But even if you did that, you couldn't definitively say that the change was caused by the capacitors because that's not the only thing that changes. Some examples:
- During the time it takes you to swap the capacitors your ears get a much needed break, so your perception changes.
- If you listen at relatively high volumes during the listening test, you will introduce a temporary threshold shift. We've probably all experienced that when we've gone from a noisy environment to a quiet one and suddenly experienced that we couldn't hear subtle sounds. That changes your perception when you test the modified equipment after a period in relative silence.
- The test of the modified equipment would take place at a different time of day or on a different day where the ambient noise is different, so a difference is heard.
- Etc, etc.
Then add that the auditory memory is pretty short. A few seconds to maybe 10-15 seconds at the most. So you simply won't have a direct comparison in your mind unless you switch back/forth at a faster rate, which isn't possible if you have to swap out components.
So any perceived change may not be due to the capacitor swap.
If you wanted to test this properly you'd build two identical pieces of equipment and swap the capacitors in one of them. That's pretty easy to do in the DIY world. Then swap back/forth. Or better yet, run a proper ABX test. Or, better still, arrange for someone else to administer the test and recruit 250 people to judge whether A is better or B or there's no difference. Then do the stats math. I think most DIYers would lose interest at this point.
Tom
I've already done that before, and so have others. Its just that once you do it, people ignore it, forget it, dismiss it, pretend it never happened, etc. That kind of dismissiveness has been going on for decades.
Then the critics go right back to their old arguments. I pointed the problem out here in this forum years ago back when Scott Wurcer and Jakob2 were still here. Some things just don't change.
Absolutely. I put them on pin headers so I could make quick changes. What an eye opener.
And what exactly were these two identical pieces of equipment that you used to swap capacitors in. Amplifiers, pre-amps, DACs, etc.
Last edited:
I don't think a tit-for-tat will get you anywhere, and that may be why this discussion refuses to die. There is also the possibility that the research does not exist because research tends to show what does work not what doesn't.
I bet it still took longer than 15 seconds, though. You also knew what changed, so the experiment is entirely sighted. While that might be an entertaining way to pass time I'd caution against drawing too many conclusions from such a test.
Tom
About 1 second. For example, testing was done with an AK4499 eval board custom Vref supply output filter. Of course it is a sensitive spot. Vref is multiplied with the dac output to produce the current for the I/V in the output stage (which amounts to convolution in the frequency domain).
Another thing I did that was soon forgotten and dismissed was to compare recordings of non-inverting audio opamp buffers in a listening test put on by PMA. He wanted ABX but it was hard, so I decided to DBT sort the opamps recordings in order of distortion by ear in a DAW. My idea was to find the two most different sounding recordings. It was very hard, and I still had one opamp to sort. So I said, forget it. instead I PM'ed PMA with my sort order before he announced which audio opamp was used for each recording. PMA gave me credit in the thread for sorting them in order of distortion solely by ear (it was impossible to do sighted before he announced which opamp and its distortion measurement for each recording). Like I said, it was very hard, but I did it anyway. People conveniently forget about that when they start demanding DBT over and over again to prove its possible for someone to hear something much easier than sorting unity gain opamp distortion by ear. IOW, people ignore what has been shown by DBT and go back to their old arguments.
I'm not sure that looking towards swapping components, is a guide to how different the same components will sound a few hours later. I think even when you do find different sounding caps, the idea they will sound different again after a few days isn't on our minds.
A more suited test, would be to build two items. Test they are the same. Then run just one of them for a week.
I don't feel such testing is blind, until I figure in some switchgear and an assistant. Though my assistants do tend to tire, after making their own snap decision. Often 'they sound the same'
A more suited test, would be to build two items. Test they are the same. Then run just one of them for a week.
I don't feel such testing is blind, until I figure in some switchgear and an assistant. Though my assistants do tend to tire, after making their own snap decision. Often 'they sound the same'
And what was the piece of equipment that these changes were being made to? You seem to be avoiding any details that might bring into question what you did or did not do.
Technically over time capacitors get worse.
So if burning them in makes them "better"
Looks like the solution all this time for " quality"
is just make a horrible power supply.....
maybe even borrow " burned in" capacitors from repair
techs that have removed them from old pre burned in systems...LOL
Or say if I re invent the wheel to be square.
thump, thump , thump , thump.
Long as I argue till the point of no end and " MAKE UP NONSENSE"
Eventually a square is likely the greatest wheel ever!!!
Imagine if you had a square wheel and you were driving on a perfectly smooth surface. The square shape would actually offer some unique advantages:
So if burning them in makes them "better"
Looks like the solution all this time for " quality"
is just make a horrible power supply.....
maybe even borrow " burned in" capacitors from repair
techs that have removed them from old pre burned in systems...LOL
Or say if I re invent the wheel to be square.
thump, thump , thump , thump.
