1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

fo.q tape

Discussion in 'audio' started by MarkieW, Jun 30, 2022.

  1. MarkieW

    MarkieW pfm Member

    well kind of I suppose, the crossover is now behaving differently, and the sound has changed. I did not "measure" the sound before, just used the Mk1 ear.
    NickofWimbledon likes this.
  2. MarkieW

    MarkieW pfm Member

    Actually I thought there were a lot of resistors....
  3. Purité Audio

    Purité Audio Trade: Purite Audio

    Just anecdote then.
  4. G T Audio

    G T Audio Trade: Manufacturer and Distributor

    It's very doubtful that any measurements would show up minor changes in tonality and timbre. Remember measurements don't show up everything we hear...

    As for the perceived changes in sound: seeing as the drivers had to be removed, changes in sound could have resulted in removing the drivers and refitting them and tightening up their securing bolts. That is more likely what is being heard, rather than some tape...
    Durmbo and Old Valve Git like this.
  5. MarkieW

    MarkieW pfm Member

    Thats your opinion. I have done the experiment and have more data
  6. Rosie

    Rosie pfm Member

    Good point . Very good point
    However we now have two users reporting significant improvements in SQ as a result of applying dampening tape to capacitors .And I would remind those interested in this thread that Noel Keywood , Editor of Hi Fi World , a chap with a good technical knowledge was experimenting with Blue-tak wrapped crossover capacitors with good results 25 -30 years ago .

    I think an open mind is a real asset in our hobby , it's only a hobby .

    Quadraspire say their largest market is not the UK , it's Japan because Japanese Audiophiles get it re racks etc
    Fergus likes this.
  7. NickofWimbledon

    NickofWimbledon Well-Known Member

    First part - I for one agree, though I'd accept 'the measurements readily available today' for 'measurements' and 'everything that reaches our ears and (more obviously) what reaches our minds' in place of 'everything we hear' if people really want more precision. Brevity is not my middle name!

    More important, isn't that exactly what's being debated, here and on other threads? Whether there can be anything to hear or measure and whether they are the same thing?

    As for the second bit, I think most Subjectivists would agree it's worth checking things like 'What else did you change at the same time?' when thinking about upgrades, from power amps to the less conventional. Even we use Occam's Razor.

    Also, @MarkieW - I agree that your results are data, because all results are data! Others clearly don't.

    Finally, @Rosie - that's my recollection too. Several people have asked why amp makers don't do some advanced version of the same if the effect is real and has been known for ages.

    'We are still debating whether it does anything after all this time" may be the answer, or

    "We reckon nearly 5% of our customers can actually tell, and what they don't know won't hurt them (or sales)".

    I think a top answer for why they don't put Blu-tak on capacitors is probably what it could do to long-term reliability or even safety.

    Let's just agree that there could be answers that don't include 'We assessed in detail and it doesn't work'.
  8. Sue Pertwee-Tyr

    Sue Pertwee-Tyr neither here nor there

    There's also the time taken to put Blu Tak, or whatever, on caps during production. I imagine the costings might bugger that up.
  9. cooky1257

    cooky1257 pfm Member

    Years ago now the Tannoy System 15 DMT Studio Monitor had damping sleeves on their HF caps and also placed the main LF inductor well away from the driver magnet and the other crossover components, they also used Kimber cable....Given the application it's a rather hostile environment inside a big monitor....

    As you can see it's a tad more 'serious' than a little foam sticker.


    Their current TOTL 'Audiophile' Domestic range doesn't bother though....ESA and MR grade Clarity caps. See link in post below.
    westsea and Rosie like this.
  10. gingermrkettle

    gingermrkettle Deep vein trombonist

    cooky1257 likes this.
  11. NickofWimbledon

    NickofWimbledon Well-Known Member

    Thanks. I need to do my homework following this and @Jim Audiomisc's links. Before I put a wet towel on my head and start trying to remember what most of this means, can I check 2 things?

    When it says: 'no variation in current is assumed here' that is obviously across the body of a capacitor, but are assuming the capacitor is being vibrated while itself getting a constant current? Also, when using music and listeners, they concluded that there was No Effect, but were they only testing for SQ differences from changing one capacitor in the HF crossover of a speaker?
  12. John Phillips

    John Phillips pfm Member

    The physics described is basically real. On the issue of capacitors vibrating audibly, forces do arise from current flowing through a film capacitor. Because of how a film capacitor is constructed those forces mostly cancel out but not entirely. A capacitor will mechanically vibrate to some extent. Figure 2 shows this but the very low absolute audio level generated is at or below "quiet domestic listening room" level. We aren't told what signal was applied to the capacitor or how far away from it the microphone was. Important details omitted :(. If 100 V ac was applied to the capacitor and the microphone was a few cm away then yes it will be audible but not from inside a loudspeaker cabinet to a normal listening position with normal signal level applied to the loudspeaker.

