Easy, peasy ... just put on one of your "bad" recordings ... 😀 , 😛
I'm being serious, actually. "Difficult" recordings are shining a very powerful torch on precisely every distortion artifact your system produces - I have a whole suite of recordings put aside, each one will very nicely highlight various weaknesses, distortion behaviour, in an unknown system - this is why "bad" recordings always sound enormously different from one system to the next ... just think about it ... if the systems were "perfect", and only the recording was "bad", then that such a recording would sound bad in precisely the same way, on every system - it would come across identically bad, on each.
That's not the case, of course - the recording is your stethoscope, probing into the frailties of the playback ...
Well, IF 600 ohms is an 'excess' load for the 2134, then 100 ohms is quite a challenge for the 4562. It should be obvious. By the way, Ron Quan measured the 2134 for crossover distortion under loading (500 ohms) and found it much better than a TLO-82 and many other IC's tested in the same way.
Karl, a thought I had was to always do at least two recordings with the "best" setup invariant as possible - one as the first recording made, and the other as the last done in the sequence. They serve as bookends to the test run, will hopefully highlight any overall drift in the experimental procedure, and provide a sanity check in the testing and listening.
Not really (/always).
Fashion/glamour models e.g. have next to perfect bilateral symmetry.
That is registered subconsciously by the eyes and brain.
Only thing most people are aware of, is that they percieve them as attractive.
The few exceptions who do not, do not experience symmetry in the median (sagittal) plane as pleasing.
Both can be verified with simple tests.
It's mating season at the catwalk again
Right Said Fred - I`m Too Sexy (The Original) - YouTube
FWIW, that's a BS requirement, there is absolutely no point for such, other than some sort of useless engineering challenge. You could as well ask for -160dB, or 20Vrms in 50ohm, at those levels it's the implementation that controls the numbers anyway.
So not sure what are you expecting for free, anything on paper/simulation would be useless, anyway. If you expect somebody to design it, build it, measure it, optimize the implementation 3-4 times (that would include PCB revisions, since you may agree you can't breadboard such an animal), not to mention that some power supplies able to withstand the output current may be required and optimized, yet another task. And everything for free and an 😎 emoticon from Mr. Marsh, good luck with that.
Last edited:
All this big waltz of claim and counter claim is silly. The test was seriously flawed. It allows anyone a 50/50 chance. And if you get it right, you can say "See? I told you so!"
Would you really except that kind of evidence for anything else?
If there were 4 recordings, then I could believe someone who could pick the good or bad ones. At just 2, there is no point.
I know that there are differences in the 2 recordings. But no one has yet convinced me they can really tell which is which by ear. I'd love to see someone pick correctly - so far, no proof.
Would you really except that kind of evidence for anything else?
If there were 4 recordings, then I could believe someone who could pick the good or bad ones. At just 2, there is no point.
I know that there are differences in the 2 recordings. But no one has yet convinced me they can really tell which is which by ear. I'd love to see someone pick correctly - so far, no proof.
That most can pick a difference is fairly clear. To make it more a test in the sense you would like it to be could be done by creating, say, 12 completely distinct recordings, where also the number using one setup vs. the other is also random, anything from 11 A, 1 B to 6 A, 6 B.
On the general subject of audibility, and transparency, of electronics there's a mob in Sweden who do it very nicely, LTS - been mentioned on the forum a couple of times - who have found only one amplifier which tested transparent ... a Bryston in the same generation as the one I listened to, of course, 😀 !!!
They do real bypass tests of the amp, while working into a fairly nasty, but realistic load ... the test of that amp is described here, ftp://bryston.com/pub/reviews/Swedish%20Review%2014B%20SST,%20part%20III.pdf.
They do real bypass tests of the amp, while working into a fairly nasty, but realistic load ... the test of that amp is described here, ftp://bryston.com/pub/reviews/Swedish%20Review%2014B%20SST,%20part%20III.pdf.
Last edited:
Thanks Mooly, for the simple x-y test that showed a difference. That is the point:
There was a difference. That is all we can really do in REALITY to show audio problems and attempt to avoid them.
There was a difference. That is all we can really do in REALITY to show audio problems and attempt to avoid them.
I didnt think you could do it, either. I admit that I probably cant do it. It would be SOTA. But JXC's suggestion has the potential to do it or come closest. David has a working variable freq oscillator and with low settling times that is this good (verified by measurments from two other independent testers) but wants a buffer so he can also drive low(er) Z loads than 100K with it but keep the low distortion so hard fought to achieve. No pain, no gain. At this point in time its a few dB here and there... all the easy means have been found. Another 10-12dB I think can be done now.
THx-RNMarsh
Last edited:
If they think more distortion is what they LIKE, then No. But I could concede that comment if they can at least hear a difference.
THx-RNMarsh
For the Hard Core ones only -
If I was to start with the body of experience and knowledge accumulated to date..... I think a faster and easier way now is to just get hold of todays test equipment - beg, borrow, or steal it (briefly and then return it). Look at the distortion under a variety of conditions... esp FFT levels and order...and as many ways as you can think of and more.... Each time you make a measurment, listen to it. Do it as often as needed to gain a memory of the affects. Do it with tubes and SS. I would hope this way would not take 10,000 hours. But over time, you will be able to know what you are hearing. Being a test equipment junky for ages, I learned what tube, IC, SS, FEt etc can do under many condions (including misbehavior... like clipping). You can also manipulate the distortion in amps if so inclined for more even order and odd order etc and listen. Coorrelate the measured data with what you hear. Soon distortion characteristics as patterns will begin to stick in your memory. At that point, you will know them when you hear them for what they are.
Easy to say.
THx-RNMarsh
If I was to start with the body of experience and knowledge accumulated to date..... I think a faster and easier way now is to just get hold of todays test equipment - beg, borrow, or steal it (briefly and then return it). Look at the distortion under a variety of conditions... esp FFT levels and order...and as many ways as you can think of and more.... Each time you make a measurment, listen to it. Do it as often as needed to gain a memory of the affects. Do it with tubes and SS. I would hope this way would not take 10,000 hours. But over time, you will be able to know what you are hearing. Being a test equipment junky for ages, I learned what tube, IC, SS, FEt etc can do under many condions (including misbehavior... like clipping). You can also manipulate the distortion in amps if so inclined for more even order and odd order etc and listen. Coorrelate the measured data with what you hear. Soon distortion characteristics as patterns will begin to stick in your memory. At that point, you will know them when you hear them for what they are.
Easy to say.
THx-RNMarsh
Last edited:
(If you get it wrong, you can state the same.
4 recordings also have a better chance of exposing the unjustified nays)
That's good advice... this one is going to be a bit different though 🙂
Yes... hmmm... we've sort of been there before in earlier tests and either some whizz with software cracks the code by whatever means... or most are non committal when it come to listening.
Your welcome. If this next one works out as intended then I hope you'll add your thoughts along the way 🙂 It should be fun.
That IC, I don't know, these IC's, hopefully you will have the chance to see for yourselves.
Ok. What was the IC p/n again? I'll test it myself under 100 Ohm load.
I'll get to increase the load to see where i cant tell the difference any more for that IC using better speakers.
THx- RNMarsh
I'll get to increase the load to see where i cant tell the difference any more for that IC using better speakers.
THx- RNMarsh
Last edited:
- Status
- Not open for further replies.