[stevenkelby]

Quote:

Originally Posted by Febs /img/forum/go_quote.gif You've missed my point. If the assertion is that "cables burn in," and one of the characteristics of a burned-in cable is improved bass, then that can be readily tested. If the test shows that there is a difference in the frequency response after burn-in, then you have established that burn-in makes a difference, and there is no need to get to the question of whether more nebulous concepts like "decay" even exist, let alone whether they can be measured.

I think that we just have an error in communication, my fault, I know I don't make much sense most of the time.

I'll try harder:

1. I did get your point. Wasn't it that "The fact is that the people have described effects of cable burn-in (i.e., "initial lack of bass") that should be easily measurable."

2. I agreed with it. "obviously if certain frequencies are reproduced at different SPL after burn in, that would be measurable"

3. You didn't get my point! Which is that I wouldn't expect a frequency response graph to show any difference after burn in.

4. The question I was asking is: Are nebulous concepts like "decay" audible, but not measurable? Do differences in decay (for example) exist between headphones, and is there a way to measure that difference. Other than with our "Golden Ears"?

Anyone have an answer to that one? I don't know but am curious!

 

[hciman77]

Quote:

Originally Posted by stevenkelby /img/forum/go_quote.gif 4. The question I was asking is: Are nebulous concepts like "decay" audible, but not measurable? Do differences in decay (for example) exist between headphones, and is there a way to measure that difference. Other than with our "Golden Ears"? Anyone have an answer to that one? I don't know but am curious!

If I understand this right Rise/fall time on a square wave could be used as a proxy for transient response/decay i.e there is a finite time to get from 10% to 90% of the wave amplitude and a finite time for the signal to drop from 90% to 10% - this is measurable, in CD players and cartridges this rise time goes from about 7 microseconds for high end cartridges to 22 microseconds (with CD it is deterministic due to the sampling frequency i.e 1/44,000s ).

Feed a square wave into some component and plug in an oscilloscope to the outputs via given cable and measure changes in the rise/fall time over time ?

Headphones are more likely to show such differences as they are mechanical devices like speakers and the diaphragms may behave very differently depending on materials, size, stiffness and what have you. The diaphragms have a finite behavior when responding to signals and returning to rest.

 

[stevenkelby]

Nice one. Surely someone can try this stuff!

How could you run that test with a set of cans? Would a mic recording the output from a new or burnt in pair, then feeding that signal through an oscilloscope possibly show any differences?

 

[artizen65]

This is my first post in these forums so I couldn't resist.

http://www.st-andrews.ac.uk/~jcgl/Sc...ect/page1.html

http://www.st-andrews.ac.uk/~jcgl/Sc...io/Analog.html

For those of you so inclined this is an execelent resource.

 

[hciman77]

Quote:

Originally Posted by artizen65 /img/forum/go_quote.gif This is my first post in these forums so I couldn't resist. http://www.st-andrews.ac.uk/~jcgl/Sc...ect/page1.html http://www.st-andrews.ac.uk/~jcgl/Sc...io/Analog.html For those of you so inclined this is an execelent resource.

So, correct me if I am wrong but the relative power losses for the same length runs of different types of speaker wire material (3m in their example) are so low i.e never more than a difference of 0.025db at 25K that they would fall into the category of unlikely to be noticed in normal use ?

So group delay means that higher frequences actually get there faster than lower frequencies - albeit by only 50ns or so - you learn something new every day

 

[tom b]

My experience is that they need about 100 hours to flow nicely and sound really good

 

[artizen65]

Quote:

Originally Posted by hciman77 /img/forum/go_quote.gif So, correct me if I am wrong but the relative power losses for the same length runs of different types of speaker wire material (3m in their example) are so low i.e never more than a difference of 0.025db at 25K that they would fall into the category of unlikely to be noticed in normal use ? So group delay means that higher frequences actually get there faster than lower frequencies - albeit by only 50ns or so - you learn something new every day

hciman,

Now i think a little story is in order. just because they use a dialectric constant of 1.5 in their example which is reletivly low for a dialectric. I sat down to listen to the system one day and it didn't sound quite right. This went on for about a week. I for the life of me couldn't figure out what was wrong. I finally went behind the system to do some checking. I found that my grandson had turned off the active shielding on my SR cables (IC's and speaker). As soon as I turned it back on it was like wow what a difference.

During my cable upgrade spree I was instantly able to hear a difference for the better in my cables. Call me crazy but IMHO cables do matter. Now wether cable burn-in makes a difference for me the verdict is still out. Only because I have never compaired a burned in cable to a non burned cable. I agree that our audible memory is not good enough to remember what something sounded like 10 days before compared to what we are listening to now.

Now I have also put cables in my system where there was no audible difference between one and the other now these have been lower end cables AR, Monster, lower end Audio Magic.

