stereo system evaluation--objective or subjective

is there a valid objective approach for evaluating the sound quality of stereo systems, or is it purely a matter of taste ?
I'll bet money that you are not really looking for an answer to this question. You already have a position on the topic you ask about and are not going to change it. This is just one of those questions designed to get everyone arguing for no reason. Nice try. Go back to Computer Audiophile.
hi don_c55:

there are no stupid questions only stupid answers.

secondly, most people evaluate stereo systems by listening.

wouldn't it be instructive to find a method that is independent of the listener ?
A persons taste has a lot to do with what sounds good to them.
Everyone goes by listening to come up with what merrits as sounding good to them.
I listen to many systems from belonging to an A'phile club here in central Florida. Most systems fall into a slot of acceptable sound to meet my taste. Others, though fitting the bill to the owners taste is outside of my own.
By hearing so many systems on a monthly basis I have come to appreciate other peoples taste that don't match my own.

If there was an objective approach for evaluating a system, it would fall short to someones taste.
Some like it warm, some like it neutral and some like it very lively.

Good topic Mrtennis.
Stereophile has technical measurements to go along with their subjective reviews... they've had it for years. And it's interesting that the reviewer's subjective views don't always align well with the objective measurements.

Let's face it, there are just too many variables. The component under review is part of a system where many complex interactions take place. When the component is sent for measurements it is not part of the same system (being measured on the bench with the technician's test equipment).

Regardless of that, if you know how to interpret the data it will provide some useful insight into the performance of the component under scrutiny.

As an audio reviewer, my particular publication does not have the facility to provide test data. So to double-check my own impressions I have two different reference systems in my home. That way I can try the review gear in both systems (with different associated gear) and see what characteristics remain constant... and also see if any outliers pop up. Often times I find that many of the sonic characteristics of the component remain constant from one system to the other... though there have been exceptions.

For whatever reason I find that system cables are the most unpredictable of my components when going from one system to the other and for me are more difficult to pin down and review than the basic hardware like amps, preamps, and DACs.

Still, I know I can depend on certain cables to do specific things and I can use them to change the tonality of a system more to my liking... kind of like an equalizer. And I'm aware that many audiophiles view system cables as "tone controls."

The simple answer is that there are entirely too many variables to find a system of quality assessment that would satisfy every listener. In my own small group of audio friends their taste in sound is all over the map and runs the gamut from the accurate and precise to the mellow and ill-focused. It seems that a lot of folks focus on different areas of performance that are important to them at the expense of other parameters.

I personally strive for neutral and natural and try to get the most detail possible without losing sight of musicality. And even this description will mean different things to different people.

Also, if you had a system that measured flat +/- 0.5dB 20Hz to 20kHz it would sound unbearably bright and unlistenable to the vast majority of audiophiles and music lovers. So maybe the thing to pursue is which curve or amount of high frequency roll off starting at what frequency would provide the ideal musical sonic balance to the highest percentage of listeners??? We can save matters of dynamic compression, transient speed, and smoothness and transparency for another time.

Frank A
Also, if you had a system that measured flat +/- 0.5dB 20Hz to 20kHz it would sound unbearably bright and unlistenable to the vast majority of audiophiles and music lovers.

Why? Do you suggest that live concerts (20Hz-20kHz +/-0dB) sound unbearably bright?

Yes, well, by the time the "Live Sound" in the concert hall reaches the listeners in the audience the high frequencies are definitely rolled off. And they are rolled off more in some concert halls than others... so which concert hall (or amount of hf roll-off would you pick?).
Kijanki, I guarantee that if you try to EQ your system so that it's flat out to 20kHz at your listening position it will sound VERY BRIGHT and you won't like it...

Don't believe me? Simply try it and see...
Great answers, save for the first.

I wouldn't dream of reinterpreting the great answers so far but would like to second the thought that objectivity is oh, so subjective. As in other fields of art, one can only judge based on experience, tastes and biases.

What I used to do when younger, when possible, was to glean as much info and tastes of a reviewer and compare it to what I heard, in a general way (much the same way as movie reviewers who I came to trust). Granted, it was much, much harder to audition the same components but aural clues figured large in listening appreciation as I learned the path to this hobby.