Long as I argue till the point of no end and " MAKE UP NONSENSE"
Eventually a square is likely the greatest wheel ever!!!
Imagine if you had a square wheel and you were driving on a perfectly smooth surface. The square shape would actually offer some unique advantages:
- Predictable Motion: Each corner of the square wheel hits the ground in a predictable pattern, providing a steady rhythm. This can be useful in applications where consistency and reliability in motion are more critical than the smoothness of the ride.
- Improved Traction: The flat surfaces of a square wheel could offer more contact with the ground, potentially providing better grip in certain situations. For example, on loose or uneven terrain, the larger surface area could help prevent slippage.
- Efficiency in Certain Tasks: In specialized scenarios like moving heavy objects over a flat, non-paved surface, the square wheel might be more efficient. The flat sides could make it easier to push or pull the object, as the wheel could "lock" into place on flat surfaces.
- Durability and Maintenance: Square wheels could be easier to manufacture and repair in some environments. If the wheel is designed with replaceable or interchangeable flat surfaces, it might be simpler to fix or replace a worn section compared to a circular wheel.
I'm not sure if the following is on topic, but I just remembered the following.
Some time ago someone on the Forum complained about his disappointing purchase of a pair of good brand headphones that didn't satisfy him.
I advised him to give it an adequate burn-in period with music signals before throwing it away.
He never replied again in that thread.
However, I was wondering about the following: if after a burn-in period we can't hear the difference in the sound of a pair of headphones (every point of view is acceptable here, since no one can prove anything yet), how can we hear the difference between two different models of headphones of the same brand and of similar built?
In short, it seems that it is not even possible to prove whether it is a question of subjectivity or objectivity.
Any thoughts?
Some time ago someone on the Forum complained about his disappointing purchase of a pair of good brand headphones that didn't satisfy him.
I advised him to give it an adequate burn-in period with music signals before throwing it away.
He never replied again in that thread.
However, I was wondering about the following: if after a burn-in period we can't hear the difference in the sound of a pair of headphones (every point of view is acceptable here, since no one can prove anything yet), how can we hear the difference between two different models of headphones of the same brand and of similar built?
In short, it seems that it is not even possible to prove whether it is a question of subjectivity or objectivity.
Any thoughts?
I liked this, sagacious and witty. 😎
Of course, even some wines have their own aging period within and no later in which they perform at their best.
Regarding "burn in":
We see currents. And the courses of these currents are constantly changing. In the case of lightning strikes and rivers, we have quasi search currents that are scanning ahead. And with constant current, with repetitions, we have "paths with less resistance".
From this we can also deduce what a current conductor or signal conductor should ideally look like, in shape and diameter, in order to reduce even the shortest search currents and secondary paths, that audibly "distort" the signal.
Physics only;-)
We see currents. And the courses of these currents are constantly changing. In the case of lightning strikes and rivers, we have quasi search currents that are scanning ahead. And with constant current, with repetitions, we have "paths with less resistance".
From this we can also deduce what a current conductor or signal conductor should ideally look like, in shape and diameter, in order to reduce even the shortest search currents and secondary paths, that audibly "distort" the signal.
Physics only;-)
Attachments
Last edited:
Burn in of loudspeakers or transducers is claimed as a subjective topic.
But has been known that mainly suspension and surround will change
after a short break in of 10 to 50 Hrs
likely broad opinions how it effects headphone transducers.
Headphones not much different
they either do or dont have as much bass.
And some have more or less fatigue at high frequency
or non that is notable.
As far as high quality that are pleasant
and work well.
Between brands like headphones, cars, planes, can openers or water bottles.
Some people have a favorite brand and will just say it is better regardless.
It is subjective yes, so will spin in circles not much differently than this one.
In capacitors, particularly electrolytic capacitors,
stabilization of the dielectric is a real phenomenon, and it is not purely subjective
as far as the actual audible difference it will remain subjective.
Assuming 10% tolerance power supply caps.
The actual capacitance does not change enough after
the dielectric has stabilized. It is still within the advertised
range.
Far a ESR and Leakage current those can change more.
But less than 10 hrs of operation to stabilize.
Leakage current being much more measurable.
Of course take a wild guess where it is more measurable.
Dirt cheap capacitors or units that have been stored extended
periods. Being most datasheets for capacitors make a big
note about storage time and temperature.
Stabilization occurs faster at rated voltage and temp.
So basically if you buy dirt cheap caps.
that have been stored for extended periods.
Then operate say 30 volts instead of 100 volts.
Leakage current would be actually measurable
change over a longer period.
up to 50 Hrs instead of 10 hrs
Also well known a cap rated for 3000 hrs
that has been under constant operation
goes way way above the hr rating.
A brand new cap being stored in very low temp
with no voltage will often fail before its rated Hrs
This is the nature of electrolytic
other film or metal types can be mainly ignored
There are also ambiguities regarding objective and subjective:
If everyone (except for a few who cannot hear differences or cannot verbalize "audible difference", or even some who are partout against objectivity, or majority) hears a difference, it is objective.
