Hacker News Reader

Comment by smackfu | Hacker News Reader

Is it better to test audio products with a meter or with double blind humans?

I've been convinced by a knowledgeable person[1] that measurements are the only useful way to measure performance. To paraphrase poorly: subjective reviews are useless since they are limited by the reviewers subjective perception, their ability to turn that perception into words, and your ability to turn those words into a concrete idea about how something sounds. Since there is no part of a speaker's performance that cannot be precisely measured, the best course of action is to learn how each measurement affects your perception and what qualities you like in a speaker. Then you can fairly accurately evaluate if you would like a speaker with just a few graphs.

(Of course this is completely impractical for almost everyone, but I'm talking about some platonic ideal of audiophile)

1: http://zaphaudio.com/

I know from experience that different headphones with similar frequency response can sound radically different. Most reviewers don't measure anything else, so that doesn't really help.

Have a look at this: http://www.nutshellhifi.com/library/FindingCG.html It applies mainly to amps, but when you learn about amps you start to realise that loud-speakers are really all part of the same big circuit.

With amps, specs only take you so far, once you get above the absolutely dirt cheap crap it doesn't really help. Using THD as one example, above 1% might indicate something wrong with the amp design... bellow doesn't really make any difference because these are just tested with sine waves (do you listen to pure sine waves? no), very low might even indicate that the design has been compromised just to get good paper specs.

Headphones and speakers can be similar, frequency range doesn't really tell you anything unless they are severely limited and thus it only really helps sift though crap.

You are correct about reviews though, it really depends on who is reviewing it, and 99.99% of the people out there have hardly anthing worth comparing what they are reviewing it to. If i've never had a pencil before, and I get a pencil and it's all crumbly I will be like "It's frickin amazin, it makes marks on paper", but that doesn't really help anyone else trying to make an informed decision. Additionally, even if you start going through audiophile forums and vet authors for their background knowledge, _even then_ you have the aspect of personal taste, musicality etc.

So yeah... if you want good audio, your fucked :) the only way is to try. And try not to go crazy. But not buying stuff with any integrated digital nonsense is a good starting point.

Depends on if you want factual technical measurements or subjective human opinions

If it's double blind isn't it objective?

Only if all humans have the same opinion about what good audio sounds like.

It's as objective as a subjective measurement can get.

It's not repeatable, and I don't know what units of measure could be used either.

Reply on news.ycombinator.com