How do you prove that one speaker is better than another?

@cal91,

Yes, although I would be concerned if a loudspeaker with an exemplary measured response sounded awful, I think I would eventually do the same.

Probably after spending an awful long time trying to get to like the better measured one first! 

Now there's another conundrum, given time, can you learn to love bad hi-fi? 

kenjit OP1,045 posts08-31-2020 4:44am

What about you? How do we know you are perfect? Because your opinion of yourself tells us so?

Because I have already been in the predicament that most of you on this forum are currently in. I was once an audiophile that sought perfect sound but no speaker was ever good enough. My standards were too high. However now I am what you might call a super audiophile which is the next level up from an ordinary audiophile. This involves severe modifications, custom tuning, hand tuning and tuning by ear. Most of you are not at this level yet and some of you may never reach this point. I am a self taught speaker tuner that strives for perfection. That is what makes me special. I do not pretend that something sounds perfect when its not. I will die a hero that the audiophile world has never seen the likes of.


This could well be the greatest exchange between any 2 audiophiles ever.

@mozartfan,

’Look Sterophile past 20 yrs+ and other hifi mags, have started this thing, pumping speakers in such a language where buyers really aren’t, sure if the review is accurate.’


That’s the problem the OP is trying to address I think.

How do we measure the playback accuracy of any loudspeaker?

How about getting a loudspeaker to play back carefully recorded test tones (or even piano notes) with weighting given to the Fletcher-Munson curves?

Maybe you could then measure the playback accuracy with highly accurate measuring equipment, repeating the process at several different volume levels to see if the accuracy increases or decreases accordingly.

Tests might reveal that one design does better at lower volume than another, but worse at a higher level. That information would be valuable depending upon the playback volume level required.

The sheer amount of work involved makes this endeavour unlikely to be done privately, but a big magazine might have the resources, if not the desire or financial inclination.

I’m also guessing legal concerns would stop Harman Kardon from ever publishing the results of their innumerable blind listening tests over the years. Would be a great read though.

It is kind of frustrating that we’re denied this in audio, as comparitive testing is accepted more or less everywhere else.

You can find comparison tests on everything from smartphone battery life, PC boot up times, the efficiency of cooling fans, 0-60 car acceleration times, fuel efficiency data etc.

Maybe we need the audio equivalent of YouTuber Project Farm? Why couldn’t he have been an audiophile?

https://www.youtube.com/channel/UC2rzsm1Qi6N1X-wuOg_p0Ng

@cal91,

'What if the speakers were proven to be the best, but you don't like the way they sound? You prefer a set of speakers that are according to the measurements not really that good. But to your ears they sound magnificent. Which speakers would you buy?'


Be careful there, you are now approaching the audiophile equivalent of the crossroads where Robert Johnson allegedly sold his soul to the Devil.

This is where you have to decide which camp you want to set foot in, the objectivists or the subjectivists?

Actually, either choice is fine just as long as you are able to acknowledge the grounds for your decision, rather than insist it's the only viable one for everyone. 

Lesser experienced would be audiophiles, or those who don't fully trust their ears, might prefer instead to place their trust in science whilst more experienced ones may be more partial to their own particular choice of sonic poison, ie desirable colorations.  

Whichever way you choose it doesn't alter the fact that one loudspeaker measured far better than the other one.

'The measurements are indisputable.' 

@kenjit,

Yes, although I would be concerned if a loudspeaker with an exemplary measured response sounded awful, I think I would eventually do the same.

’impossible for an exemplary measuring speaker to sound awful.’


I would like to think so too.

I’m pretty certain so would the designers at B&W, Tannoy, Harbeth, Magico, PMC, Wilson, Sonus Faber, ATC, and especially you’d think at Harman Kardon/ Revel/ JBL etc.

They’re all doing their best to produce what they regard as the highest quality loudspeakers (to a price) with some of the most advanced testing facilities and software that they can get access to.

Unfortunately the only people doing comparitive tests are reviewers, most with conflicting interests.

Even worse, a few of them might even discreetly be in the employ of certain manufacturers.

A far from ideal state of affairs.

Hence the need for the OP to ask the original question. How can we know? Who can we trust?

Perhaps someone with a background in film or broadcasting work could chime in to explain how large organisations decide which loudspeakers to use for production work?

The days of buying subjectively are drawing to an end. 

This is the age of almost limitless information and we have specs for literally everything. You can't find a smartphone review without being bombarded by various specs. Same for cars. Same for washing machines. Same for hedge cutters. Same for lawn mowers. 

So why not loudspeakers? Is it really a good idea to spend thousands after just a casual listen in what might be less than optimal conditions?

Performance counts and loudspeakers are not exempt. The coming generation will laugh when it reads that we used to buy them by listening alone.

The only ones I ever bought by listening alone were my first ones. Every subsequent pair needed increasing amounts of research beforehand.

Any savvy designers out there will need to get ahead of the game and start dishing out all the specs they dare to.

Me, I'm definitely in the Steve Jobs camp of research when it comes to buying new stuff.

What's more, it seems to work.

https://www.google.com/amp/s/9to5mac.com/2011/03/07/how-steve-jobs-picks-a-washerdryer/amp/

@bluemoodriver,

Wise words. We're nowhere near anything like a perfect loudspeaker.

Still, there's nothing wrong with knowing as many performance stats as possible. 

You'd do the same before buying an Audi, BMW, Lexus, Mercedes, Porsche or Tesla etc, wouldn't you?

In fact you'll have to look at all the data/stats for an all electric vehicle before very long - whether you want to or not. 

I think the point the OP is making is that it's no longer viable for manufacturers to promote listening as the primary means of establishing the superiority of one design over another.

How on earth are we supposed to make progress if we carry on like that?

Sure it's a viable approach if we're only planning to buy an all in one system from Walmart etc.

However, since we're audiophiles striving for a far superior performance, we hope, then some form of measurement and data might be more appropriate.

Especially if we're planning on spending more than $100 on a system. In any case it's easy enough to get a refund or an exchange from Walmart or Amazon etc. No serious risk of lasting buyers remorse.

On the other hand if I'm planning to spend over $2000 on a pair of loudspeakers then I'd expect them to measure well, wouldn't you?

I'd also want to know exactly where the weaknesses were so in order to prioritise them. Unacceptable ones might include excessive sibilance, a bass response that starts to fall off a cliff at around 50Hz, a sawtooth frequency response, a suckout between 4-7kHz, an excessive cabinet contribution or transparency, an overly focussed sweetspot, driver dispersion anomalies etc

All of those things are far from easy to determine by just listening alone, even for the most experienced amongst us. You'd need a lot of time and a carefully selected range of music. Good powers of concentration and a decent musical memory would help too.

Surely at the very least, auditioning has to be a combination of both listening and data, just as the BBC did with their perennial favourite, the LS 3/5A. A design popular to this day and one that measures extremely well within its operating requirements.

If it's only to be one or other, listening or measurements, then I'd go with the data.

I think I could put up with a loudspeaker that measures extremely well.