Head-Fi.org › Forums › Equipment Forums › Sound Science › Method for quantifing the differences in cables?
New Posts  All Forums:Forum Nav:

Method for quantifing the differences in cables?

post #1 of 12
9/4/09 at 10:50am
Thread Starter 
Sorry if this has been discussed before, but I'm very confused. All these discussions about DBT, A/B switch tests - why is it not possible to quantify the difference between 2 pieces of sound equipment?

Method:

1) Play a set of sound waves in known frequencies into an amp.
2) Record the output accurately.
3) Change the connecting cables / power supply.
4) repeat steps 1) and 2)
5) Analyse and compare the two recordings, with each other and with the original audio.

If cables really colour the sound, you'll easily be able to see the differences when you compare outputs. What am I missing?
Reply
post #2 of 12
9/4/09 at 12:34pm
You're missing that the believers will come up with a raft of reasons why your test is invalid unless it confirms their point of view.

If your test shows no difference, then your equipment is bad or not as sensitive as their ears, you did the test wrong, etc. etc.

I agree with your test proposal. That would definitively show whether there was a difference. However, the problem is getting people to accept the test results. Some refuse to accept test results, no matter how carefully done.
Reply
post #3 of 12
9/4/09 at 12:40pm
You're missing that the un-believers will dismiss the differences because "they are not audible".
Reply
post #4 of 12
9/4/09 at 2:00pm
How about

1. insert cable, amp, source, transducer, MUSIC
2. listen
3. evaluate using your processor and hearing devices
4. decide on cable, amp, source, transducer, MUSIC
5. listen

Repeat if necessary
Reply
post #5 of 12
9/4/09 at 2:14pm
Quote:
Originally Posted by echelon View Post
Sorry if this has been discussed before, but I'm very confused. All these discussions about DBT, A/B switch tests - why is it not possible to quantify the difference between 2 pieces of sound equipment?

Method:

1) Play a set of sound waves in known frequencies into an amp.
2) Record the output accurately.
3) Change the connecting cables / power supply.
4) repeat steps 1) and 2)
5) Analyse and compare the two recordings, with each other and with the original audio.

If cables really colour the sound, you'll easily be able to see the differences when you compare outputs. What am I missing?

youre not missing anything. we actually have a member who performed that exact test on i believe 6 different cables, of varying materials and prices - all came out the same, within 0.01db of each other at the worst, iirc

try to search for his thread - it's really interesting
Reply
post #6 of 12
9/4/09 at 2:41pm
Nope - that's a decent way to record differences provided you have low noise equipment. A few people here have done this and elsewhere on the internet. No-one found a significant difference.
Reply
post #7 of 12
9/4/09 at 3:19pm
I don't think the frequency response is the issue nor the magnitude. Nick has confirmed there are no differences in amplitude or frequency response. We need to define the differences in sound another way. How do you measure warmth, metallic edge, etc.?

Better is something that should be stricken from the dialog. That causes most of the disagreement. Different is a better word. Copper braided, solid copper, silver plated copper, stranded silver, solid silver, etc. does have different sound that can be characterized. We need to find the criteria that makes the sound different. Obviously frequency response and amplitude isn't it.
Reply
post #8 of 12
9/4/09 at 3:52pm
You can pump a single frequency in and see if it's altered when it comes out. Make it more comprehensive and do it for an entire music file.
Reply
post #9 of 12
9/4/09 at 4:08pm
Quote:
Originally Posted by Happy Camper View Post
I don't think the frequency response is the issue nor the magnitude. Nick has confirmed there are no differences in amplitude or frequency response. We need to define the differences in sound another way. How do you measure warmth, metallic edge, etc.?
the irony is, frequency response IS what causes those characteristics you just mentioned!

warmth typically refers to sweet or recessed highs, and slightly increased levels of bass/mid-bass

a metallic edge often refers to low levels of harmonics in relation to the fundamental

both of these can only be measured via FR
Reply
post #10 of 12
9/4/09 at 5:08pm
Quote:
Originally Posted by Happy Camper View Post
I don't think the frequency response is the issue nor the magnitude. Nick has confirmed there are no differences in amplitude or frequency response. ........... We need to find the criteria that makes the sound different. Obviously frequency response and amplitude isn't it.
More to the point, what criteria CAN be audibly different between different cables, all cables, even two supposedly identical cables will measure slightly differently, for FR these differences will (Normally) be so small as to be of dubious audibility apart from cables with pathological behavior.

So then what audible criteria can differ sufficiently to be audibly different between different cables and be objectively meausurable as well, hint it isnt RLC, these can be substantially and yet still inaudibly different.

Personally I am yet to be convinced that any "normal" cables will be audibly different under unsighted conditions, given that the thresholds for stimuli differentiation are quite well known and so far (AFAIK) nobody has come up with any normal cables which show such relatively extreme differences in these parameters, viz ms of group delay, multiple 10ths of a db amplitude difference over an octave and so on.

I just remembered there was one successful DBT of analog cables, this test used 1m of one cable and 6m of another and used the uneq'ed and un-level-matched output of a MM cart as the signal, possibly an extreme case but a positive result still.
Reply
post #11 of 12
9/5/09 at 12:09am
Quote:
Originally Posted by El_Doug View Post
the irony is, frequency response IS what causes those characteristics you just mentioned!

warmth typically refers to sweet or recessed highs, and slightly increased levels of bass/mid-bass

a metallic edge often refers to low levels of harmonics in relation to the fundamental

both of these can only be measured via FR
Exactly, the sound we hear is simply the result of small variations in pressure over the ear with time. Assuming the test equipment has sufficient sensitivity to measure this variation to below the sensitivity of the human ear at a sufficient sample rate to resolve frequencies up to the maximum range of human hearing, then any audible difference will show up in either the amplitude or phase of the resulting frequency response spectrum. As far as I know neither of these assumptions is unreasonable for a decent test setup, so if no difference can be measured, then there is no audible difference there in the first place. Besides, if the large differences in sound that are often attributed to cables were real, there would be no problem measuring them even with a very modest test setup.

/Awaits response that says you can't say that what is audible is limited sound alone or something to that effect...
Reply
post #12 of 12
9/5/09 at 1:10am
Better yet is a test that purports to use gradually more expensive and more elaborate cables while using the same gauge and element composition of wire. People will likely notice a difference based on presentation alone. That's more important in determining differences, since the participants will swear to hear differences, they can be thought of as actual psychological phenomena rather than microscopic technical differences.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Method for quantifing the differences in cables?