As far as group delay at the speed of light 50ns is huge. That is why some speaker manufactures time align their their drivers. Others phase align their speakers. Other manufactures don't care. I had a chance to have a discussion with a speaker designer last year at RMAF and I was complaining that most of the rooms were on the bright side. He in no uncertain terms said it was because the highs were arriving before the rest of the signal.

Ok lets surmise that you are listening at 5 watts with the initial pluck of the guitar string you might go to 50 watts even with a minimal .025 db loss it could affect the overall audible dynamics of the music.

Don't get me wrong I think the whiter paper I posted is an execelent resourse.

 

[stevenkelby]

Great info in those links too, thanks!

 

[65535]

20 amps out to BURN them in sufficiently. Voltage is your choice, you can usually get about 20 amp surge before a 115v breaker trips.

(Other than corrosion thats the only thing that changes in wires physically, placebo on the other hand...)

 

[hciman77]

Quote:

Originally Posted by artizen65 /img/forum/go_quote.gif Ok lets surmise that you are listening at 5 watts with the initial pluck of the guitar string you might go to 50 watts even with a minimal .025 db loss it could affect the overall audible dynamics of the music.

If the difference between two cables of the same gauge is 0.025db at listening levels it is way below what is known to be detectable in level discrimination tests which is 0.25db (Toole and Olive 1988) and that only in pink noise studies in lab conditions i.e low ambient noise, with music discrimination is no better than 0.5db (Jestaedt, et. al. 1977) (at 80db levels) at 5db it goes up to 1.5db.

Of course if you make the cable very thin and/or very long you can get big time signal loss, but why would anyone rational do that ?

Quote:

As far as group delay at the speed of light 50ns is huge. That is why some speaker manufactures time align their their drivers. Others phase align their speakers. Other manufactures don't care. I had a chance to have a discussion with a speaker designer last year at RMAF and I was complaining that most of the rooms were on the bright side. He in no uncertain terms said it was because the highs were arriving before the rest of the signal

Sadly our perceptual abilities are really not that good. Group delay in audio is not detectable below 1ms at any frequency (Blauert and Laws 1978) , 50ns is way below that, so that isnt really an issue.

 

[bigshot]

Quote:

Originally Posted by artizen65 /img/forum/go_quote.gif I had a chance to have a discussion with a speaker designer last year at RMAF and I was complaining that most of the rooms were on the bright side. He in no uncertain terms said it was because the highs were arriving before the rest of the signal.

That is completely absurd. What company does he design speakers for so I can avoid them!

Quote:

Originally Posted by artizen65 /img/forum/go_quote.gif Ok lets surmise that you are listening at 5 watts with the initial pluck of the guitar string you might go to 50 watts even with a minimal .025 db loss it could affect the overall audible dynamics of the music.

Do you have any idea what a .025 dB's difference sounds like? There is absolutely no way you would even be able to detect a difference with direct A/B comparison and the best equipment in the world. You probably would have a great deal of trouble telling the difference between 2-3 dBs, which is several orders of magnitude higher. A delay of 50 nanoseconds in a particular frequency would also be completely inaudible. Not even close. A little definition of terms is in order... a nanosecond is 1 billionth of a second. You might just be able to hear a couple thousandths of a second delay. It would take more than that to start messing with the music itself.

Go to your computer and use an audio editing program to start adjusting a track. Boost a frequency by 1dB and see if you can hear it. Add a delay of a microsecond or two using a digital reverb. See what it really sounds like. Ramp it up until you actually *can* hear a difference and see how those numbers compare to specs. It's important to understand what these numbers mean in audible terms. When salesmen throw bogus numbers at you to try to convince you to buy their product, you have to be able to sort the bull pucky out.

See ya
Steve

 

[artizen65]

Don't have much time but we are talking db in an electrical signal which is way different than db in sound pressure levels.

 

[hciman77]

Quote:

Originally Posted by artizen65 /img/forum/go_quote.gif Don't have much time but we are talking db in an electrical signal which is way different than db in sound pressure levels.

Okay, in electrical terms to get an audible difference in FR for CD players you are talking about one dropping at least 1 - 2 db off the 16 - 20K range. A 0.025db roll off is just not going to be detectable. CD players are considered flat if they hit 20 - 20K at +/- 0.5db.

 

[bigshot]

As I understand it, the Bell Labs created the decibel as a tenth of a bell- a bell being the amount of signal loss caused by transmitting a signal through one mile of telephone cable. You would definitely be able to hear a bell's difference- a mile, and just barely hear 2db- about a quarter mile of cable. How much loss do you think there is in your six inches of audio cable?

Math was never my strong suit, but even I can figure out this word problem!

See ya
Steve