|
General Asylum General audio topics that don't fit into specific categories. |
|
|
|
General Asylum General audio topics that don't fit into specific categories. |
|
In Reply to: Speaker Measurements vs. Subjective Listening posted by lenw on February 18, 2007 at 11:26:57:
"How much attention should we pay to speaker measurements when they're provided in reviews such as Stereophile?"I would pay more to speaker measurements in magazines just like Stereophile than from those from manufacturers. Why? They do their testing the SAME WAY for EVERY SPEAKER and use the SAME TEST GEAR. This is, quite simply, the only conceivable way to make any kind of "paper comparisons" between something as subjective as how a loudpeaker sounds or "voices". Only an experience speaker designer can cut through through the marketing haze and see a lot of these measurements for what they really are. There are many different ways to present measurements so that they show the speakers strengths while concealing potential weaknesses.
I noticed that on of the criterion for the "ideal speakers" was that they need to be "time coherent". This is interesting. Does this mean that the music, when played back, will occur at the exact same time as it was recorded? No. I believe this is a "buzzword" mixture of two OTHER criterion: a)time aligned and b)phase coherent.
Physical alignment is easy. Align the drivers so that the "time of flight" from each of the drivers acoustic center to the listening (or test) position is identical. Time alignment is another story. In addition to compensating for driver physical locations, you may also need to have additional physical offsets for time delays associated with crossover topologies that are not transient perfect. A speaker that has all speakers with zero-offset with respect to the listening position (and no additional time alignment) MUST have a transient perfect crossover to be considered "time aligned". (Basically, lining up the acoustic centers of drivers does NOT guarantee time alignment.)
Now, phase coherence on the other hand is a subject of much debate as to the audibility of a non-constant group delay. The "group delay" (rate of change of phase with respect to frequency) is only constant for a 1st order Buttworth crossover (and will remain so only if the ACOUSTIC response is not a higher order). With a constant group delay, a single physical offset is capable of compensating for crossover delay. With a non-constant group delay, physical alignment delay compensation for is only valid at the crossover point, and becomes "less valid" as you go further away from Fc into the remainder of the crossover overlap region. (This is my understanding of it anyways).
If indeed your requirement for phase coherency was an absolute, we would probably be looking at about 1% of all speakers ever made. Don't forget that we can't just eliminate all crossover orders higher than 1st order electric... we also need to get rid of speakers that do not have a 1st order ACOUSTIC response. (Even speakers with 1st order electric crossovers cannot change the fact that drivers have 2nd order rolloffs at their frequency extremes. This is why a 1st order electric filter can result in as high as a third order acoustic rolloff).
So. What does all this mean? It means that if Stereophile says a speaker has a strange anomaly in the frequency domain, you might want to wonder why. But what happens if Stereophile reviews a speaker with a higher order crossover that is NOT phase coherent but has excellent transient response due to good design and they think it sounds incredible? Should you write that speaker off of your list and consider the boys at Sterephile to be liars? Or should you listen for yourself and decide how important crossover topology is?
Let's face it. If you are going to buy speakers based on design philosophy, you either know enough to be designing and building your OWN speakers, or you are simply caught up in the highly subjective realm of conflicting design ideals. There are no "Ideal" loudspeakers anyways (despite marketing claims to this effect). Different configurations and sizes of drivers and crossover topologies each have their own set of pros and cons. 1st order acoustic designs are novel indeed, but are by no means a panacea of production based on this single design criterion alone. In fact, quite a few "concessions" need to be made to MEET this incredibly difficult ideal.
To be honest, if you are designing and building your own loudspeakers, measurements are critical to stay on the path of the chosen design philosophy. But if you can't design speakers from scratch yourself, gettting hung up on measurements will likely only result in you making "snap judgments" that may be based on a misinterpretation of the data. The fact that manufacturers do not conform to a standardized test regimen for their products adds a whole other dimension of complexity to this.
I say stick with a/b comparisons of speakers that you like. Make note of what happens when the switch is made. Did something change for the better? Did something bad show up? Did something good become "less good"?
So. How much stock should we place in reviews? Although measurements are a good indicator, so are the subjective analysis of the reviewers in a lot of cases. Sometimes what the reviewers report is bang on - all these guys DO is listen to speakers, and they are often very good at using subjective terminology to describe what they are hearing. If 5/6 reviewers say that a speaker has a "bloomy" or "bloated" bottom end, maybe this is true. In any case, listen FIRST before you READ. If you read reviews FIRST you might influence your own perception by "projecting" what the reviewers said with what you are hearing. Your brain will be looking for similarities and differences between what you are hearing and "stored data" about what someone ELSE had to say. This is counter productive. You should be only listening to what YOU are hearing and not referring back to what was said by others. This extra "processing" will only serve to cloud your perception further. That is just my opinion. If my friend tells me to go listen to "awesome" sounding speakers I am walking in EXPECTING awesome sounding speakers, and expectation bias can be more of a factor than a lot of people around these parts will readily admit.
Just some food for thought...
One problem with a/b comparisons is that some guys go through great lengths in demos to ensure the speakers have the same setup position, and even move the speakers for each iteration. This makes sense from a testing philosophy (implement as many controls are you can to validate the comparison) with one rather large caveat emptor: who is to say the both speakers operate BEST in the that particular position? :o)
Kinda complicated isn't it? It really doesn't have to be...
Show your dealer you are serious. Tell him if he spends the time showing you a number of different speakers and compares your final "two best" for you, you'll make a purchase at that point. Ignore measurements and just find that speaker that does most justice to the music you are most familiar with. Tell him your budget constraints going in so he does not throw something into the mix that you can't or won't afford.
Falling in love with a $100,00 speaker you can't buy is just a wee bit masochistic, now isn't it?
Hope this helped.
Cheers,
Presto
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Here is my take on it... (long) - Presto 16:14:57 02/18/07 (0)
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