    That doesn't address your first question. It's not for me to say how someone should read this paper but the key question to me is has the paper demonstrated that this physics matters? Or is a mountain being made from a molehill? If anyone is of the persuasion that human hearing is perfect and everything is audible then this will not be a relevant question.

    On your second question the system used for listening is not clear enough. It looks like they were bi-wiring a B&W 805 loudspeaker. The power amplifier connection to the LF terminals is direct. But they write "LF driver" - is that right? If so what happened to the LF crossover? The connection to the HF terminals goes via one of two possible capacitors in series. Again they write "HF driver". Did they leave the LF crossover in place and just bypass the HF crossover in the loudspeaker? Or did they leave both in place and merely put one of two different capacitors in series with the HF crossover? The text only talks of connections to the two drivers without distinction. Maybe just the HF crossover was bypassed and replaced by capacitor A or capacitor B; and AIUI the B&W 805 HF crossover is simply a series capacitor so that would make sense. I assume this was done properly but the text could be clearer.

    However, the preparatory listening tests failed to demonstrate that an electrolytic capacitor could be distinguished from a film capacitor. A difference described as "fairly audible" failed to materialise in the listening test. Moving on from concerns about missing details and lack of clarity, that causes alarm bells to start ringing for me.

    Then to the main test. We have some graphs aggregating what the listeners reported as differences in "clarity" and "spatial information" between capacitors A and B and a "preference" but no connection at all is made between the results individually or in aggregate and either A or B, let alone any statistics to demonstrate significance. No useful conclusion is possible AFAICS. This does not mean that differences are inaudible - just that ISTM this paper hasn't demonstrated it.
  13. gingermrkettle

    gingermrkettle Deep vein trombonist

    Although given that there is a stated sample size you could fairly easily do a chi-squared test on those results to check statistical significance, which given the results of the preference survey would appear to show that a random allocation of preference in this case does not fit the data. If so then we would have a situation where ABX testing does not show statistical significance for the effect of a variable but another protocol could.
  14. John Phillips

    John Phillips pfm Member

    I wonder if you may be assuming something I didn't see explicitly in the paper?

    Take just the "clarity" graph. The labels are just "Duller", "Same" and "Clearer". Are you assuming they should be "A duller than B", "A same as B" and "A clearer than B"? There's no indication that I can see that they aren't "One of A or B was duller", "A same as B" and "One of A or B was clearer".

    Unless I missed it, the results haven't been explicitly connected to either A or B. It is possible (even probable) that the results do favour the sponsoring manufacturer's product because this is really marketing and can't be reliably read as good science. If so you can do some statistics as you say. But the paper does not make the connection explicitly and I never make assumptions about what is not stated when reading technical papers. It's too easy to fool yourself.
    NickofWimbledon likes this.
  15. sq225917

    sq225917 Bit of this, bit of that

    Guy on diyaudio built a cap tester years ago and measured the output from various types and then nulled them with software, worse case was lytics down around -100db best case was film caps around -125db

    Ranging from sunk so far below signal to be utterly inaudible to why are we even testing this.

    Same with resistors.

    Coils on the other hand.....
  16. gingermrkettle

    gingermrkettle Deep vein trombonist

    Yes, it isn't exactly clear on the specific characteristics although the preference graph does specify which characteristic has the higher number of preferences from the sample population.

    We do though come back to the question of whether this is showing a statistically significant difference from a properly designed experiment (if there is a clear preference then there must be a difference by definition) in the second methodology when ABX did not show any difference.
  17. sq225917

    sq225917 Bit of this, bit of that

    Yeh, they cheated.
  18. martin clark

    martin clark pinko bodger

    </ my snip from rest of a very fine post>

    Something I'll add here, though in absolute terms the noise level is in the weeds - if you've ever bench-tested an amplifier of moderate power, say at 50w or more into a dummy load, case-off - then yes, you will usually be able to hear the test-tone from the circuit board! regardless of type of par.

    The electrostatic forces in capacitors can be surprisingly -large; but sim also occurs in every other part on the board, to a lesser extent. There's no magic here, just basic physics.*

    & yes I've tried sticking sticky-stuff on various parts when doing such; and no, it doesn't do anything useful, either; again, just basic physics.

    * e.g. in fact , Wima even markets a ' LN' = low-noise, variant of some of their large film caps for industrial use in inverters & the like... sim approach to clarity caps - denser cases/ encapsulation, tighter winding, 'potting' the part to stiffen the assembly / restrict the internally-generated vibration.

Share This Page


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice