|
Post by johneppstein on Nov 18, 2018 13:34:05 GMT -6
I think 99.9% of that kind of stuff is pure, unadulterated illusion. USE YOUR ILLUSION! Illusion is our business.
Seriously.
|
|
|
Post by ragan on Nov 18, 2018 13:34:06 GMT -6
If he's a fraudulent dude who fakes stuff, then f*** him. Doesn't mean there's anything wrong with null theory (there isn't). Though as svart points out, there's a lot of systematic interplay that muddies the waters. And though it's still not actually the case that it's the cables imparting a 'sound', it could sometimes be the case that the resultant 'sound' is in fact different. Audibly or not is an entirely different question. I think expectation bias is probably an exponentially larger factor in what makes people believe in stuff like Cable Sonics or Extra Fancy Three Foot IEC Cables and whatnot. I think 99.9% of that kind of stuff is pure, unadulterated illusion. I think there might be a possibility that in a large studio situation where you might have literally miles of wire running between different rooms and pieces of gear there might actually be an audible difference between some high tech cables and standard quality, especially if you get into using stuff like pure silver conductors. For short runs, probably not so much. But I really would like to see a null test between a mile of expensive silver cable and a mile of plain copper. The test would need to be conducted using a full range music signal.
In general though it's not something I worry about. I don't have anywhere near the budget for it to be a real issue.
Also, systematic idiosyncrasies of the connected components aside, the qualities of the cables are easy to compare whether it's a foot or 1000 miles. It's the same figures for resistance, capacitance and inductance either way, just add some zeros. I mean, providing their context is the same. Traveling thousands of miles would, practically speaking, involve temperature, humidity and a bunch of other things changing. And maybe there's stuff I just don't know yet. Well, certainly there is, lots. But maybe there's stuff that directly pertains to this question of conductor types in studios. And now, no joke, I have to get back to studying for tomorrow's physics midterm where I'm, quite literally, calculating resistance, capacitance and inductance for various scenarios...
|
|
|
Post by johneppstein on Nov 18, 2018 13:49:16 GMT -6
I think there might be a possibility that in a large studio situation where you might have literally miles of wire running between different rooms and pieces of gear there might actually be an audible difference between some high tech cables and standard quality, especially if you get into using stuff like pure silver conductors. For short runs, probably not so much. But I really would like to see a null test between a mile of expensive silver cable and a mile of plain copper. The test would need to be conducted using a full range music signal.
In general though it's not something I worry about. I don't have anywhere near the budget for it to be a real issue.
Also, systematic idiosyncrasies of the connected components aside, the qualities of the cables are easy to compare whether it's a foot or 1000 miles. It's the same figures for resistance, capacitance and inductance either way. No.
Discrepancies are cumulative. So are things like noise pickup, harmonic distortion, etc.. Something that may be beyond our range of measurement is a small (short) sample may become significant over a greater length. And given that the ear is more sensitive in certain aspects than conventional test gear who's to say that a measurement that shows up with a very long sample might not have some slight audible effect in a short one? One might believe that it shouldn't, but that's making an assumption. Assumptions, of either a positive or negative type, are not scientific. A truly scientific conclusion would be "We don't know."
As an aside, one might assume that harmonic distortion should not be a factor in wire. However it is known that, e.g. carbon resistors are capable of inducing measurable distortion in some circuits. It's a very short jump from the composition of a resistor to the composition of wire - a wire is, after all, a type of resistor. I'm not saying that wire does induce significant distortion in a signal - but because we can't prove that it does (or haven't yet) doesn't mean that it does not.
It is wise to remember that all our science is actually just an approximation - which is why things are steadily updated over time; the approximation gets better and but it's still an approximation. And because it's an approximation it's possible to prove what is but it is not possible to prove what isn't unless it's a case of two possibilities that are clearly mutually exclusive.
|
|
|
Post by ragan on Nov 18, 2018 14:10:48 GMT -6
Also, systematic idiosyncrasies of the connected components aside, the qualities of the cables are easy to compare whether it's a foot or 1000 miles. It's the same figures for resistance, capacitance and inductance either way. No.
