Home
AudioAsylum Trader
General Asylum: Too Generalized... by Todd Krieger

General audio topics that don't fit into specific categories.

For Sale Ads

FAQ / News / Events

 

Too Generalized...

129.239.26.4


[ Follow Ups ] Thread:  [ Display   All   Email ] [ General Asylum ]
[ Alert Moderator ]

"I have seen a number of posts/responses over the past several weeks referring to questions about one tweek or another and whether these have been proven to work. First off, I don't believe that its my job to prove with a scientific certainty that something will or won't work for individual systems. There are too many variables to predict with any certainty that someone will or won't notice a change (let alone an improvement) if they try this or that. The information offered by me and others is being offered to the group to try on their own and see if they notice an improvement or change."

I've personally wondered why people get hung up over audible difference having to be quantified, unless he has no faith in his own hearing perception...

When I hear a difference, I really don't care if it can be measured, unless it can somehow indicate which sample is the more-accurate one.

Conversely, I will often try to pick up measured differences in my listening. Sometimes I can, sometimes I cannot.

"The reason for this post is more about certain attitudes being displayed by the "scientific" crowd that believes that unless something can be reproduced in some testing methodology, its not valid."

That is only an issue if what you're supposed to quantify is a requirement given by a customer. Such as flight performance or safety requirements of an aircraft... Then it becomes a big deal. For audio, the only requirements that should be measureable are ones to assure the unit won't hurt the user or his associated equipment in normal use. And the fact the normal specs are at least reasonable.

"As I have state before here, I come from the sciences background (Chem E)and understand the scientific methodology; but I also understand its limitations. That doesn't mean I don't try to use science in my design or testing; what it does mean is that I will rarely try to prove something with scientific certainty. There is a huge difference between the two."

The problem is in audio, there is very little correlation between subjective sonic preference and measured performance. It's nice to have equipment that excels in both domains, and I often use it as a factor in shopping for specific equipment.

"What the proponents of the scientific method often fail to realize is that the testing methodology itself will sometimes affect and predetermine the outcome of the phenominium being tested."

Case and point- I've determined over the years the impact test hardware has on the sonics of an audio system. Many test methods involve adding hardware, which I adamantly believe obscures subtle differences between electronics and cables. Or in other words, the "sonic signature" of the test equipment gets interjected into the system.

"This is one of the reasons you will sometimes get different results doing the same test in controlled environments. Its also why I dislike double blind testing in audio applications. In a double blind test the participants rarely know the system or the room and haven't fine tuned their ears to its particular sonic characteristics."

That, plus what I mentioned above... Plus the fact in the case of electronics and cables, I have **never** been able to come to an accurate conclusion with **only** short term listening- I am often impressed with something at first-listen, but cannot stand the product listening long-term. Most "audio tests" require omission of long-term testing, which I say is **the** biggest flaw of controlled audio testing.

"Also, added to that the interaction between people and the impact of their physical presence in a room and I have no difficulty believing that they will not hear a difference."

Now *that* I would call a stretch...

"It would be difficult to do otherwise. Do you wonder why the majority of audio engineers couldn't hear the impact on the music of a watermarking schema proposed in the mid 80s that was later decried by the audiophile community? Take a look at the test methodology and you will see why."

I'm not familiar with this "watermarking schema"...

"I vastly prefer to have a piece of equipment (or a tweek) tried in various peoples systems. They know the sound of the baseline (their system) and can more easily tell if something changed or not. Assuming I've picked knowledgeable people, I will get useful feedback from these individuals which tells me something about how the equipment interacts with other equipment and rooms. I will get far more useful information from this than any sort of double blind panel. More importantly, I am more concerned about people's reaction to something than in validating it scientifically."

This is a double-edged sword. For example, you may have two friends with 100 dB/m efficiency 8-ohm horn speakers, and tell you this five-watt SET amp is fantastic. If you are not aware that specific types of amps have certain interface requirements and limitations, you may conclude that one of these things would be perfect for your 82 dB/m 3-ohm power-hogs... Science may not be the end-all, but it should not be totally disdained either.

"I could go on and on about the deficiencies of how people have chosen to apply the scientific method to problems in audio, but suffice it to say that much of this has been pysudoscience."

Maybe in regard to subtle sonic differences, but definitely not for addressing component interaction. Scientific ignorance is often the cause of people frying their audio systems... (And in extreme cases, burning their houses down...)

"Although some chemists and physicist would argue, math is the only true science. When it comes to audio, the scientific method is even less reliable since we do not have precise language to ascribe what is being heard, we know very little about how we hear, what the human hearing is accutely attuned to, and sound measurements correspondance to human hearing."

Some scientific measurements, such as THD and damping factor, have little relevance to perceived sonic quality, but I would not ignore a speaker frequency-response curve that has major aberrations, even though one may be initially-impressed with the sound. More-often than not, a "suckout" in a speaker's frequency response will eventually not only get noticed, but will often drive the listener crazy once he's aware of the problem, but cannot listen around the problem.

"Bottom line, don't look for scientific certainty when it comes to audio. Take what is said in the spirit in which it is given- an observation by one or more people about something that changed their system for the better or worse. And whatever you do, ignore those who say something can't be because it is not scientifically proveable."

This approach should **not** be generalized. See my above comments when science is totally-disdained from the component-selection process. What you said is mostly-true with subtle changes in cables and amps, but can cause mucho problems if you happen to interface to components that are not compatible with each other.

"These are people who like to be content in knowing that something can't exist because there is no scientific study to prove that it does exist. These are the same people who have never heard the difference between an Onkyo reciever and a Jeff Rowland integrated because there are no measurements which show it so."

I would not really worry about such people- Many cannot hear differences, and I would not ridicule them for that. Remember- It's what *you* like what counts, and if someone else bashes what you like, for whatever reason, that's *his* problem... Too often, opinions of others end up making people change their systems for the worse.

And conversely those who perceive everything to sound the same have **their** favorite components too- And if **they** like it, opinions from you or me about how they choose their components don't matter either. It goes both ways.


This post is made possible by the generous support of people like you and our sponsors:
  Atma-Sphere Music Systems, Inc.  



Topic - Scientific Method Versus High End- Long (Rant?) - ggraff 07:52:59 04/8/02 ( 216)