Comment by Groxx | Hacker News Reader

Comment by Groxx | Hacker News Reader

Parent Full thread

>A few years ago, audiophiles dismissed iTunes’ 128-Kbps resolution as anemic, even though it supposedly passed rigid blind testing against full-bandwidth CD tracks of the same song. The sound is compressed, connoisseurs said. The high end is mangled. Good work, audiophiles: Online stores have cranked up the audio quality to a fat 256 Kbps. To most ears, it’s indistinguishable from a CD. (Actually, most ears are listening through crummy earbuds anyway, but whatever.)

There is so much wrong with those sentences, I'm not sure where to start.

1) Who the hell did they test their 128kbps on? I can tell the difference, even with crummy earbuds. With my Grados, 192 is still distinguishable, and I'm far from skilled in audiophile terms.

The rest of the article seems to miss a major point in this whole affair: how we got here.

Thank you, pirates and audiophiles, for allowing me to buy high-quality, DRM-free music. Could you go after the video industry now, pretty please?

My theory is that these blind tests are using a random sample of music, unfamiliar to the person being tested. It is not a random pop song that I care about sounding good, it is my favorite music that I have listened to 100 times. For these tracks, picking up distortions is easy and it really ruins the experience.

You have Grados, that puts you in the top 1%.

I can clearly tell the difference between most 128kbps MP3s and CDs on my stock (bose, ugh) car stereo.

iTunes sells AAC, not MP3, as far as I know. My _opinion_ is that 128kbps AAC is about the same quality as MP3@V2.

I don't buy music from iTunes because AAC isn't compatible with all my devices.

It is true as you say that other CODECs are "better" than MP3.

"""I can tell the difference, even with crummy earbuds."""

It's not enough to be able to "tell the difference". You also have to be able to tell the difference in random A/B tests.

If you are like 99.9 percent of the population, you will not be able.

A test is reproductible. Your anecdotal testimony means nothing at all.

Random A/B with different songs in A and B: nobody can. You can get a statistical success off raw sound "quality", and nothing else. What if they played Lo-Fi music on one?

Random A/B with the same song, one CD and one 128kbps MP3: not a problem identifying the two with iPod earbuds. Even if I'd never heard it before. With good headphones, the difference leaps out, it's utterly unmistakable.

Reply on news.ycombinator.com