If all participants who hear a difference describe it in the same way, categorize it in the same way, such as trial 1 = black or trial 2 = grey, it is objective.
If half of the participants prefer trial 1, black, and the other half prefer trial 2, gray, then it is subjective.
Listening, hearing is a measurement process with objective results. The terms objective and subjective are used synonymously for e.g. low-complexity procedures and high-complexity procedures or also synonymously for quantitative method and qualitative method. This results from a lack of training in methods: research methods; study criticism... a lack of scientific education.
If everyone (except for a few who cannot hear differences or cannot verbalize "audible difference", or even some who are partout against objectivity, or majority) hears a difference, it is objective.
If all participants who hear a difference describe it in the same way, categorize it in the same way, such as trial 1 = black or trial 2 = grey, it is objective.
If half of the participants prefer trial 1, black, and the other half prefer trial 2, gray, then it is subjective.
Listening, hearing is a measurement process with objective results. The terms objective and subjective are used synonymously for e.g. low-complexity procedures and high-complexity procedures or also synonymously for quantitative method and qualitative method. This results from a lack of training in methods: research methods; study criticism... a lack of scientific education.
@WhiteDragon
Interesting comment of yours.
In particular:
In my opinion, we can call this whatever we want, both "stabilize" or even "burn-in".
And from what I understand (IF, I understand) it is even measurable, so it would be objective.
But all doubts remain if we think that it is also audible, so eventually it it would be subjective.
However, what I continue to ask myself is this: why, if on the one hand we accept the fact that capacitors need a "stabilization" or "burn-in" period, can we not simultaneously accept the fact that, just like some wines, they have a "use" period within which their electrical and sonic performance changes?
And if they do change, as I'm convinced they do, why should we exclude a priori that someone could not hear its effects?
On the other hand, the same thing will probably happen for resistors, diodes, or whatever, or not?
Interesting comment of yours.
In particular:
In my opinion, we can call this whatever we want, both "stabilize" or even "burn-in".
And from what I understand (IF, I understand) it is even measurable, so it would be objective.
But all doubts remain if we think that it is also audible, so eventually it it would be subjective.
However, what I continue to ask myself is this: why, if on the one hand we accept the fact that capacitors need a "stabilization" or "burn-in" period, can we not simultaneously accept the fact that, just like some wines, they have a "use" period within which their electrical and sonic performance changes?
And if they do change, as I'm convinced they do, why should we exclude a priori that someone could not hear its effects?
On the other hand, the same thing will probably happen for resistors, diodes, or whatever, or not?
For example, there is red-green blindness. If, for example, we wanted to display lines in visual measurement procedures with these colors, these participants will not perform well. Many of them might even start to argue and quibble;-)
Let's assume that many audio discourse participants have limited hearing;-)
Many audio discourse participants have an education that has ZERO to do with audio. Their education includes only visible lines and numbers, a purely conceptual and highly limited view that does not even correspond to a model, not even a symbol for audio, hearing. Whoever claims that these lines and numbers correspond to the subject of audio, whoever does not even notice this misstatement, and treatment of a subject that misses the subject...-)
A common electrician's training only includes: Close circuit: current - open circuit: no current. Even the highly complex process and form of current is not taught. Those trained call quickly "stop trolling" when they are confronted with their horizon of experience and knowledge...-)
Let's assume that many audio discourse participants have limited hearing;-)
Many audio discourse participants have an education that has ZERO to do with audio. Their education includes only visible lines and numbers, a purely conceptual and highly limited view that does not even correspond to a model, not even a symbol for audio, hearing. Whoever claims that these lines and numbers correspond to the subject of audio, whoever does not even notice this misstatement, and treatment of a subject that misses the subject...-)
A common electrician's training only includes: Close circuit: current - open circuit: no current. Even the highly complex process and form of current is not taught. Those trained call quickly "stop trolling" when they are confronted with their horizon of experience and knowledge...-)
stabilize as in specifically the dielectric material only in electrolytic type.
as mentioned it is perfectly known phenomena
with electrochemical process to make these types.
less than 3 to 10 hrs to stabilize
at rated voltage.
the actual measurement of capacitance change
is negligible
it is within tolerance.
ESR and leakage also change within first 3 to 10 hrs
also negligible
again only issue is with dirt cheap capacitors
which are run below voltage tolerance.
measurement of " burn in" changes will only
be obvious with current leakage.
as mentioned in my first comment.
Since Diy power supply uses 2x to even 4x
more capacitance needed anyways.
Non of this is audible in any way shape or form
or your capacitors are just pure garbage.
logical explanation for op's issue was either
a difficult 4 ohm load or lower.
Usually a speaker that actually makes bass
amplifier performance yes is noticeable.
Could be unstable have offset issues or not enough
output devices. Cooking single pair
with overbias doesnt help either