked the original file. Been deadly silent for days. Are the rest of you cowards, deaf, or just not interested?
Well it looks like this test has fizzled out. I've gotten only 4 people to guess which material is which - three of whom correctly picked the original file. Been deadly silent for days. Are the rest of you cowards, deaf, or just not interested?
I've just done a little bit of DiffMaker type of investigation, and there's a couple of things of interest which may or may not help people: in the DA/AD loop there's a little bit of phase shift at the low end; and, the right channel has more distortion than the left in the midrange, not dramatically so, but definitely visible ...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.