 Now, I got bashed by some Readers for the last reportage from the KSS meeting devoted to mains plugs (report HERE) I did really! Ohohoho! There were not that many critical mails, but all touched upon the issue of methodology. We conducted the listening test the classic way, just like I test the products normally, with long samples and without blind tests. And yes, we did not the blind test. And this was no “sloppiness”, “negligence”, “try to circumvent scientific reasoning”, “felony”, etc, just citing the most fitting descriptions found in the emails. If somebody is not convinced then I am not able to “convert” him or her and I will not even try. But if you are open for arguments, a few short words in that case. )
First of all, in the boundaries of KSS we never tested devices in the blinded tests regime. We used this methodology only while testing discs. This was a conscious choice, based on the two years of experience I gained organizing “Small Audio Shows” for “Audio”. In almost all editions of this popular cycle we organized double blind test ABX testing. It was done the following way: we marked two compared elements like RCA plugs as A and B, without informing the test persons about their names and without showing them. Then the same short fragment of a musical piece and switching occurred – ABX, BAX, and so on, informing the listeners about the sequence. The listeners had to point what is the X – A or B. In every round we made ten to fifteen sequences of switching. The participant received specially created cards, where they wrote what they thought is X in every test. To be able to prove the difference, that the two products really differ from each other, it is necessary to have 8 out of 10 right. It is however accepted, that the lowest borderline is 7 out of 10. Experience showed, that differences can be spotted that way, as the mentioned plugs did change the sound. But… CONFESSION OF FAITH The conduct of this kind of experiments – all tests should be treated like that – is not only depending on their technical preparation, but also on contact with people. And one of the things I learned during those two years, is that the stress during the experiment is so big, that a part of the takes is useless. Many times it was so that – seemingly – experienced listeners could not distinguish clearly different sounding components from each other. The tests needed to be repeated twice or trice (still without revealing the true names of the AB components) to lead to trustworthy results. Why? First of all the participants had to accustom to the system and the room. Seemingly not a big thing, as not the system but its part was tested, but the reorientation did not happen automatically and needed time. Hence the improper answers in the first attempt. What is the conclusion? That psychology is has as much an influence on the test as with any other, and the claim that this test is the “only one trustworthy” is faulty from the beginning. If we have to repeat the test, or conduct it three times, and if we do not hear the differences then, should we repeat it ten times? As many times as needed for the participants to “learn” the system, the music, the room, etc. So maybe twenty? I exaggerate a bit, because at the second round the participants were so tired, that they almost fell on their faces, but I want to tell you, that this kind of tests has the same problems like the normal AB tests. It also shows, that is known even in the camp of the ABX only believers: the listener is the same component of the test as its subject – he or she is not only the tester but also the tested. I resolved the problem inviting proven listeners, that turned out to have “golden” ears, what eliminates them in some way from the equation. Unfortunately not completely. )
The only thing that – in some way and not always – can be taken from an ABX test is if the difference exists or not. There is no way about defining the KIND of the change. Short samples just prevent that, and longer are useless from the ABX methodology point of view, even for establishing main things. In my opinion, in music, and audio is a part of that, only long listening sessions make sense. Only when we allow the emotional transmission of a musical piece reach us we can assess how it does sound. This is only my opinion, not a dogma, but it has support of four years of experiments – two in ABX an AB tests, and then another two with AB and open comparisons. In the end all tests in world magazines I value for professionalism and consistency in conducting tests, are done in a classic way, knowing what is tested and in long time periods. Yes, blind listening test can be valuable, but it does not account for everything. „[Meyer and Moran] in the October issue of Journal of the AES report the results of a double blind test comparing SACD and DVD-Audio recordings [recorded in native formats – redaction note] with their 44,1/16 versions. In general – no changes were detected. [...].”
