Describe The Sounds Of Your Favorite Caps

[okgb]

I can understand if someone is stocking up , or bulk buying
wanting to get the " best " overall choice , I think is a valid question

 

[CJ]

when you roll up a bunch of tin foil for a cap, you also are making a mini inductor,

can this change the sound? probably,

open up a wima cap, the foil is put down in layers instead of rolled,

i am guessing this is to lower the inductance,

yes it is a subjective topic as everybody has different ears and they also listen to different types of music,

quality and price do not always make a difference,

replacing the 250 pf ceramic in any fender amp with a silver mica does not really help, and i thought it sounded a little harsh with the mica but this is where you need some fancy equipment, to verify that your morning wake and bake are not playing tricks on your experiments,

 

[JohnRoberts]

Highly parallel versus serial internal structure reduces ESR and ESL (equivalent series L and R).  For these terms to be audible they need to significant in the context of the capacitors reactive impedance at frequency of interest. These are mainly issues in power supply and perhaps passive speaker crossovers where there is significant current flowing or at very high frequency where reactive impedance is low.

I suspect (ASSume) the sonic character that you call a subjective difference between caps is related to "voltage coefficient", or a measure of actual amount of capacitance change WRT terminal voltage. This mainly expresses in audio filters where there is a pole frequency in the middle of the audible passband, so the subject caps has significant AC terminal voltage, and a changing capacitance with that terminal voltage that introduces nonlinearity and distortion products that could be audible if large enough.

Of course good caps (COG, NPO) have very small voltage coefficients.

JR

 

[ricardo]

...  it's better to work on basic skills first ...

Here's Basic Skill no 1 for Listening Tests.

Make 2 IDENTICAL units.  They have to measure AND sound identical.  Then modify one & compare.  If you don't do this, you are better off sending me US$500 for a sample of my virgin crafted Unobtainium capacitors.

When you start doing Blind Listening Tests properly, the first thing you will find is

• Most (all?) self-declared Golden Pinnae are deaf.
• Much later on, when you have spent a lot of time & money on Listening Tests and amassed a lot of experience, you will find the true Golden Pinnae.  It will probably be your mother or girlfriend.  Also be prepared to be amazed at what these people can distinguish under some conditions some of the time. (A sure sign of a deaf wannabe Golden Pinnae is ranting about chalk & cheese.)
• Only then will you determine the true sound of capacitors etc

PS  Better to make 3 identical units.  Then you can slip in the odd comparison with the 2 remaining identical units and see who talks about chalk & cheese.

 

[lassoharp]

Make 2 IDENTICAL units.  They have to measure AND sound identical.  Then modify one & compare.  If you don't do this, you are better off sending me US$500 for a sample of my virgin crafted Unobtainium capacitors.

If the testing circuit is going to be a tube circuit I would suggest trying to use only a single transformerless circuit for the following reasons.

1) Getting two identical (design & component wise) sounding circuits may be difficult due to the differences in the tubes.  I very rarely see two tubes that give same voltage set under a given circuit.  They will also likely have different noise & distortion characteristics.

2) Basically the same thing if attempting to use transformers.  Plus the expense of using two transformers and the trouble of finding two that measure identical.

3) Time will be saved on the numerous measurements & expense required for proper matching including getting them to hold identical gain.  Even with a single circuit the tube characteristics will change over time, so be careful to allow a consistent warm up time before testing begins.


Since there will likely be only one or two coupling caps in the circuit, a switch can be used to switch between 2 or more caps for personal listening and live auditions.  If there are real differences the sonic signature should translate to a digital recording which IMO will be needed to allow a greater number of listeners access and also help eliminate the psychological factors involved in the in person auditions.  To further remove psychological biases for online tests, I would suggest simply asking participants to listen to the samples and describe what if any differences they can hear and just ask them to pick the ones they like the best, without any mention of caps.  I for one would probably have a pre bias if told the test was regarding caps.  For the reasons that Ricardo stated, I would try to keep the participants limited to those who record and mix music professionally.

 

 

[Blue Jinn]

Bang! when I hook them up backwards.

 

[okgb]

that's a Bad sound right ?  Don't hook them up backwards

 

[etheory]

that's a Bad sound right ?  Don't hook them up backwards

I don't know, I think it would be good for accurate sound reproduction of the 1812 Overture by Tchaikovsky, especially if timed correctly....

 

[okgb]

But you need a big cap for the Royal ending !

 

[JohnRoberts]

But you need a big cap for the Royal ending !

Speaking of big caps... the Toyota Hybrids that just raced at Le Mans used capacitors to store energy from braking on the racetrack. The Audi's that won used flywheels to store energy. Curious that neither "hybrid" used batteries....

Now I guess those Toyota caps would be "fast", something like 150 mph average lap speeds.

JR

 

[Kingston]

sure caps have a sound. especially audible when comparing two channels of whatever unit that has 5-10 of certain type in a signal path.

But what really blew my mind was when several years a go I made one of my passive EQ's a "mastering unit". The only thing I did was replacing four (brand new) carbon pots with equal rotary switches. I have since read that carbon pots "can be noisy". But that noise didn't show in any of the measurements I did. Maybe I simply lack the precision measurement devices that would be needed.