Discrepancies are cumulative. So are things like noise pickup, harmonic distortion, etc.. I don't follow. If there's a relationship (ie, such and such piece of gear exhibits such and such when presented with such and such parameters), why wouldn't you be able to factor that in? These "discrepancies" are introduced by something - conductor behavior in the stuff we're talking about. There will be a relationship there. Doesn't need to be linear of course, could be any curve you can imagine, but I don't follow how it would be mysterious and incalculable. Or at least I don't see why it would necessarily be mysterious in incalculable.
|
|
|
Post by Martin John Butler on Nov 18, 2018 15:55:33 GMT -6
Over 30 years ago my "expectation bias" was in line with Ragan's thinking. I was adamantly convinced there was no way in hell changing to more expensive speaker wire would improve the sound I got on my stereo system. I used standard Monster Cable speaker wire then, and on another system, lamp cord. When I was getting something at a Hi-Fi shop, the owner suggested I upgrade my speaker cables. He recommended some basic, but slightly higher priced speaker wire by Tributaries. It cost a little more than twice what the Monster wire cost. I thought he was crazy, really. He said, try this, and if you don't like it, you can bring it back, no worries, no questions asked. So I thought it was worth trying, if only out of curiosity.
So, I put on all sorts of music, one particular Tom Waits song I liked had some unintelligible lyrics. Instantly, I could understand the words I couldn't understand previously. I was very familiar with that song, yet I could now hear clearly words that I figured were typical Tom Waits mumbling sounds before. Then there was a certain part in another track I thought were real strings. It was now painfully obvious they were not real, but from a keyboard. This went on and on with different styles and genres.
So, if you think there can't be any difference for scientific reasons, that's fine with me, just don't tell me I'm imagining things, because I'm not. There are people who've proven they can discern the difference in blind tests, so it's not that there are no differences, it's that not everyone notices them or cares to.
|
|
|
Post by notneeson on Nov 18, 2018 16:52:19 GMT -6
Martin, there’s no shortage of top shelf mastering guys who are into fancy cables. You’re not alone in trusting your experience over theory, or the general audio bb shit show of uninformed opinions for that matter.
But, the day you realize the eq you just labored to dial in is actually in bypass, is the day you realize how easy it is to fool yourself in audio. And it’s a very common experience among AEs.
Trust your gut, but question your assumptions. That’s my motto, anyway.
|
|
|
Post by Martin John Butler on Nov 18, 2018 16:55:48 GMT -6
Not a bad policy notneeson :-)
|
|
|
Post by drbill on Nov 18, 2018 16:59:54 GMT -6
I'll take it one step further.....
If you BELIEVE that one piece of gear over another is better - or you believe that one cable over another is better -- or you are sure that one interface is superior over another -- and if it inspires you to make better music and / or be more creative.....that's enough.
No null tests needed.
Only belief. And you will make better music - even with equal or even - gasp - inferior gear.
We work and create in the etherial, not in a lab. At least I do. I know some prefer to work in a lab over a creative lair where catching fairy dust is paramount, and that's 100% OK with me. I'll leave the null tests to them, while I grab my belief's and run with them. Takes both kinds to make the world go round.
|
|
|
Post by notneeson on Nov 18, 2018 17:22:17 GMT -6
I'll take it one step further..... If you BELIEVE that one piece of gear over another is better - or you believe that one cable over another is better -- or you are sure that one interface is superior over another -- and if it inspires you to make better music and / or be more creative.....that's enough. No null tests needed. Only belief. And you will make better music - even with equal or even - gasp - inferior gear. We work and create in the etherial, not in a lab. At least I do. I know some prefer to work in a lab over a creative lair where catching fairy dust is paramount, and that's 100% OK with me. I'll leave the null tests to them, while I grab my belief's and run with them. Takes both kinds to make the world go round. And when you’re paid to record other people’s music, maintaining that belief (in the gear, the studio, your skills, the work itself) is the most important and often most difficult part of the job.
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Nov 18, 2018 17:35:57 GMT -6
I think there might be a possibility that in a large studio situation where you might have literally miles of wire running between different rooms and pieces of gear there might actually be an audible difference between some high tech cables and standard quality, especially if you get into using stuff like pure silver conductors. For short runs, probably not so much. But I really would like to see a null test between a mile of expensive silver cable and a mile of plain copper. The test would need to be conducted using a full range music signal.
In general though it's not something I worry about. I don't have anywhere near the budget for it to be a real issue.
