What's a positive attribute?

rwortman -- Sat, 02/06/2010 - 22:28

I am specifically talking about the attribute reviewers refer to as "revealing". In every review I have read this is treated as a positive no matter what it is referring to. I disagree. If we are listening to a very high quality speaker that is capable of resolving differences in source components that other speakers cannot, it would seem that this is a very good thing. On the other hand, if two amplifiers sound wonderful driving these speakers and one changes its sound with every change of speaker wire and power cord while the other sounds the same no matter what we do to it, why would we conclude that the "revealing" amplifier is superior? Maybe the other one has a better designed power supply and output section that is more tolerant of load changes and power input. Similarly if we have two preamps or CD players that have equally impressive sound and one sounds different with every change of feet, or object you put on top of it and the other does not, does this make the former the superior piece because it is more "revealing". I think not. The better design should be more resistant to vibration and not have every component in it acting as a microphone. Perhaps this is valued because many reviewers and audiophiles are inveterate tweakers and components that don't respond to these ministrations are simply no fun. I would rather listen to music than fuss about with my system. I think that components should be judged on the basis of how they sound and their ability to resolve the details in the music, not on whether or not they respond to the popular tweaks of the day.

Tom Martin -- Wed, 02/10/2010 - 20:54

True enough.

I think the general use of "revealing" in our reviews is intended to convey the idea that the component under review is sufficiently resolving to show differences between ancillary components (as opposed to covering up the differences). As you say, a component could be hyper-sensitive to different ancillary components, but we try not to use "revealing" to indicate that. For example, when a speaker seems to be difficult to drive, we don't say it is revealing. We say it is difficult to drive. One standard for distinguishing between compatibility sensitivity and resolving power is to test whether the component reveals things heard previously or dimly. The opposite is also true: a product that tends to make all ancillary equipment sound similar is usually either low resolution or colored in some way.

Reviewers will use some test methods that are not intended to mirror the consumer's usage pattern. For example, we all have test discs that are demanding and therefore useful in revealing certain distortions. But we don't necessarily recommend these discs for your listening. An extreme example of this is the use of sweep sine waves to test for rattles and resonances. I'd be the last person to suggest that you listen to these for pleasure! And, yet, they are very efficient for finding out certain problems.

There is, inevitably, a gray area where compatibility issues could be confused with resolving power. So, the final and fundamental test is whether the resolving power goes hand-in-hand with delivering a sense of the absolute sound. High resolution is really high resolution only if it supports the goal of accurate music reproduction. Compatibility problems tend not to create musically consonant distortions.

I would say the hypothetical cases you describe should not be articulated, if they occurred in the review process, as "this component is revealing" because there is no musical benefit and therefore no real gain in resolution. Rather we should describe what you outline as "sensitivity to ancillaries."

CEO and Editorial Director, Nextscreen LLC

Keladrin -- Wed, 07/11/2012 - 08:16

Audio reviews are fraught with overuse of euphemisms and descriptors that are completely open to interpretation. Sometimes it is like the reviewer is describing a wine rather than music accuracy. If descriptor terms are used then these should be explained so we know the baseline otherwise the review reads like a poor work of fiction that means little to all apart from the reviewer and is about as much use as a horoscope. Examples are the descriptors: analytical, clinical, revealing, glassy, finesse, shimmering, sweet. These are nothing but meaningless colourful metaphors designed for dramatic effect. More objective terms like: dynamic, coherent, accurate, distortion free, immediate are a bit more meaningful.

The most useful terms (as long as they are explained to the uninformed) are: Good transients, accurate dispersion, low harmonic distortion, good dynamic range, accurate low-level detail, good frequency smoothness, accurate coherence, realistic volume dynamics, good bass extension, low hangover as they are linked to specific issues with all loudspeakers and how well they are addressed.

The dynamic range (Ratio of loudest undistorted volume to average background noise level) is very informative if we understand what it means as it is not just to do with the undistorted volume available but the quality of the low-level detail - this is probably what 'resolution' tries (but fails) to refer to in some cases. For speakers that have adequate maximum volume performance it is often the low-level detail (the ability to resolve the low volume portions accurately) that separates the best from the rest, despite common perceptions.

Now the term 'analytical'/'clinical' - what exactly does this imply? Surely something that resolves all the details and gives an accurate representation is better than the opposite? Surely a better term is 'accurate' if this is what is meant - this is rarely used. Loudspeaker comparisons are also a bit of a waste of time as there is no agreed standard or benchmark. Surely the benchmark should be unadulterated live music listened to in the sweet spot of well designed acoustic space? All loudspeakers will substantially degrade the signal in similar but subtley different ways so the last thing you want to do is use a speaker as a benchmark.

Why is the term 'accuracy' never used in any context? Something to think about...