DL What turns out from that? Exactly what I described in the beginning: despite the “objective” pretences of the ABX tests they are equally susceptible to external influences like other kinds of tests. Ok, I know, this should be the reportage from a KSS meeting, but where, if not here, I can explain my standpoint? Finally it is not about that everybody would now nod me, but about making my point of view, and thus the point of view of “HIGH Fidelity OnLine” is known. And even if we do not share this point of view, then it is better to know what we are discussing than leave everything in the mist of understatements and hints, that turn into semi-truth. If there will be a need for that, I will repeat this “confession of faith” at every occasion, so that everything is clear. Repeating after a classic – if we have to differ, let us differ beautifully… SOUND )
The fifty first meeting of KSS was organized shortly after the fiftieth, but some thing changed in the meantime. First of all some of the participants from the former meeting came to the conclusion, that they cannot live without the beryllium plugs. Going further, we came to the conclusion that we need to exchange the wall sockets to dedicated Oyaide or Furutech, voting with their wallets (or bank accounts). For as far as I know, we are not alone. Mains plugs are very exotic parts of the audio chain, but isolators, I want to talk about, or better said supports, (mechanical isolators) aren’t the center of audiophile world either. Now having tested the plugs, the supports were a piece of cake – effortless. The listening session was conducted in the same system as before, so I will not describe it again. Our goal was to determine “if” and if yes, then “to what extent” and “how” the separation of audio components from the surface they are placed on influences the sound. This time the most obvious candidate for this was taken for testing, namely the CD transport. In the test five sets of absorbers were examined compared to the standard cones supplied with the Lektor, made by Monacor. Three of the tested isolators, of Finite Elemente brand, were supplied by the distributor, the company Audio System. )
All the Cera supports are the products of the German company Finite Elemente, known mostly for their phenomenal, being my object of desire and silent dream, equipment supports (I mean the furniture) Pagode. Their success is based mostly on the attention paid to the detail, like isolators. All Cera supports are made around the same concept: we have a base, where the cradle for a ceramic ball is milled. The main element has a rubber ring around. On top of that another, broader element comes, also with a milled cradle for the ball. The complete structure is held together by the mentioned rubber ring, and the interface between the surface and the device placed on the isolator is the ceramic ball. In the biggest element, or actually “feet” CeraBase we have three such balls, and the element is held together by the central pin and rubber ring. Simple and effective, because you do not have to think about the pad – in the Cera we have the pad and isolator combined. Nordost is also a “rigid” isolator, but composed of two elements – a big main part with a sharp pin and a pad. There are two kinds of Pulsar Points available – aluminum and titanium one – we tested the latter. And finally the Vibrapods. Those are “soft” supports from two elements – a rubber, broad ring with a ply and a rubber cone with a steel ball in its end. The cone is supported by the mentioned ply. Because I babbled out a bit in the beginning, so I will come down quickly to the point: absolutely unrivalled, in our opinion, were the CeraPuc supports, that I bough just after the test, and other participants are aiming to get them. The sound became nobler with them in place, became more focused. Interestingly, and I confirmed it with my Lektor, the bass sounds stronger with them – it is condensed and contoured, and in general more present. The sound gained on depth and dynamics. In systems, where there is a problem with the attack of the sound, so there where the sound is a bit hard, CeraPuc will underline this problem. But if we have a not fully controlled bas, then the supports will discipline it a bit, but will also point to its problems. The smaller CeraBalls go in the same direction, but the changes were not that dramatic. This is the same direction, maybe even the sound gets more disciplined, but without the smoothness and naturalness opened by the CeraPucs. The CeraBase was tested with the stands that supported the Sonus Faber Electra (description HERE). The effect was similar like with the CeraPuc – improvement of vividness, coherence, without sharpening, and even with a slight smoothing of the treble. We tested two settings – under the stands on the marble supports and under the marble supports and on the floor. The last version was disastrous, the first one – super. )
This is of course our subjective opinion, but we devoted quite some time and effort to the supports, repeating listening tests, listening to the same piece many times, so in our opinion this is how it really is. If I had to establish a hierarchy of the isolators listened to, then it would be like this (starting with the best): CeraPuc – Pulsar Point – CeraBall/Vibrapod. CeraBase gives the improvement similar to the Ceraball, but maybe not as striking. Interesting was the reaction to those “revelation” of Mr. Waszczyszyn, the constructor and owner of Ancient Audio, a man, that I would not suspect of exaltation or wackiness. This is a flesh and blood engineer, listening and hearing, but devoted to undisputable facts. Hard facts. You should have seen his face when I collected the toys (Cera are packed in nice cases). The question he asked is probably a basic one for a constructor: “How should I now listen and tune my devices? If such an insignificant thing changes the sound is such a fundamental way, then what should I take as a reference in designing? I hope that you now understand the problems of a constructor, that encounters dozens of such things, and those are only the small things!” That is the reason, that is worth to try isolators in your own system by yourself. Just please remember, that the system must be well composed before you try. It is not even about the cost, but about synergy – the supports will not correct the errors but can expose the assets.
|
||||
| © Copyright HIGH Fidelity 2007, Created by B |