Also, systematic idiosyncrasies of the connected components aside, the qualities of the cables are easy to compare whether it's a foot or 1000 miles. It's the same figures for resistance, capacitance and inductance either way, just add some zeros. I mean, providing their context is the same. Traveling thousands of miles would, practically speaking, involve temperature, humidity and a bunch of other things changing. And maybe there's stuff I just don't know yet. Well, certainly there is, lots. But maybe there's stuff that directly pertains to this question of conductor types in studios. And now, no joke, I have to get back to studying for tomorrow's physics midterm where I'm, quite literally, calculating resistance, capacitance and inductance for various scenarios... I would not make the assumption that all are a linear function, I would assume at least some would be an algorithmic function.
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Nov 18, 2018 17:40:47 GMT -6
There are some cables expensive cables designed based on science and imperical data, using audio test equipment by HP, Tecktronis AP ect. Part of the problem with the implementation of most null testing is the use of audio interfaces rather than calibrated test gear. On the other hand some expensive cables are pure snake oil dipped generic copper in pretty packaging.
|
|
|
Post by drbill on Nov 18, 2018 17:50:53 GMT -6
I'll take it one step further..... If you BELIEVE that one piece of gear over another is better - or you believe that one cable over another is better -- or you are sure that one interface is superior over another -- and if it inspires you to make better music and / or be more creative.....that's enough. No null tests needed. Only belief. And you will make better music - even with equal or even - gasp - inferior gear. We work and create in the etherial, not in a lab. At least I do. I know some prefer to work in a lab over a creative lair where catching fairy dust is paramount, and that's 100% OK with me. I'll leave the null tests to them, while I grab my belief's and run with them. Takes both kinds to make the world go round. And when you’re paid to record other people’s music, maintaining that belief (in the gear, the studio, your skills, the work itself) is the most important and often most difficult part of the job. Yeah, it certainly can be. Experience, a diverse variety of gigs under your belt, and plenty of personal experimentation on your own time are the friends and armor that will get you out alive.... Clients trust us for our experience and expertise. If I told them "I'm not sure, let's do a null test and see if they are the same"....haha....well, that would probably be my last gig with them. If you commit and believe, it fuels the session, you make progress, and the chance of magic can happen. Your clients trust in you increases, and you make more money and music while your reputation grows. In my experience, that never happens during scientific discovery. Which CAN be useful on your own time, but never with clients around.
|
|
|
Post by drbill on Nov 18, 2018 17:58:46 GMT -6
Part of the problem with the implementation of most null testing is the use of audio interfaces rather than calibrated test gear. I say that's most if not all of the problem. Martin's anecdotal evidence may or may not be audible / discernible on system 1, while being completely audible / discernible on system 2. We are often measuring with coarse tools that are not designed for the job we set before them.
|
|
|
Post by wiz on Nov 18, 2018 18:19:51 GMT -6
There are some cables expensive cables designed based on science and imperical data, using audio test equipment by HP, Tecktronis AP ect. Part of the problem with the implementation of most null testing is the use of audio interfaces rather than calibrated test gear. On the other hand some expensive cables are pure snake oil dipped generic copper in pretty packaging. I am actually a qualified Metrologist. Its interesting the discussion is mostly personal, which is a little sad...comments about weight etc Thats just mean. The part I found most interesting in the video, is when he moved the cables apart , noise was picked up by one, and then when moving the cables back together, the noise was gone. That to me is one of the biggest telling features, and the place where you can learn a shit load. Anyone want to do a better test, rather than just criticise something that someone has put a lot of work into and put up for free and is offering a perspective... We all know how we have been fooled in our audio jobs, happens to me almost daily. I found the video interesting, I have learned from Ethan.. he did design and build his own audio console in the early 70s. He may have questionable ethics, dont know, wasnt part of it.. I take people at their word... cheers Wiz (who is also fatter than he was 10 years ago, and I have a little bald spot now, and grey pubes.)
|
|
|
Post by ragan on Nov 18, 2018 19:06:30 GMT -6
Also, systematic idiosyncrasies of the connected components aside, the qualities of the cables are easy to compare whether it's a foot or 1000 miles. It's the same figures for resistance, capacitance and inductance either way, just add some zeros. I mean, providing their context is the same. Traveling thousands of miles would, practically speaking, involve temperature, humidity and a bunch of other things changing. And maybe there's stuff I just don't know yet. Well, certainly there is, lots. But maybe there's stuff that directly pertains to this question of conductor types in studios. And now, no joke, I have to get back to studying for tomorrow's physics midterm where I'm, quite literally, calculating resistance, capacitance and inductance for various scenarios... I would not make the assumption that all are a linear function, I would assume at least some would be an algorithmic function. Right, the relationship could follow any curve or distribution I maginable. Never meant anything had to be linear.
|
|
|
Post by ragan on Nov 18, 2018 19:09:32 GMT -6
I'll take it one step further..... If you BELIEVE that one piece of gear over another is better - or you believe that one cable over another is better -- or you are sure that one interface is superior over another -- and if it inspires you to make better music and / or be more creative.....that's enough. No null tests needed. Only belief. And you will make better music - even with equal or even - gasp - inferior gear. We work and create in the etherial, not in a lab. At least I do. I know some prefer to work in a lab over a creative lair where catching fairy dust is paramount, and that's 100% OK with me. I'll leave the null tests to them, while I grab my belief's and run with them. Takes both kinds to make the world go round. Good insight for sure. For my part, I kinda approach it like this. There are *so many* things to fuss about in audio engineering (or anything else). And for many of them, we just can’t get at a reliable, objective answer. In the cases where we can get an objective answer?? Hoo-ray! I want it! Then I can check it off the list of concerns and unknowns and get back to the fairy dust, where the real fun is.
|
|
|
Post by Martin John Butler on Nov 18, 2018 19:31:15 GMT -6
Part of the problem with the implementation of most null testing is the use of audio interfaces rather than calibrated test gear. I say that's most if not all of the problem. Martin's anecdotal evidence may or may not be audible / discernible on system 1, while being completely audible / discernible on system 2. We are often measuring with coarse tools that are not designed for the job we set before them. Very true, thanks drbill, Eric. There are so many variables when it comes to audio quality, I mostly prefer the "spend some time with it approach". First impressions usually matter the most, but sometimes it takes a while for you to focus on something that you eventually hear easily. This happened to me while doing my new album, which took a few years. My ears learned to notice things that previously bugged me but I didn't know what it was exactly, now it's quite apparent immediately.
|
|
|
Post by johneppstein on Nov 18, 2018 20:12:37 GMT -6
No.
Discrepancies are cumulative. So are things like noise pickup, harmonic distortion, etc.. I don't follow. If there's a relationship (ie, such and such piece of gear exhibits such and such when presented with such and such parameters), why wouldn't you be able to factor that in? These "discrepancies" are introduced by something - conductor behavior in the stuff we're talking about. There will be a relationship there. Doesn't need to be linear of course, could be any curve you can imagine, but I don't follow how it would be mysterious and incalculable. Or at least I don't see why it would necessarily be mysterious in incalculable. Seriously?
All test gear has limiting parameters.
If a small sample of the substance being tested has an anomaly - deviance from ideal spec, cumulative distortion, cumulative error, whatever that falls outside the limits of you test gear it won't register on the output - but that doesn't mean it isn't there and it doesn't mean that it might not be detectable if a large enough quantity of the substance/ device/ whatever being tested were connected in series.
Connection in series = discrepancies are additive. Connections in parallel = discrepancies are averaged.
When discrepancies are additive they might or might not be obvious when measured in series. When measured in parallel it's less likely.
Therefore when a mile of X cable is tested for effects on the signal it far more likely that you'll detect an error than when measuring a smaller sample, OR EVEN when measuring a larger sample in parallel, which will tend to mask discrepancies.
Why wouldn't you be able to factor it in? Because NO test equipment is even remotely perfect and maybe what you're looking for is beyonmd the ability of your test gear to measure.
Sewriously, it's pretty obvious. If your distortion analyzer is accurate to 1% of THD it won't detect something with .5% THD - but is you series two of the devices under test together your analyzer might then detect that the two seriesed devices have a cumulative distortion of 1%, which is detectible to your (imperrfect) gear. And ALL test gear is imperfect.
The question of whether your ear is more sensitive to a given effect than your test gear is another question entirely. Ethan would say no. But then Ethan's ear apperas to be about as sensitive as two tin cans and a piece of string.
It's really pretty obvious.
|
|
|
Post by johneppstein on Nov 18, 2018 20:17:33 GMT -6
Martin, there’s no shortage of top shelf mastering guys who are into fancy cables. You’re not alone in trusting your experience over theory, or the general audio bb shit show of uninformed opinions for that matter. But, the day you realize the eq you just labored to dial in is actually in bypass, is the day you realize how easy it is to fool yourself in audio. And it’s a very common experience among AEs. Trust your gut, but question your assumptions. That’s my motto, anyway. Not all people are idiots.
|
|
|
Post by johneppstein on Nov 18, 2018 20:29:01 GMT -6
I found the video interesting, I have learned from Ethan.. he did design and build his own audio console in the early 70s. I have so far been unable to verify that.
.You don't know the half of it. I actually had to retain a lawyer to get him to desist from outright slander. And unauthorised reproduction of copyrighted material given him in coinfidence.
|
|
|
Post by ragan on Nov 18, 2018 20:55:36 GMT -6
I don't follow. If there's a relationship (ie, such and such piece of gear exhibits such and such when presented with such and such parameters), why wouldn't you be able to factor that in? These "discrepancies" are introduced by something - conductor behavior in the stuff we're talking about. There will be a relationship there. Doesn't need to be linear of course, could be any curve you can imagine, but I don't follow how it would be mysterious and incalculable. Or at least I don't see why it would necessarily be mysterious in incalculable. Seriously?
All test gear has limiting parameters.
If a small sample of the substance being tested has an anomaly - deviance from ideal spec, cumulative distortion, cumulative error, whatever that falls outside the limits of you test gear it won't register on the output - but that doesn't mean it isn't there and it doesn't mean that it might not be detectable if a large enough quantity of the substance/ device/ whatever being tested were connected in series.
Connection in series = discrepancies are additive. Connections in parallel = discrepancies are averaged.
When discrepancies are additive they might or might not be obvious when measured in series. When measured in parallel it's less likely.
Therefore when a mile of X cable is tested for effects on the signal it far more likely that you'll detect an error than when measuring a smaller sample, OR EVEN when measuring a larger sample in parallel, which will tend to mask discrepancies.
Why wouldn't you be able to factor it in? Because NO test equipment is even remotely perfect and maybe what you're looking for is beyonmd the ability of your test gear to measure.
Sewriously, it's pretty obvious. If your distortion analyzer is accurate to 1% of THD it won't detect something with .5% THD - but is you series two of the devices under test together your analyzer might then detect that the two seriesed devices have a cumulative distortion of 1%, which is detectible to your (imperrfect) gear. And ALL test gear is imperfect.
The question of whether your ear is more sensitive to a given effect than your test gear is another question entirely. Ethan would say no. But then Ethan's ear apperas to be about as sensitive as two tin cans and a piece of string.
It's really pretty obvious.
Heheh. Welll I’m terribly sorry to be so dense as to not get your Super Obvious John Eppstein Points™️ If you mean uncertainties, it’s actually more complicated than you’re suggesting to propagate them properly. You have to use partial derivatives and add error ratios in quadrature if you want the uncertainties to end up correct after you calculate whatever it is you’re after. If you're just talking about test gear sensitivity, that's not going to effect how accurate your theory is, it'll just effect the specific degree to which a given measured quantity (and thus calculations resulting from it) is accurate. You're using "discrepancy", which would normally mean the difference between expected and measured values. Why does that, to your mind, make anything less able to be figured out? I feel like you're jumbling a few concepts together here. At any rate, none of this is what I was talking about. I was speaking generally about relationships. When there are causal relationships, even complex ones, we can often model them. And where there is error and uncertainty, we bring that along so that the model isn’t overstepping its bounds as far as what it can and can't say. None of this means you’re wrong about whatever it is you think is unable to be calculated, but it might mean it's a coincidence if you're right.
|
|
|
Post by Martin John Butler on Nov 18, 2018 21:09:54 GMT -6
I'm beginning to think I accidentally came to a Modern Science website ;-)
Have a good night guys.
|
|
|
Post by Ward on Nov 18, 2018 21:45:20 GMT -6
I'm beginning to think I accidentally came to a Modern Science website ;-) Have a good night guys. I thought I accidentally logged into gearslutz
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Nov 18, 2018 22:08:54 GMT -6
I'm beginning to think I accidentally came to a Modern Science website ;-) Have a good night guys. Knowing the science as well as the art of what we do will only make it easier to get where you want, knowing the science behind the tools you use only makes you use them with more confidence. I am the first to admit often what we love about some gear is what it dose wrong if we are looking for true signal integrity. Knowing an understanding what makes us like it even if what we like is distortion helps us get that tone again and again.
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Nov 18, 2018 22:10:07 GMT -6
I'm beginning to think I accidentally came to a Modern Science website ;-) Have a good night guys. I thought I accidentally logged into gearslutz Nah then someone would have said something about somebody else’s mother or that they liked amateur colonoscopy.
|
|