Reply to thread

Message

<blockquote data-quote="GordonW" data-source="post: 2250275" data-attributes="member: 568981">Most people seem to misunderstand the applicability of an ABX test.Even David Clark, the INVENTOR of the modern ABX test, had to admit to Gary Galo at the 1991 AES meeting, that an ABX test CANNOT rule out differences between equipment... only that the ABX can only CONFIRM a difference. In other words, there is NO WAY an ABX test can prove a negative result. These are FUNDAMENTAL LIMITATIONS of the ABX methodology.There are quite possibly differences that are beyond the scope of the limited test specifications. These limitations of ABX testing have been known in audio engineering circles, for at least the last 16 years.Besides, the &quot;apples to apples&quot; test requirements fly in the face of reality anyhow. NOT ALL AMPS actually HAVE the same frequency response... and once you put an EQ in line with one amp to compensate for another, all bets are off anyway, due to the addition of the &quot;sonic signature&quot; of the EQ itserlf. And ,the difference between noise floor behaviour is not something to be IGNORED, but one of the PRIME REASONS people pick one amp design over another.In addition, the &quot;no clipping&quot; requirements of the ABX test methodology mentioned above, fly in the face of VIRTUALLy ALL real-world music listening. Unless you have THOUSANDS of watts and/or VERY EFFICIENT speakers (ie, 95 dB sensitivity or higher, you WILL instantaneously clip MOST ANY amp on transients. Most &quot;live&quot; music (ie, recordings without excessive compression) have 30-40 dB &quot;crest factor&quot;... ie, if you&#039;re listening to an average of 80 dB, you have PEAKS of 110 dB or higher. It&#039;s not PERCIEVED to be that loud on average... because, the average level IS only around 80 dB. But, the AMP sees that 110 dB peak as 1000X the power demand, of the 80 dB nominal level... and voila, momentary clipping.In short, one of the fundamental problems in modern society is the tendency to over-simplfy issues... which is bad enough when it&#039;s simply politics or day-to-day social behaviour. But when such over-simplification invades science, it&#039;s a bad day for science...Regards,Gordon.</blockquote>

[QUOTE="GordonW, post: 2250275, member: 568981"] Most people seem to misunderstand the applicability of an ABX test. Even David Clark, the INVENTOR of the modern ABX test, had to admit to Gary Galo at the 1991 AES meeting, that an ABX test CANNOT rule out differences between equipment... only that the ABX can only CONFIRM a difference. In other words, there is NO WAY an ABX test can prove a negative result. These are FUNDAMENTAL LIMITATIONS of the ABX methodology.There are quite possibly differences that are beyond the scope of the limited test specifications. These limitations of ABX testing have been known in audio engineering circles, for at least the last 16 years. Besides, the "apples to apples" test requirements fly in the face of reality anyhow. NOT ALL AMPS actually HAVE the same frequency response... and once you put an EQ in line with one amp to compensate for another, all bets are off anyway, due to the addition of the "sonic signature" of the EQ itserlf. And ,the difference between noise floor behaviour is not something to be IGNORED, but one of the PRIME REASONS people pick one amp design over another. In addition, the "no clipping" requirements of the ABX test methodology mentioned above, fly in the face of VIRTUALLy ALL real-world music listening. Unless you have THOUSANDS of watts and/or VERY EFFICIENT speakers (ie, 95 dB sensitivity or higher, you WILL instantaneously clip MOST ANY amp on transients. Most "live" music (ie, recordings without excessive compression) have 30-40 dB "crest factor"... ie, if you're listening to an average of 80 dB, you have PEAKS of 110 dB or higher. It's not PERCIEVED to be that loud on average... because, the average level IS only around 80 dB. But, the AMP sees that 110 dB peak as 1000X the power demand, of the 80 dB nominal level... and voila, momentary clipping. In short, one of the fundamental problems in modern society is the tendency to over-simplfy issues... which is bad enough when it's simply politics or day-to-day social behaviour. But when such over-simplification invades science, it's a bad day for science... Regards, Gordon. [/QUOTE]

Verification