In the end it was every man for himself as I learned that what I liked, others didn't. What I disliked, others did. Consensus was akin to herding cats.

All the best,
"Herding Cats," yes Nonoise, that's an excellent analogy and I dare say, "poetic" as it applies to audiophilia.
Plato, There are many speakers that measure pretty flat to 20kHz and are not bright at all. Many headphones have flat response to 20kHz without "unbearable" brightness. Frequency response is not a very good indicator of the sound since it doesn't show phase shift, affecting summing of harmonics, or effects of Transient Intermodulation that can produce a lot of odd harmonics responsible for system brightness. There are also tweeter's distortions at its lower end or 8-11kHz where sibilants are. It is very complex issue and frequency response alone is not the best indicator of the sound. I cannot even hear above 16kHz where my speakers measure in the room with CD test tones at about the same level as at 1kHz (Radio Shack meter at listening position facing straight between speakers, response corrected)
Kijanki, I think you just want to argue with me for the sake of arguing so this will be my last reply to this thread. The simple truth is that although many speakers are reported to have highs flat to 20kHz, in actual use in an actual listening room the sound will be decidedly rolled off at that frequency by the time it bounces off walls, carpet and furniture, and goes through door openings. This, not to mention that unless the tweeters are aimed directly at the listener you're not going to get anywhere near flat 20kHz response.

Plus I wouldn't call the Ratshack meter anywhere near precise.

But hey, it's your right to believe what you want to believe. I just don't happen to agree with you.

Saying that I argue for sake of arguing is rude, therefore I don't care to discuss anything with you either.
i have a reason for asking this question.

it would be useful to have some set of measurements that could assist in discriminating potentially poor sounding, from potentially good sounding components, as it is impossible to audition many components.

there is always the possibility of missing out on a decent piece of equipment because it is inaccessible. as well, one may waste ones time auditioning a component that does not match well with the other components of a stereo system.

it would seem that there is no algorithm available to assist in the process of increasing the probability of eliminating "unwanted" components and ferreting out those which have promise.
evaluation is subjective
performance is objective

Measured performance if done correctly should be a factor in evaluation but that alone will not likely determine results of individual human evaluations which in the end is all that matters.
Taste? No, we all hear differently so in the end 'ears' rule. A while back someone did a study on identical twins and found even they had hearing differences which were more environmentally created. Add into the mix that different cultures and those speaking different native languages hear differently, e. g. 'the British sound'. Accent matter.
Isn't it interesting that some British recording artists lose their accent while singing and some Americans gain one :)
Or is it just my ears?
"wouldn't be instructive to find a method that is independent of the listener?"


How do you eliminate the listener, idiot?

That would mean removing subjectivity and using only technical data. I wonder to what extent design engineers use their ears. There is a fairly large margin of predictability towards a particular desired outcome but I would think a lot of trial and error goes into prototyping any piece of gear before a final rendition is settled on. So I guess I've answered my own question.
The only logical way to evaluate a TV is look at the picture and compare it to the picture of other TVs sitting next to it, no? Why would evaluating anything audio related be any different?


That's true, but evaluating TV picture is not that simple. Where it comes to sharp/soft or cold/warm static picture anybody can see it right away, but when the same picture looks very different at 240Hz refresh rate it become complex. My TV has this ability and I hate it. Everything looks like home video made. Somebody must love it, otherwise they wouldn't sell it. I also remember TV set that was sharpening slowly when picture got still (faces). It looked very weird in the store but customer I talked to couldn't see it. My Samsung LED TV has edge lighting that supposed to be very uneven - I cannot see that. My TV picture is almost perfect to me but some people, I talked to, believe that analog technology was better. Sound is perhaps even more complex.
again, nobody addressed my point:

are there any specs that can be used to screen components

are there any parameters which could be used to discriminate between components that one might want to audition from those to avoid ?

for example, there are many preamps out there. it is impossible to listen to all of them. it may be possible to establish a probability that, given criteria, one could feel confident that it is worth listening to some, or , perhaps, it's hopeless.

The answer is no. Technical measurements are objective but often gear that measures best sounds the worst.
Stereo system evaluation -- objective or subjective Yes,it's a combination of both.There should be enough information in the above posts to answer the complicated yet very simple question.