Anyway, the channel with rotary switches sounded much better. I don't know what exactly, except "transients are sharper".

I've also compared lots and lots of opamps, DOA and chip. There are clear audible differences there, even if very detailed FFT plots are roughly equal shape.

It would make my life much easier if I could measure these things I'm hearing. Currently I consider voicing a unit a bit of a black art, and the thought makes me uneasy.

 

[ricardo]

It would make my life much easier if I could measure these things I'm hearing.

A Listening Test IS a measurement.

You can do it properly by eg starting with what I suggested earlier (theres a lot more) or do a really slipshod measurement like most wannabe Golden Pinnae.  Your measuring instrument is your Listening Test Panel.  It has an accuracy which has to be checked and calibrated regularly, just like any other instrument.

If you don't do all this  ...  :-\  I give up.

What you need are the most musical capacitors in the known universe.  Hand carved by virgins from solid Unobtainium.  Send me US$500 in used bank notes for a sample. 8)

 

[Kingston]

You can do it properly by eg starting with what I suggested earlier (theres a lot more) or do a really slipshod measurement like most wannabe Golden Pinnae.  Your measuring instrument is your Listening Test Panel.  It has an accuracy which has to be checked and calibrated regularly, just like any other instrument.

I know this well, but like I said it makes me uneasy. It's on my long term to do list to create a lab suite of tests that correspond with what I'm hearing.

Also, I don't want to get an opinion from a measurement device, but quantifiable scientific data.

 

[ricardo]

Also, I don't want to get an opinion from a measurement device, but quantifiable scientific data.

A properly conducted Listening Test is a measurement that produces quantifiable scientific data.  Here are some examples.

• Nobody can tell the difference
• Some people, under some circumstances, can reliably tell the difference
• Some people who claim to be Golden Pinnae can't tell the difference under any circumstances though they insist the differences are huge

This tells you whether the difference is worth investigating and who to listen to.  It may be only John Atkinson can tell the difference and only on a test signal which leaves him with headache for the next 24 hrs.

When you pass this stage, (ie establish there is an audible difference, at least for some people under some circumstances which are relevant) only then do you ask for opinions.  The preferences themselves may not be scientific but statistical data on opinions IS quantifiable scientific data.  eg

• Of the people who can tell the difference, all prefer one option
• The people who can tell the difference are divided in their preferences
• The overall preference of these true Golden Pinnae changes depending on the music (circumstances)
• The people who can reliably tell the difference have no preference

All the examples quoted (including the JA anecdote) are true and taken from more than 10yrs of using Double Blind Listening Tests ala Lipsh*tz & Vanderkooy bla bla for the design of speakers & electronics.

 

[kevinkace]

Lipshitz

Someone needs to tone back on the language filter...

 

[RATMNL]

Ok, following article comes down to what my Studio-Technology teacher (http://www.grimmaudio.com/pro_products.htm) said when I told him I wanted to run a Cap A/B test... He got Mad! don't know how much of it has been adequately proven, but I put stock in his words.

http://www.avguide.com/forums/blind-listening-tests-are-flawed-editorial?page=1

(he did not write this, but it summarizes: The answer is that blind listening tests fundamentally distort the listening process and are worthless in determining the audibility of a certain phenomenon.)

 

[ricardo]

The answer is that blind listening tests fundamentally distort the listening process and are worthless in determining the audibility of a certain phenomenon.)

Why should not knowing what you are listening to make the test worthless?  Is that because when you don't know that one option costs 1000x more than the others, you can't tell it sounds better?  Maybe, if its not a Blind Listening Test, it is actually a Deaf Listening Test.

There are serious faults with tests conducted on many people at the same time.  You need to test each person individually.  But have you tried devising a Blind Test where all your concerns are addressed?  I suggest you try.  Surely it can't be that difficult? 

It will take some time before your test has the rigour of L&V but long before that, you will at least find out who in your test panel is deaf.  Just be prepared to find that its you.  8)

You can do Blind Tests with 'single-presentation observational listening protocols practiced in product reviewing'.  I have done a lot of this.  The logistics & cost go up but the principles are the same; the Blind Listener must not know what he's listening to but needs to have enough control over presentations so he can distinguish small differences.

BTW, when you start doing Blind Tests you also find how easy it is to skew the test, inadvertently or deliberately (ie cheat).  But it costs so much to do Blind Tests properly that its not worthwhile cheating and wasting the info.

 

[okgb]

Almost like a troll , haven't heard from the o.p. instigator recently
what's matter decided caps are too expensive or you can't tell ?

 

[JohnRoberts]

It appears to be an itch that several wanted to scratch... and a common theme.

Double blind listening tests are great unless you don't like the results. I've told this story several times but it seems appropriate to repeat it here. I recall a double blind test of studio monitors put on by a recording magazine using engineers and producers from the left coast. Long story short, several pulled out of the test and refused to let their names be mentioned in print after being informed that they had selected AMR (Peavey) monitors as the best sounding. Oops... 

JR

PS: Speakers can sound pretty different, capacitors not so much when properly used.

 

[pucho812]

my favorite caps only make sounds when they explode