We need to talk about SPEAKERS

[Grahame]

Interesting article on El Reg http://www.theregister.co.uk/2014/07/02/feature_the_future_loudspeaker_design/

 

We need to talk about SPEAKERS: Soz, 'audiophiles', only IT will break the sound barrier

 

Design, DSPs and the debunking of traditional hi-fi

 

 

Covers a lot of ground, the importance of time domain accuracy, and the relationship to the human hearing system.

 

Some of the ideas may be applicable to headphones (eh, Tyll?)

 

Author seems to have the credentials.

 

Sound opinions, or just clickbait? (the commentards seem to have bitten)

 

Discuss

 

[RudeWolf]

Well, the clickbait does run strong throughout the article, but it isn't out of tone with the rest of what Vulture puts out.

 

That being said I agree that much of the speaker design is still stuck in the seventies. It is strange that with all the yammering in the article about the time domain accuracy not once was analyzed the actual enclosure geometry. I do think that speaker designers have to employ computer assisted analysis to keep their birdhouses up to date.

 

As for DSP assisted sound production - I still remain skeptical. In two weeks I'll try to audition what Devialet has done with its SAM technology and what can it do for the already brilliant KEF LS50. In addition to that I have some of the best DSP guys operating in my country - Real Sound Lab and SonarWorks. I've been trying to set up an interview with them since forever as they claim to be developing an app based solution for smart devices that does headphone correction. I've heard what these guys can do with two compression drivers mounted inside steel buckets and it was pretty impressive.

 

Yet I feel that we mustn't overestimate what DSP can do. I'm pretty sure the physical speaker design very much persists in the equation. Maybe you could design for DSP and get better results than using traditional methods. Who knows...

[nikongod]

There is no real benefit to speaker MFR's to follow current trends. The reason is simple. 

 

By following current trends/best practices in designing "objectively better" speakers (computer modeling, measurement systems based on computer guestimation, digital crossovers, etc) EVERYONE will arrive at a largely similar end result - which I'm not entirely convinced would add significantly to the cost of the thing. At that point, what differentiates one guy from the next? NOTHING! This quickly turns into a race to the bottom, which favors large companies who can compete on economics of scale, and all of these small to medium sized MFRs are left with no business. 

 

The best thing for a small-medium MFR and the industry that supports them to do is accept and embrace that every speaker sounds sooooooooo different that you might even think you were listening to a different recording by simply switching speakers. After this truth is accepted the little MFR can sell an obviously fucked up speaker to some guy who FINALLY got tired of the particular way that his old speakers were fucked up. A few years down the line he will get tired of these speakers and get another set of differently fucked up speakers, but so long as the money keeps flowing...

 

Nobody is enraged by the fact that they never heard the saxophone in their favorite recording on their old speakers. They can hear it on the speakers they have now (but not the trumpet) so these speakers must be better. They are delighted!

 

Its not that speakers suck (even though they do) the same thing happens in headphones. Every brand has "their sound."